Impartial Prioritieshttps://impartial-priorities.org/2022-07-16T00:00:00+00:00Chaining Retroactive Funders to Borrow Against Unlikely Utopias2022-04-19T00:00:00+00:002022-07-16T00:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2022-04-19:/chaining-retroactive-funders.html<p>There is no qualitative distinction between investors and retroactive funders on an impact market. Rather they will <em>de facto</em> fall along a spectrum of how altruistic they are. That is because investors will (1) expect investments into well-defined prize contests to be less risky than fully speculative investments, and will (2) expect more time to pass before they can exit fully speculative investments, so that a counterfactual riskless benchmark investment represents a higher threshold for them to consider impact markets at all.</p><h2 id="recap-impact-markets"><a class="toclink" href="#recap-impact-markets">Recap: Impact Markets</a></h2>
<p><img alt="Impact market" src="/images/toward-impact-markets/impact-market.png"></p>
<p>For a more comprehensive explanation of impact markets, see <a href="https://forum.effectivealtruism.org/posts/7kqL4G5badqjskYQs/toward-impact-markets-1">Toward Impact Markets</a>.</p>
<p>In short, an altruistic retroactive funder announces that they will pay for impact (or “outcomes”) they approve of. It resembles a <a href="https://forum.effectivealtruism.org/posts/2cCDhxmG36m3ybYbq/impact-prizes-as-an-alternative-to-certificates-of-impact">prize competition</a> in this way. But (1) they’ll pay proportional to how much they value the impact and not only the top <em>n</em> submissions; (2) the impact remains resellable by default; and (3) seed investors offer to pay the people who are vying for the prizes or provide them with anything else they need and receive in return rights to the impact and thus prize money.</p>
<p>It is analogous to the startup ecosystem: Big companies like Google want to acquire small companies with great staff or a great product. Founders try to start these small companies but often can’t do so (as quickly) without the seed funding and network of venture capital firms. When the exit happens (if it happens), the founders may not even any longer own the majority of the company because they’ve sold so much of it to the investors.</p>
<p>The benefits are particularly strong for high-impact charities and hits-based funders:</p>
<ol>
<li>If a hits-based funder usually funds projects that have a 1 in 10 chance of success and switches to retroactive funding, they save:<ol>
<li>the money from 9 in 10 of the grants,</li>
<li>the time from 9 in 10 of the due diligence processes, and</li>
<li>the risk from accidentally funding projects that then generate bad <span class="caps">PR</span>.</li>
</ol>
</li>
<li>Investors can thus speculate on making around 10x return on their successful investments, and they can further increase their expected returns:<ol>
<li>by specializing in a narrow area (such as <span class="caps">AI</span> safety) to make excellent predictions about which project will succeed,</li>
<li>by providing founders with their networks in those areas,</li>
<li>by buying resources at a bulk discount that founders need (such as compute credits), and</li>
<li>by finding founders that none of the other investors or funders are aware of to negotiate deals with them where they receive a large share of their impact certificate/s.</li>
</ol>
</li>
<li>Charities can attract top talent and align incentives with top talent who may not be fully sold on the charity’s mission:<ol>
<li>by promising them a share in all impact sold,</li>
<li>by locking that share up in a vesting contract,</li>
<li>by (possibly) sharing rights to the impact with another company that is the current employer of the talent so that they don’t need to quit and can draw on the infrastructure of the other company. (In fact, somewhat value-aligned companies may be interested in becoming investors themselves if they want to retain talent who want to work on prosocial applications of their knowledge.)</li>
</ol>
</li>
<li>Individual researchers can attract funding for their work even without the personal ties to funders, e.g., because they are in a different geographic region and better at their research than at networking.</li>
</ol>
<h2 id="profitability-of-impact-markets"><a class="toclink" href="#profitability-of-impact-markets">Profitability of Impact Markets</a></h2>
<p><img alt="Profitability of impact markets" src="/images/windfall-clauses-and-retroactive-funders/profitability.png"></p>
<p>We think about this in terms of the riskless benchmark <span class="math">\(B\)</span> and the ratios <span class="math">\(r_c\)</span> and <span class="math">\(r_p\)</span>. The benchmark <span class="math">\(B\)</span> is a return – e.g., <span class="math">\(B = 1.1\)</span> for a 10% profit – that an investor expects over some time period. An investment is interesting for the investor if it is more profitable than <span class="math">\(B\)</span>. <span class="math">\(r_c = \frac{c_f}{c_i}\)</span> is the ratio of the costs that funder and investor face respectively. This includes, for the funder, the cost of the grant, the time cost of the due diligence, the reputational risk if the due diligence misses something, and, for the investor, the cost of the grant minus savings thanks to shared infrastructure, economies of scale, etc. <span class="math">\(r_p = \frac{p_i}{p_f}\)</span> (note that enumerator and denominator are the other way around) is the ratio of the probabilities that investor and funder respectively assign to the project success. The investor may specifically select projects where they have private information (e.g., thanks to their network) that give them greater confidence in the project’s success than they expect the funder to have.</p>
<p>Hence, investments are interesting if <span class="math">\(r_c \cdot r_p > B\)</span>.</p>
<p>The graph shows the benchmark of an investment with 30% riskless profit compared to the maximum profit from various project configurations. It elucidates that an investor who can help realize a project more cheaply than the funder or thinks that it is more likely to succeed, can outperform the funder in a range of scenarios. These are scenarios where one or both parties can reap the gains from trade and save time or money.</p>
<p>The square between 0 and 1 on both axes is largely irrelevant. These are scenarios where the investor would have to pay more than the funder or is less optimistic about the project. Those are obviously uninteresting. But also just outside that square and around the edges, there are areas where the investor may not be interested because their edge (in terms of the <span class="math">\(r_p\)</span> and <span class="math">\(r_c\)</span> ratios is too small. Then again a riskless 30% <span class="caps">APY</span> is a high benchmark.</p>
<p>A few examples:</p>
<p>If a charity already has a track record of doing something really well 10 out 10 times in the past, there is very little risk involved when they try it for an 11th time:</p>
<p>Maybe an investor thinks they’re 99.5% likely to succeed and the funder thinks they are at least 99% likely to succeed, and the action costs $1m for either and takes a year.</p>
<p>That’s <span class="math">\(r_p = 1.005\)</span> and <span class="math">\(r_c = 1\)</span>. It is only interesting for an investor who cannot otherwise invest the money at 0.5% profit per year.</p>
<ol>
<li>It’ll be worth little to the funder: If they value the impact at 99% probability at $1m, they’ll pay $1m/99% ≈ $1.01m for it, so $10k premium.</li>
<li>If an investor offers to carry that tiny amount of risk, they’ll want it to exceed their 10–30% benchmark after a year, or else a standard <span class="caps">ETF</span> investment would be more profitable to them. That’s at least a $100k premium.</li>
<li>A bid of a $10k premium (minus the overhead of the whole transaction) from the funder but an ask of $100k premium from the investor means that there’ll be no deal.</li>
</ol>
<p>But consider a case where someone has no track record:</p>
<p>The investor thinks they are 20% likely to succeed. The funder thinks that they’re 10% likely to succeed. The action costs $1m for both and takes a year.</p>
<p>That’s <span class="math">\(r_p = 2\)</span> and <span class="math">\(r_c = 1\)</span>. It’ll be interesting unless someone has a benchmark of more than 100% per year.</p>
<ol>
<li>The funder will pay up to $1m/10% = $10m for the riskless impact.</li>
<li>That’s a 1000% return (or 900% profit) for the investor with 20% probability, so 100% profit in expectation, which beats most benchmark investments. Even if their riskless benchmark is as high as 30%, they’ll accept offers over 650% return. Naturally, these investors have to be fairly risk neutral or make many such investments. (If they are somewhat altruistic, they can consider the difference between the risk neutral and their actual utility in money a donation.)</li>
<li>Funder and investor will meet somewhere at or below 1000%.</li>
</ol>
<p>It is easy to create an analogous example for the case where funder and investor make the same probability assessments but where the grant size is so small that the investor, who already knows the project, can fund it at half the price compared to the funder who would have to spend a lot of time on due diligence.</p>
<h2 id="impact-timeline"><a class="toclink" href="#impact-timeline">Impact Timeline</a></h2>
<p>A typical product that is suitable for impact markets is a scientific paper. Papers, like many other projects, have the property that they often get stuck in the ideation phase, sometimes have to be abandoned (for other reasons than being an interesting negative result) during the research, sometimes don’t make it past the reviewers, and sometimes turn out to have been a bad idea only decades later.</p>
<p>When an investor wants to invest into a paper that has not been written, but which they are highly optimistic about, they may see these futures:</p>
<p><img alt="Timing of retroactive funding" src="/images/windfall-clauses-and-retroactive-funders/timing-of-retro-funding.png"></p>
<p>The <em>x</em> axis is the time (in years), the <em>y</em> axis is the <a href="https://forum.effectivealtruism.org/posts/7kqL4G5badqjskYQs/toward-impact-markets-1#Attributed_Impact">Attributed Impact</a> (proportional to dollars), blue lines are possible futures, and the red line is the median future.</p>
<p>There are two big clusters: all the futures in which the paper gets written, published, and read, versus and all the futures in which it either never gets finished or gets read by too few people.</p>
<p>One to three years into the process, it becomes clear in which cluster a given future falls, particularly if it falls into the upper cluster. (Otherwise there’s a bit of a halting problem because it might still take off.) Maybe the paper has been published on arXiv and is making rounds among other researchers in the field.</p>
<p>After 10 years, the majority of the impact has become clear and the remaining uncertainty over the value of the Attributed Impact of the paper is low.</p>
<p>After 15 years, we’re asymptotically approaching something that looks like a ceiling on the Attributed Impact of the paper. Experts have hardly updated on its value anymore in years, so their confidence increases that they’ve homed in on its “true” value. (“True” in the intersubjective sense of Attributed Impact, not in any objectivist sense.)</p>
<p>This is a vastly idealized example. In practice it may be that a published paper that used to be held in high regard suddenly turns out to have been wrong, an infohazard, plagiarized, etc. Or it may be that it’s suddenly noticed that a decade-old forgotten-about paper (that had high ambitions at the time but seemed to fall short) contains key answers to an important new problem.</p>
<h2 id="timing-of-retroactive-funding"><a class="toclink" href="#timing-of-retroactive-funding">Timing of Retroactive Funding</a></h2>
<p>If an investor is a specialist in some small field and profits from economies of scale in the field (e.g., the compute credits bought in bulk that we mention above), then they may expect to make a 10× profit from each retro funding that they receive. That’s the difference between the size of the retro funding at which the retro funder breaks even (ignoring interest) and the cost to the investor. We assume for simplicity that monetary and time costs (grants and due diligence) are the same. So, 2 · 10<em>i</em> − 2 · 5<em>i</em> = 10<em>i</em>, where <em>i</em> is the average seed investment. (We’re using the parameters from above where retro funders save 10× from making fewer grants and 10× from saving time spent on due diligence. We also assume that a patient, well-networked, specialized investor has twice the hit rate of the generalist funder.)</p>
<p>If, counterfactually, they would’ve invested this money at 30% <span class="caps">APY</span>, the impact market ceases to be interesting for them if they expect the retro funding to take longer than 8–9 (≈ 10.6 ≈ 1.3<sup>9</sup>) years: 2 · (1 / <em>rate<sub>funder</sub></em>) − 2 · (1 / <em>rate<sub>investor</sub></em>) = (1 + <em>apy</em>)<sup><em>years</em></sup>.</p>
<p>Here we’re comparing an investor at different hit rates to a retro funder who would otherwise have a 10% hit rate under four counterfactual market scenarios. The impact market is profitable for any number of years less than the break-even point.</p>
<p><img alt="Break-even vs. hit rate without profit" src="/images/windfall-clauses-and-retroactive-funders/break-even-vs-hit-rate-without-profit.png"></p>
<p>If the retro funder wants to save money, they can pay out less, but will need to do so earlier. For simplicity, the following chart is only for the scenario with a counterfactual 30% <span class="caps">APY</span>: 2 · (1 / <em>rate<sub>funder</sub></em>) · (1 − <em>savings</em>) − 2 · (1 / <em>rate<sub>investor</sub></em>) = 130%<sup><em>years</em></sup>.</p>
<p><img alt="Break-even vs. hit rate without profit" src="/images/windfall-clauses-and-retroactive-funders/break-even-vs-hit-rate-with-profit.png"></p>
<p>A retro funder needs to take this into account when deciding how much certainty they want to buy from the investors. More added certainty comes at a higher price. They can regulate this through the size of their retro funding or through the timing. Depending on the impact in question there are usually certain sweet spots that they can aim for, and do so transparently so that investors know what time horizons to speculate on.</p>
<p>When it comes to our example above, it seems fairly clear whether the paper was a success (was written, published, and read by some people) after about 2–3 years. So one sweet spot may be to wait for the moment of publication (as a draft or after peer review) or after the initial public reception can be gauged. The second is interesting because investors may be well-positioned to help with the promotion.</p>
<p>But there are other options – less profitable options much later.</p>
<h2 id="dissolving-retroactivity"><a class="toclink" href="#dissolving-retroactivity">Dissolving Retroactivity</a></h2>
<p>We can imagine a chain of retro funders from a particular set of futures into the present: Someone makes a binding commitment that if they are successful in making a lot of money – say, their business is successful – they will use the money or a fraction of it to buy back impact that has previously been bought by a certain set of existing retro funders who the person trusts. They can continually add new ones to this set.</p>
<p>This can also be formulated as a prize contest: If I’m successful, I’ll use that budget to buy impact from my favorite retro funders at a reasonable bid price. If 1 in 5 projects still fail between the time when the retro funder bought them and the time when the success happens, the retro funder may buy them at 120% of the price that the previous retro funder paid.</p>
<p>Under this framing there is no qualitative difference anymore between investors and earlier retro funders. They’re all just different investors with different attitudes toward risk or preferences about how they weigh the profit vs. the social bottom line of their investments. (Some of them may choose to consume their certificates, though, to signal that they’ll never resell them.) There may even be investors who choose to invest into “whatever project person X will do next,” so earlier than the abovementioned seed investors.</p>
<p>A startup may be interested in making such a commitment because they have the choice to either do the research in-house or at least pay for it immediately or to pay for it later and only if they are successful. Since startup success is typically Pareto distributed, they’ll have vastly more money in the futures where they are successful than they have now or in unsuccessful futures. So this deal should be interesting for most startups.</p>
<p>For investors it’s a question of whether they want to expose themselves more to the field or to a particular team. If they’re excited about the team behind the startup and trust that team to do well regardless of what field they go into, they’ll want to invest directly into the startup. But if they’re more agnostic about all the teams in a field but are very excited about the field, they may prefer investing into the research projects to bet on the retro funding.</p>
<h2 id="example"><a class="toclink" href="#example">Example</a></h2>
<ol>
<li>Cultured meat (or “cell/clean/c meat”) startups may require a lot more research to be done on how to scale their production and make it cost-competitive. But they don’t yet have the money to do all of that research in-house.</li>
<li>They commit to investing a large portion of the money they’ll make from going public into buying impact. Specifically they hash out particular terms with an organization like Founders Pledge that stipulate what impact related to cultured meat research they will buy from which retro funders.</li>
<li>The promise of great potential future riches boosts funding and opens up hiring pools.</li>
<li>Eventually, the now more likely future might happen, and the large budget from the exits serves to buy most of the impact from the retro funders.</li>
</ol>
<h2 id="conclusion"><a class="toclink" href="#conclusion">Conclusion</a></h2>
<p>We’ve received a grant via the Future Fund Regranting Program to work on this. If you’d like to join our discussions, <a href="https://discord.gg/7zMNNDSxWv">please join our Discord</a>.</p>
<p>Thanks to my cofounder Dony for reviewing the draft of this post! He gets 1% of the impact of it; I claim the rest.</p>
<script type="text/javascript">if (!document.getElementById('mathjaxscript_pelican_#%@#$@#')) {
var align = "center",
indent = "0em",
linebreak = "false";
if (false) {
align = (screen.width < 768) ? "left" : align;
indent = (screen.width < 768) ? "0em" : indent;
linebreak = (screen.width < 768) ? 'true' : linebreak;
}
var mathjaxscript = document.createElement('script');
mathjaxscript.id = 'mathjaxscript_pelican_#%@#$@#';
mathjaxscript.type = 'text/javascript';
mathjaxscript.src = 'https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.3/latest.js?config=TeX-AMS-MML_HTMLorMML';
var configscript = document.createElement('script');
configscript.type = 'text/x-mathjax-config';
configscript[(window.opera ? "innerHTML" : "text")] =
"MathJax.Hub.Config({" +
" config: ['MMLorHTML.js']," +
" TeX: { extensions: ['AMSmath.js','AMSsymbols.js','noErrors.js','noUndefined.js'], equationNumbers: { autoNumber: 'none' } }," +
" jax: ['input/TeX','input/MathML','output/HTML-CSS']," +
" extensions: ['tex2jax.js','mml2jax.js','MathMenu.js','MathZoom.js']," +
" displayAlign: '"+ align +"'," +
" displayIndent: '"+ indent +"'," +
" showMathMenu: true," +
" messageStyle: 'normal'," +
" tex2jax: { " +
" inlineMath: [ ['\\\\(','\\\\)'] ], " +
" displayMath: [ ['$$','$$'] ]," +
" processEscapes: true," +
" preview: 'TeX'," +
" }, " +
" 'HTML-CSS': { " +
" availableFonts: ['STIX', 'TeX']," +
" preferredFont: 'STIX'," +
" styles: { '.MathJax_Display, .MathJax .mo, .MathJax .mi, .MathJax .mn': {color: 'inherit ! important'} }," +
" linebreaks: { automatic: "+ linebreak +", width: '90% container' }," +
" }, " +
"}); " +
"if ('default' !== 'default') {" +
"MathJax.Hub.Register.StartupHook('HTML-CSS Jax Ready',function () {" +
"var VARIANT = MathJax.OutputJax['HTML-CSS'].FONTDATA.VARIANT;" +
"VARIANT['normal'].fonts.unshift('MathJax_default');" +
"VARIANT['bold'].fonts.unshift('MathJax_default-bold');" +
"VARIANT['italic'].fonts.unshift('MathJax_default-italic');" +
"VARIANT['-tex-mathit'].fonts.unshift('MathJax_default-italic');" +
"});" +
"MathJax.Hub.Register.StartupHook('SVG Jax Ready',function () {" +
"var VARIANT = MathJax.OutputJax.SVG.FONTDATA.VARIANT;" +
"VARIANT['normal'].fonts.unshift('MathJax_default');" +
"VARIANT['bold'].fonts.unshift('MathJax_default-bold');" +
"VARIANT['italic'].fonts.unshift('MathJax_default-italic');" +
"VARIANT['-tex-mathit'].fonts.unshift('MathJax_default-italic');" +
"});" +
"}";
(document.body || document.getElementsByTagName('head')[0]).appendChild(configscript);
(document.body || document.getElementsByTagName('head')[0]).appendChild(mathjaxscript);
}
</script>Toward Impact Markets2022-03-14T00:00:00+00:002022-03-14T00:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2022-03-14:/toward-impact-markets.html<p>We want to create a market to trade “impact certificates” – a variation on <a href="https://en.wikipedia.org/wiki/Social_impact_bond">social impact bonds</a>.</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#summary">Summary</a></li>
<li><a href="#impact-markets">Impact Markets</a><ul>
<li><a href="#retroactive-funding">Retroactive Funding</a><ul>
<li><a href="#reduced-uncertainty">Reduced Uncertainty</a></li>
<li><a href="#reduced-evaluation-time">Reduced Evaluation Time</a></li>
<li><a href="#longer-market-exposure">Longer Market Exposure</a></li>
<li><a href="#earning-a-double-bottom-line">Earning a Double Bottom Line</a></li>
<li><a href="#spotting-blindspots-for-funding-opportunities">Spotting Blindspots for Funding Opportunities</a></li>
</ul>
</li>
<li><a href="#impact-certificates">Impact Certificates</a></li>
<li><a href="#impact-market-participants">Impact Market Participants</a></li>
<li><a href="#auction-platform">Auction Platform</a></li>
<li><a href="#impact-stock-and-derivatives">Impact Stock and Derivatives</a></li>
</ul>
</li>
<li><a href="#benefits">Benefits</a><ul>
<li><a href="#benefits-for-funders">Benefits for Funders</a><ul>
<li><a href="#greater-maximum-scale">Greater Maximum Scale</a></li>
<li><a href="#more-funding-opportunities">More Funding Opportunities</a></li>
<li><a href="#greater-hiring-pools">Greater Hiring Pools</a></li>
<li><a href="#more-priorities-research">More Priorities Research</a></li>
<li><a href="#parking-money">Parking Money</a></li>
</ul>
</li>
<li><a href="#benefits-for-charities">Benefits for Charities</a><ul>
<li><a href="#cheaper-liquidity">Cheaper Liquidity</a></li>
<li><a href="#aligning-incentives">Aligning Incentives</a></li>
<li><a href="#incentivizing-excellence">Incentivizing Excellence</a></li>
</ul>
</li>
</ul>
</li>
<li><a href="#risks">Risks</a><ul>
<li><a href="#impact-and-profit-distribution-mismatch">Impact and Profit Distribution Mismatch</a></li>
<li><a href="#moral-cooperation-failure">Moral Cooperation Failure</a></li>
<li><a href="#manipulation-of-the-default">Manipulation of the Default</a></li>
<li><a href="#drawbacks-of-verifiability">Drawbacks of Verifiability</a></li>
<li><a href="#noise">Noise</a></li>
</ul>
</li>
<li><a href="#solutions">Solutions</a><ul>
<li><a href="#curation">Curation</a></li>
<li><a href="#impact-attribution-norm-formerly-attributed-impact">Impact Attribution Norm (formerly “Attributed Impact”)</a><ul>
<li><a href="#definition">Definition</a></li>
<li><a href="#in-plain-terms">In Plain Terms</a></li>
<li><a href="#commentary">Commentary</a></li>
<li><a href="#timeline">Timeline</a></li>
<li><a href="#example">Example</a></li>
<li><a href="#pricing-impact">Pricing Impact</a></li>
<li><a href="#limitations">Limitations</a></li>
</ul>
</li>
<li><a href="#responsible-retroactive-funders">Responsible Retroactive Funders</a></li>
<li><a href="#pot-of-money">Pot of Money</a><ul>
<li><a href="#auction-platform_1">Auction Platform</a></li>
</ul>
</li>
<li><a href="#auditors">Auditors</a></li>
<li><a href="#shorting">Shorting</a></li>
<li><a href="#expose-certificates">Exposé Certificates</a></li>
<li><a href="#targeted-marketing">Targeted Marketing</a></li>
</ul>
</li>
<li><a href="#current-work">Current Work</a></li>
<li><a href="#acknowledgements">Acknowledgements</a></li>
</ul>
</div>
<h2 id="summary"><a class="toclink" href="#summary">Summary</a></h2>
<p>You can think of an impact market as a prize contest. There is a prize pool and an announcement that prizes will be awarded to incentivize certain contributions that the prize committee wants to see. Additionally, it’s possible for contestants to attract seed investments and collaborators to help realize their contributions in return for a promise to share the prize money.</p>
<p>Therefore, the backbone of the market are the prize committees, which we call <em>retroactive funders</em> or <em>retro funders</em>. They are, for example, foundations with a team of sophisticated altruistic grantmakers. They are transparent about the sorts of contributions (or <em>impact</em>) that they want to see and reward it when they see it. Some altruists or charities will be in a position to complete such actions without help, but others will need to fundraise or partner with others first. An impact market is where this fundraising and partnering can take place.</p>
<p>Impact markets promise to partially solve a range of pressing problems that altruists face. First and foremost, they allow grantmakers to delegate the “startup picking” aspect of their work to profit-oriented investors, and then only invest time and money in the minority of cases in which a project meets their bar for success. (See sections Retroactive Funding and Benefits.) These and other benefits can boost the efficiency of the funding allocation of hits-based funders by one or two orders of magnitude.</p>
<p>However, impact markets also bear a number of (1) benign risks that can cause them to fail and (2) more dangerous risks that can cause them to be harmful. (See section Risks.) We think it will be possible to address the risks in the second category, but we would like to be more certain of that. (See section Solutions.) Small-scale, sandboxed experiments among highly informed participants may help us test impact markets while limiting their risks, which is what some parties are now attempting. (See section Current Work.)</p>
<p>We hope that this article will give readers a chance to critique our plans and point out further risks or weaknesses in our defenses against known risks. If you would like to help, <a href="https://discord.gg/7zMNNDSxWv">please join our Discord</a>, comment, or contact us some other way.</p>
<h2 id="impact-markets"><a class="toclink" href="#impact-markets">Impact Markets</a></h2>
<p>Our current best guess is that impact markets will take the shape of four incremental additions to existing financial markets.</p>
<ol>
<li><strong>Retroactive funding.</strong> First there needs to be a norm of and ideally several institutions for retroactive funding. They have a lot of responsibility for the health of the market and need to monitor it closely. We don’t think that it’s generally feasible for them to operationalize up front exactly what the impact looks like that they want to buy (say, by pointing to particular metrics) but rather want to rely on indirect normativity: Market participants can read about them and talk to them and try to understand their values. Ideally there are many retro funders who are all highly qualified and responsible and who are getting more numerous over time and whose capital increases. Some of these retro funders may have been started with the explicit goal to incentivize more retro funders by retro-funding existing retro funders. (We’ll use the terms “retro funder” and “to retro-fund” as convenient short forms.)</li>
<li><strong>Contracts.</strong> Anyone who founds, funds, or otherwise supports a project – a supporter – and wants to participate in its success needs to negotiate their inclusion in a contract (an <em>impact certificate</em>) that documents what share of the impact of the project they own. Retroactive funders will later require this proof if they are interested in buying the impact from the supporter. (Whether it’s possible to produce such proof retroactively, when it is needed, depends on the specifics of the case.)</li>
<li><strong>Auction platform.</strong> Seed investors are probably hesitant to invest into impact if they have to fear that it’ll be hard or take a long time for them to find a buyer for it. An auction platform can streamline the interaction between the buyers and the sellers of impact.</li>
<li><strong>Impact stock and derivatives.</strong> Finally, at some later date, charities (incorporated as, e.g., public benefit corporations) may decide to create their own stock. Alternatively, some third parties may create derivatives that track the market capitalization of all impact certificates of one charity. These will make some forms of investment easier, and charities could use them for mission hedging. Seed investments no longer need to go through each impact certificate once a charity has stock.</li>
</ol>
<table>
<tr>
<td>
<strong>Notes</strong>
<ol>
<li>This whole document culminates in a particular definition of the <em>Impact Attribution Norm</em> (formerly <em>Attributed Impact</em>). All mentions of the “Impact Attribution Norm” and just “impact” refer, by default, to the Impact Attribution Norm according to that definition. Verbs such as “liking” or “valuing” impact also by default refer to valuing the Impact Attribution Norm of an action.
<li>What we call “charity” is a charity in spirit but may be an Impact <span class="caps">DAO</span> or an entity incorporated as for-profit if that is necessary. Examples: Against Malaria Foundation, Wave, Protocol Labs.
<li>What we call a “project” is a precisely defined undertaking of, for example, a charity. Examples: one distribution of long-lasting insecticide-treated bednets, an epic in a software development process, a conference.
<li>We want impact markets to be interesting for (1) altruistic consequentialists and (2) profit-oriented capitalists. (Most people are a bit of both.) If it ends up being useful for collectors too, all the better, but we want impact markets to work regardless of whether collectors get interested in them.
</li>
</ol>
</td>
</tr>
</table>
<h3 id="retroactive-funding"><a class="toclink" href="#retroactive-funding">Retroactive Funding</a></h3>
<p>Retroactive funders benefit in various ways. <a href="https://impartial-priorities.org/chaining-retroactive-funders.html">This follow-up article quantifies the benefits and uses a clearer example.</a> (<a href="https://forum.effectivealtruism.org/posts/a8tfqSXJ7QL8neEBk/retroactive-funding-impact-quantification-and-maximization?commentId=Bs95ZaoaidBPyBK2x">Here’s another attempt</a> that indicates that impact markets may be 60–11,000 times as efficient as prospective funding, or 500 times in the median case.)</p>
<h4 id="reduced-uncertainty"><a class="toclink" href="#reduced-uncertainty">Reduced Uncertainty</a></h4>
<p>Retroactive funders only fund projects that have achieved a certain degree of success.</p>
<p>Projects usually go through two phases, a phase where finding out whether it’ll be successful is a question that management experts and priorities research experts need to collaborate on and a phase where the management experts are not needed anymore.</p>
<p>Let’s say that the project is a proposed conference a year in the future.</p>
<p>In the first phase the project can fail because the team behind it splits up, because the team procrastinates until all venues are booked, or because a pandemic strikes. It can also fail because the attendees come away excited about Homeopaths Without Borders or because one attendee took out vacation days to attend and their replacement accidentally triggered a Dead Hand mechanism.</p>
<p>The second phase starts when the conference is over, the venue is cleaned, and the feedback from the attendees is in. In this second phase all the uncertainty concerning the team, their procrastination, and the pandemic is gone. What remains is just the second type of uncertainty. (Or some part of it since the clumsy replacement has probably returned to their accustomed job away from the Dead Hand mechanism.) The uncertainty at this stage is strictly lower than the uncertainty during the first phase.</p>
<p>Let’s suppose that a retroactive funder has the requirement on a conference that it must’ve happened, that not too many attendees died from Covid, and that the feedback was at least 90% positive. Let’s further suppose that they would have to spend one day per proposed future conference to assess whether the team behind it will reach that bar, and that then they’ll be right in 1 in 5 cases.</p>
<p>Not having to do all of this means saving 5 times 1 day of work per successful conference for the evaluation and keeping the grant money for longer. So even if they buy the impact of the 1 in 5 successful conferences at 5× the price, they still make a profit, altruistically speaking.</p>
<p>The lower the probability that they assign to the projects’ success, the greater the efficiency improvement that impact markets provide.</p>
<h4 id="reduced-evaluation-time"><a class="toclink" href="#reduced-evaluation-time">Reduced Evaluation Time</a></h4>
<p>Let’s suppose that the funder supports a lot of conferences that follow roughly the same pattern, so that can be said to implement the same intervention, and that the funder has invested enough time up front to now be quite confident in the value of the intervention.</p>
<p>If the 5 days for the team evaluation are half the time that they would, counterfactually, spend per project (counting the up-front investment into finding the intervention), then they can buy the successful conferences at (2 × 5)× the price of the counterfactual grant and still save money. (The factor 5 is from the subsection above.)</p>
<h4 id="longer-market-exposure"><a class="toclink" href="#longer-market-exposure">Longer Market Exposure</a></h4>
<p>Since they invest later, they can keep their money invested in yield-generating assets for longer. In most cases this will have a small influence, say, 1.1× per year over one or two years.</p>
<h4 id="earning-a-double-bottom-line"><a class="toclink" href="#earning-a-double-bottom-line">Earning a Double Bottom Line</a></h4>
<p>But there are even more benefits once impact markets are better established. Some retroactive funders may be quite conservative with their investments, so that they don’t think that their retroactively purchased impact certificates are (in impact terms) hundreds of times more valuable than the average impact certificate that gets purchased by other retroactive funders. These retroactive funders may be interested in doubling as a speculator in other parts of the impact market.</p>
<p>Usually they’d have to park their money for a few years, decades, or centuries in classic stocks, bonds, or ETFs, which generate comparatively little impact just to maintain their liquidity and thus option value. With impact markets that becomes unnecessary as they can put money in charities (essentially becoming investors too) and then take it out again when they want to use it for their retroactive funding.</p>
<p>As retroactive funders themselves they may also have the expertise to predict what other retroactive funders will want to buy, so that they can be among the most successful investors on the market. But we’d still think that the social bottom line will do the heavy lifting in that portfolio, so that it’s probably not so attractive for retroactive funders who think that an actual extra dollar for their retroactive funding is worth a lot more than a dollar to one of another retroactive funder’s favorite charities. For example, it might be that they estimate that their monetary profits will shrink by 1 percentage point but that the social bottom line will be similar to what the Against Malaria Foundation would have achieved with ten times that money. A GiveWell-type funder would be excited about that investment opportunity while a donor who typically supports organizations like <span class="caps">MIRI</span> would want to avoid it.</p>
<p>Finally, retroactive funders may burn their shares in impact certificates if speculators start avoiding them because they fear that the supply is too great. But if that is not a worry, other future retroactive funders may also just buy shares from earlier retroactive funders in a never-ending chain of retroactive funding. Such a chain will also signal that the retroactive funding is likely to only ever increase, which would boost long-term investment into impact certificates.</p>
<p>The gains from this parking are hard to pin down. The worst-case is close to 1× since it’s optional and funders won’t do it if it’s not useful for them. In the other direction it’s probably capped at around 20× since retroactive funding would not be attractive in worlds in which seed funding is (still) that effective even on the margin.</p>
<h4 id="spotting-blindspots-for-funding-opportunities"><a class="toclink" href="#spotting-blindspots-for-funding-opportunities">Spotting Blindspots for Funding Opportunities</a></h4>
<p>If the space is big with hundreds of promising charities with dozens of certificates each at any given time, then funders may be overwhelmed and miss great retro-funding opportunities if they are outside their social circles, in another country and language, or even just very unusual. A market made up of speculators from many different industries and countries may be better at recognizing such niche opportunities than any one funder. They will also be incentivized to make sure that the funder knows about them.</p>
<p>Conversely, retro funders need to be wary of the temptation to ignore unpopular impact certificates. Rewarding the sorts of speculators who are smart enough to notice great funding opportunities that no one else notices is exactly the sort of mechanism that makes impact markets valuable for prioritization.</p>
<h3 id="impact-certificates"><a class="toclink" href="#impact-certificates">Impact Certificates</a></h3>
<p>Impact certificates, in our model, are contracts that regulate how to delimit and distribute some unit of change to the supply of a public good. (See below for the Impact Attribution Norm that refines this “change to the supply of a public good.”) Owen Cotton-Barratt and I plan to publish a set of rules that will provide more clarity here. (You can find a sneak peek in our <a href="https://www.youtube.com/watch?v=ZiDV56o5M7Q&list=PLhuBigpl7lqtMdPkejuo3mHdLFX53ftXJ&index=5">Funding the Commons talk</a>.) They aim to smooth out some initial friction that a marketplace for impact certificates may face. We’ll summarize them briefly in the following.</p>
<table>
<tr>
<td><strong>The guiding metaphor.</strong> A winery is typically credited with producing wine (which corresponds to impact). This assumes that employees, letters of the land, states, ancestors, meteorologists, insects, et al. forfeit their claim to the wine one way or another, either by custom or through a contract. Wine can flow freely or evaporate, but it can also be bottled and sold. (The full bottle corresponds to an impact certificate.) The bottling, in turn, can be undone, which is usually followed by the consumption of the wine. That consumption, we assume, is final.
</td>
</tr>
</table>
<p>This metaphor should give a rough overview of the spirit of the rules.</p>
<p>The rules propose that impact certificates be:</p>
<ol>
<li>as clear and concrete as possible, and</li>
<li>issued only by all those actors collectively who own classic legal rights in that which created the impact.</li>
</ol>
<p>They also recommend that the market system make them:</p>
<ol>
<li>consumable in the sense that one can permanently reverse the issue, and</li>
<li>incremental in the sense that they don’t require changes to existing financial markets.</li>
</ol>
<p>Generally, we think that markets may well home in on these norms over time by themselves, but we’re currently in a good position to prevent all the frictions that would come with that. We think that in particular the first two rules are crucial. We’re more on the fence about whether rules 3 and 4 are really needed.</p>
<p>The rest is commentary:</p>
<p><strong>The first two rules</strong> jointly limit the kinds of outputs that can be bottled up in impact certificates. The limits are not hard, but greater ambiguity will make it harder to find buyers for an impact certificate: Buyers of overly vague impact certificates may make losses if others later sell impact certificates that seem to substantially overlap. No one will quite know who owns the overlapping bits, and bid prices will drop. These buyers will later take good care not to buy overly vague certificates.</p>
<p>Ambiguity can have two sources: confusion over what is being sold and confusion over who (else) might make claims to that which is being sold. We therefore advise issuers of impact certificates to word them carefully and to make contracts in writing with whoever might make a claim on the impact. Note that such contracts needn’t specify a concrete fractional allocation but can also specify an algorithm (such as <a href="https://sourcecred.io/">SourceCred</a>) that is to be used to determine the allocation at the time of the issue.</p>
<p>Eventually, we imagine, aggregator and auditor firms will emerge that are independent of any issuers. Aggregators will compile all information on all impact certificates that have ever been issued on any market so that duplicates can be exposed. Issuers can then get audits to prove to buyers that they are honest and are not trying to double-spend their impact. (For example, someone might first sell their impact from “A fundraiser for Rethink Priorities,” and later sell their impact from “A fundraiser at a vertical farming expo” when really those are about one and the same fundraiser.)</p>
<table>
<tr>
<td>
<p><strong>Examples.</strong> A good definition may sound broadly like, “Within the 264th year of the era of Aelius Antoninus, I, Hypatia, will ghost-write and publish a paper on behalf of Ada Lovelace that proves (once she is born) that Vingean reflection is possible. I’m selling half of the impact. The other half will belong to Ms. Lovelace. There are no funders outside the market, and there are no conflicts of interest to disclose. In fact, our interests will be similar.”</p>
<p>(Or with reduced specificity: “Within the year N, I, Person A, will publish a paper that proves X, coauthored with Person B. I’m selling half of the impact. The other half belongs to Person B. There are no funders outside the market, and there are no conflicts of interest to disclose.”)</p>
<p>A paper has a clear owner, there are no funders who don’t participate in the market (including Hypatia herself), it is clear what she’ll do, how long it will take, and that afterwards the paper will be public, and it is clear that the uncertainty over the deeper merits of the work will take many centuries to be resolved. (Ada Lovelace lived about 1,400 years after Hypatia.) All this certainty and uncertainty can be priced in by the market.</p>
<p>A middling definition may sound like, “I, Eve, will distribute 1,000 copies of the attached Vegan Outreach leaflet at Barbican Station in London between April 1 and August 1, 2022. Vegan Outreach has waived all claims on the impact. I will sell 90% of the impact and retain 10% to appease anyone who has switched to lower-suffering behaviors in response to the leaflet but is not content to yield all of their impact to me.”</p>
<p>This definition is highly concrete but there is no clear concept of ownership or rights or responsibility to ground it in. As a result, some of Eve’s newly made vegans who have started to go leafleting themselves may want to sell their impact from their own leafleting. The valuation of Eve’s certificate would have to be enormous to appease them all with just 10% of it. Speculators should price in that there’s a lot of fuzziness here about how much of the impact is owned by who.</p>
<p>A really bad definition is something like, “I, Hancock, will use my superpowers to do something for animal rights.” It is vague in almost all the ways it can be vague. There is a broad idea of a strategy in there, but speculators will have no idea whether they should price it like a corporate campaign or like leafleting. Worse, the definition also leaves the door open to very harmful activities, so that, regardless of the actual outcomes, the Impact Attribution Norm is likely negative. A less extreme version of this is a hypothetical impact certificate for a whole organization that is still active or the whole life of a live person. Any prediction as to how these may change their strategies over the coming decades will have high variance so that the Impact Attribution Norm will almost inevitably be low.</p>
</td>
</tr>
</table>
<p><strong>The third rule</strong>, the consumption mechanism, allows certificate owners to influence prices by verifiably signaling that they will never sell the certificate. Possible tax-exemptions could be tied to consumption, which may be a minor concern for most investors, but some may invest through a legal entity that can only make tax-exempt grants. The consumption mechanism has also been called “dedication” and “burn.” We’re not convinced that this mechanism is necessary at first, but time will tell. One indicator that it is necessary will be if investors don’t expect impact certificates to be sufficiently deflationary or are worried that retro funders with large holdings might suddenly sell out of their positions.</p>
<table>
<tr>
<td><strong>Tax exemption.</strong> Note that some organizations in the crypto space seem to function like (our phrasing) “optional donor advised funds” in that they have tax exemptions in various countries and can write donation receipts but will only do so if the donor chooses to donate the money, which they don’t have to do. A retro funder could interact with the market through such an entity. If they choose to resell their impact certificates, they will not receive a donation receipt, but when they choose to burn their certificate, they get it. The donation receipt will of course then be from that intermediary rather than from the charity that actually received the money.
</td>
</tr>
</table>
<p><strong>The fourth rule</strong> aims to allay worries that impact markets, if successful, will cause revolutionary changes to existing allocation mechanisms. Avoiding such changes avoids opposition, which should make it much easier to institute impact markets. Conversely, even fervent supporters of impact markets may find proposals unrealistic that require changes to existing, established markets.</p>
<h3 id="impact-market-participants"><a class="toclink" href="#impact-market-participants">Impact Market Participants</a></h3>
<p>The key idea behind impact markets is that we want to reward supporters (such as founders, investors, and advisors to charities) who proved a particular aptitude for predicting what interventions will later be seen as impactful by sophisticated altruists.</p>
<p>There are three types of fundamentals on an impact market:</p>
<ol>
<li><strong>Projects:</strong> These are verifiable, clearly ownable, and clearly delimited sets of actions that are detailed in an impact certificate. They benefit from:<ol>
<li>Seed funding,</li>
<li>More options for aligning incentives.</li>
</ol>
</li>
<li><strong>Supporters:</strong> These include founders, advisors, seed funders, et al. of a charity. Everyone who negotiates a share in the impact certificate. We also use the terms <em>investor</em> and <em>speculator</em> for the seed funders. They benefit from: 3. The seed funding of the project (founders and employees), 4. Being able to make an exit (founders and other shareholders), 5. Passive income (investors).</li>
<li><strong>Retroactive funders:</strong> The backbone of the market. The funders who buy impact certificates once the projects have matured to a point where the retroactive funders are confident enough in their evaluation of the projects’ impact. They reward all supporters of a successful project generously, because they are thankful for not having to bear the risks of all the projects that fail even by their own lights. They benefit from: 6. Not carrying the risk for failing projects (5–10x monetary savings), 7. Not having to do charity evaluation (team, market, etc.) in addition to intervention evaluation (2x time savings), 8. Investing later in time (1.5x monetary savings), 9. Parking money (in the role of a speculator) where it generates a double bottom line (1–20x savings?), 10. Being able to draw on the market to notice blindspots in the in-house prioritization.</li>
</ol>
<p><img alt="Impact market diagram" src="/images/toward-impact-markets/impact-market.png"></p>
<p>Let’s suppose the Against Malaria Foundation (<span class="caps">AMF</span>) wants to do a distribution of long-lasting insecticide-treated bednets with its old distribution partner Concern Universal (<span class="caps">CU</span>). Let’s further suppose that it wants to use impact certificates for this and not its usual funding channels.</p>
<p>(This is an example only. We didn’t talk to <span class="caps">AMF</span> or <span class="caps">CU</span> about any of this and don’t actually want any charities that don’t have expertise in innovative financial products to become involved so early in the process! Impact markets are probably also not suitable for such safe investments as <span class="caps">AMF</span>’s net distributions.)</p>
<ol>
<li><span class="caps">AMF</span> ascertains that some retroactive funders are still interested in more net distributions.</li>
<li>The <span class="caps">AMF</span> staff decide how they want to split the Impact Attribution Norm from the distribution among themselves, the legal entity <span class="caps">AMF</span>, all their partners, contributors, and advisors, and their funders. These decisions can be adjusted in later negotiations. (we’ll treat <span class="caps">AMF</span>, the legal entity, and all its staff as just “<span class="caps">AMF</span>” in the following to save space.)</li>
<li><span class="caps">AMF</span> decides on the details of the auction process, that is, what fraction of the profits from each sale should go to <span class="caps">AMF</span> and what the minimum percentage raise is between bids.</li>
<li><span class="caps">AMF</span> writes an impact certificate.</li>
<li><span class="caps">AMF</span> contacts <span class="caps">CU</span>, negotiates their split with them, and sends them their fraction of the impact certificate. (we’ll assume here that the distribution partner has its own source of seed funding just to make its case interestingly different from <span class="caps">AMF</span>’s.)</li>
<li><span class="caps">AMF</span> contacts a venture capital firm. The <span class="caps">VC</span> agrees to fund the distribution. They negotiate the split, and <span class="caps">AMF</span> sends them their fraction of the impact certificate.</li>
<li>A year later the distribution gets completed successfully.</li>
<li>A retro funder notices the successful distribution, is ecstatic, and buys fractions of the certificate from all holders. <span class="caps">AMF</span> has no work with this: Any holder just has to give their certificate (or a fraction thereof) to the buyer.</li>
</ol>
<p>What we seek to demonstrate here is that impact certificates (the way we construe them) are not magic. They don’t need to be imbued with any deep meaning beyond that of any other contract.</p>
<h3 id="auction-platform"><a class="toclink" href="#auction-platform">Auction Platform</a></h3>
<p>We want to create an auction platform to streamline the whole process. (<a href="https://www.impactcerts.com/">You can keep up-to-date on the progress here.</a>)</p>
<p>Our ideas for MVPs differ in two main ways: (1) The <span class="caps">MVP</span> can either emulate the classic market mechanisms where multiple shareholders co-own a project, or it can use the Harberger tax auction where the original issuer is paid from the profit of each sale, sales are forced, and fractionalizing ownership among multiple shareholders is optional; and (2) it can be aimed at a crypto audience and run on a blockchain, or it can be aimed at a non-crypto audience, not use a blockchain, and avoid monetary transfers. (Note that we use auction and market synonymously here.)</p>
<p>Either way, the critical mechanism is that there’s someone (a retro funder) who rewards earlier supporters (founders, funders, advisors, et al.) if they made what has turned out to be good calls.</p>
<p>The classic, non-Harberger auction (or market) has the advantages that:</p>
<ol>
<li><strong>Alignment.</strong> You can align incentives with cofounders, employees, advisors, partner organizations, et al. by giving them shares in the impact.</li>
<li><strong>Seed funding.</strong> You can get investments by selling parts of the impact to early investors.</li>
<li><strong>Exit.</strong> You can participate in the success of your project by keeping some fraction of the impact until the project has come to fruition (by your assessment) and only then exit. As a founder you will be unusually convinced of it (because if more people were, it would typically already exist), and by extension, you will be more optimistic about it than your investors.Hence you’ll want to retain as much of your impact as you can afford until the day has come when you have the proof that will convince the investors too.</li>
<li><strong>Scaling down.</strong> You can start big projects by getting several investors who each buy smaller shares of the impact. Conversely, even smaller investors can get exposure to big projects by buying small shares in them.</li>
</ol>
<p>The Harberger auction shares two of them even without shared ownership:</p>
<ol>
<li><strong>Alignment.</strong> You can align incentives with cofounders, employees, advisors, partner organizations, et al. by making separate contracts with them that they’ll receive part of the profit share. That would be similar to an ordinary participation.</li>
<li><strong>Seed funding.</strong> You can get seed funding by selling all of your impact to early investors. You can set the minimum bid so that the funding is enough to bootstrap the project, but you can’t have any further fundraising rounds.</li>
<li><strong>Simplicity.</strong> You don’t need to countertrade your investors once you have the seed funding. You’ll simply profit from each of their sales automatically.</li>
</ol>
<p>With shared ownership, the Harberger auction shares all the benefits of the classic system, but in practice there will be trade-offs:</p>
<ol>
<li>A larger issuer share (or tax) will discourage speculation because it’s subtracted from the profits of the speculator. The market will be less liquid as a result. But issuers can set an arbitrarily small share to manage that risk. <a href="https://docs.projectserum.com/appendix/fees">Crypto exchanges</a> seem to work well with taker fees around and slightly below 0.05%.</li>
<li>But a small share also means small profits. Prices can’t decrease in the Harberger auction, so the profits will be proportional to the price. (If they could decrease, they’d be proportional to the volume, and charities had an incentive to create volatility.) In the classic system, you and your collaborators may collectively manage to retain 10–30% of your impact until you exit, which you have to time such that you don’t sell too early. In the Harberger system you don’t have to worry about the timing, but you’ll all collectively only “retain” (through the tax) 0.05% unless you want to sacrifice market liquidity. Even if you choose to set the tax at 1%, that’s still 10x the cost, plus the reduced liquidity, in exchange for simplicity.</li>
</ol>
<p>All in all, a Harberger auction with fractional ownership seems like it maximizes option value because issuers can choose to turn it into a classic auction by setting their tax to 0.</p>
<p>The current state is as follows:</p>
<p><strong>Blockchain (web3).</strong> Our first testing ground was a <a href="https://www.impactcerts.com/">smart contract on an Ethereum test net</a> that implements an auction as described above. It does not currently support fractional impact certificates, but we think that that could be added in the form of a separate third-party smart contract. In general, a blockchain-based or web3 solution would have a number of advantages:</p>
<ol>
<li>It makes it easy to interface with the existing web3 ecosystem around public goods funding.</li>
<li>The existing ecosystem is itself a great testing ground for innovative solutions because everyone is unusually knowledgeable of innovative financial tools, so we’re at less of a risk of making mistakes that would cause other people to lose money.</li>
<li>It allows for great scale as transactions are fast and frictionless, and, depending on the blockchain, also cheap.</li>
<li>It makes it easier to market the solution because users don’t need to trust us if they can read the code and there are audits.</li>
</ol>
<p>(You can watch a demo in <a href="https://www.youtube.com/watch?v=NTIAdn0Oms8&t=21m38s">our talk at Funding the Commons <span class="caps">II</span></a>.)</p>
<p><strong>Spreadsheet (web2).</strong> Another <span class="caps">MVP</span> that we have in mind is a spreadsheet (via Google Sheets) that tracks transactions between issuers, speculators, and retroactive funders (any pairwise permutations). One sheet simply records the transactions; another summarizes the transactions such that they show the latest configuration of ownership; and a third sheet displays some statistics, such as the price history, market capitalization, volume, etc. We haven’t created this spreadsheet yet, but we expect that it’ll be hard to scale for various reasons related to error-reporting and because spreadsheets get slow when there’s a lot of data in them.</p>
<p>Crucially, the spreadsheet would not execute the monetary transactions themselves, it would just record them. A transaction is then said to be pending when one side has entered it but the other hasn’t confirmed it. It is only considered for the statistics once both sides have confirmed it. Google Sheets allows owners to protect ranges and individual cells such that only one person can edit them, so that aspect should be fairly failsafe.</p>
<p>A variation on the same theme is a spreadsheet that only tracks what projects accept seed funding and what projects have been completed. All seed-funding transactions could then happen over a platform like Kickstarter or Indigogo. Once the project is completed, a retro funder can consider it and transfer their money either directly to all contributors or to the project founder so that they can forward it to all contributors.</p>
<p>We could test the spreadsheet solution by allowing people to use it to trade on retro funding for something very safe and easily verifiable such as <span class="caps">EA</span> Forum articles. One could also call this a “proportional prize pool for prescient philanthropists” because they are basically prizes for people who’ve made good calls – as founders, funders, supporters, or similar – except that not only the top 3 or top 10 get prizes but almost everyone who clears some minimal bar of the retro funder and has invested enough that the prize is worth the transfer overhead.</p>
<p>A web2 solution like that would have a few advantages too:</p>
<ol>
<li>It is quick to iterate on as it doesn’t require audits. It won’t handle financial transactions, and even if it did from the users’ perspective, they would really be handled by a payment gateway like Datatrans.</li>
<li>It would not handle financial transactions, so many legal risks from those shift away from us to the users of the system, who we’ll need to warn to do their research.</li>
<li>It’ll be more frictionless for the sorts of users who haven’t used web3 solutions before.</li>
</ol>
<p>Personally, we’re also a bit annoyed by people talking so much about valuing impact certificates in and of themselves like that’s important, so we like solutions that don’t obviously contain any one component that is an impact certificate. (We see an impact certificate like any other contract that we value only for the monetary or altruistic bottom line that it nets “us” – scare quotes because only the monetary bottom line is agent-relative.)</p>
<p>We are fairly convinced that the blockchain-based solution is going to be the culmination of our efforts one day, but we’re ambivalent over which <span class="caps">MVP</span> will allow us to test the market more quickly and productively. (Since we first published this document, we’ve focused on web2 and have graduated from the spreadsheet to a proper web app.)</p>
<h3 id="impact-stock-and-derivatives"><a class="toclink" href="#impact-stock-and-derivatives">Impact Stock and Derivatives</a></h3>
<p>At later stages, we imagine that investors will find it overly complicated to use individual auctions to bid on countless projects. Some charities may also prefer to use other kinds of auctions at least for some of their projects. Plus, the trust that a project will be impactful is probably going to be tied to the team behind it, the charity, so that people, especially those who are lay people when it comes to the interventions of the charity, will prefer to invest into charities instead of individual projects.</p>
<p>We imagine that this can be achieved with derivatives (such as perpetual futures) that track the market capitalization of all impact certificates that a charity has issued. Alternatively, and if that is legally permissible for the charity, it could issue its own stock, which it could also use to hedge against volatility in the flow of donations that it gets through other channels.</p>
<h2 id="benefits"><a class="toclink" href="#benefits">Benefits</a></h2>
<p>(We briefly covered this section in <a href="https://www.youtube.com/watch?v=NTIAdn0Oms8&t=1m30s">our talk at Funding the Commons <span class="caps">II</span></a>.)</p>
<p>We think that impact markets are most suitable for hits-based funding for individuals or young charity startups without impressive track records. <a href="https://impartial-priorities.org/chaining-retroactive-funders.html">This article makes this argument in greater detail.</a> In short, reference classes of projects where success probabilities over 80–90% are common make it hard for investors to think that they have enough private information about a particular project that they will accept the low success-conditioned rewards that retro funders will pay in such cases.</p>
<p>The following table gives an overview – i.e. rough tendencies based on broad simplifications – of who will find which benefit to be interesting:</p>
<p><img alt="Comparison of which benefits are interesting for funders vs. charities" src="/images/toward-impact-markets/benefits.png"></p>
<h3 id="benefits-for-funders"><a class="toclink" href="#benefits-for-funders">Benefits for Funders</a></h3>
<h4 id="greater-maximum-scale"><a class="toclink" href="#greater-maximum-scale">Greater Maximum Scale</a></h4>
<p><span class="caps">EA</span> funders seem to be reaching the limits of their scale. There is a vast difference between the time investment that GiveWell and <span class="caps">ACE</span> have found appropriate per dollar moved to a funding opportunity compared to the time investment that hits-based funders such as the Future Fund currently find appropriate. They are probably not wrong about that given their constraints. But impact markets can change these constraints.</p>
<p>Specifically, if a funder currently expects 10% of their grantees to succeed (by their lights), then impact markets can allow them to scale up by 10x just thanks to the grantmaking capacity that retroactive funding conserves. They may even save money in the process. <a href="https://impartial-priorities.org/chaining-retroactive-funders.html">This article quantifies the benefit some more.</a></p>
<h4 id="more-funding-opportunities"><a class="toclink" href="#more-funding-opportunities">More Funding Opportunities</a></h4>
<p>Funders are usually based in a specific location and have particular social circles. It is very costly for them to build up trust with someone outside those circles if they even learn about the person or group in the first place. That limits their access to funding opportunities to a tiny fraction of the landscape.</p>
<p>Much less trust is needed if founders and investors pitch already successful projects to the funders.</p>
<p>And the investors are already almost everywhere around the world in a variety of social circles, are expert networkers, and already have expertise in startup picking, which they can apply to charity startups.</p>
<h4 id="greater-hiring-pools"><a class="toclink" href="#greater-hiring-pools">Greater Hiring Pools</a></h4>
<p>Currently grantmakers who work for funders have to be highly trusted by the people with the money, have to be experts in priorities research (or at least in understanding and applying the latest priorities research), have to have a comprehensive overview of the spaces in which they’ll do their grantmaking, and have to be great at startup picking, including all the social skill that it takes to quickly form an accurate opinion on a founder team. That’s a lot of constraints, and some of them feel intuitively anticorrelated so that it’s even harder to find people that combine all of them on a high level.</p>
<p>With impact markets, they only need to be trusted experts in priorities research. They don’t need to have the comprehensive overview of the spaces because countless investors are their eyes and ears. And they don’t need to be great judges of character and project-specific skill because the projects that get pitched to them already succeeded to some appreciable extent.</p>
<p>That should expand the hiring pool for funders many times over.</p>
<h4 id="more-priorities-research"><a class="toclink" href="#more-priorities-research">More Priorities Research</a></h4>
<p>Normal investors will have incentives to publish priorities research that they do after they have invested in order to increase the price at which the certificate or certificates trade. Or they will at least publish the good bits that swayed them to invest.</p>
<p>Short-sellers will do the opposite and publish their exposés.</p>
<p>These can contain important considerations that we’ve previously missed. Failing that, they may help communicate important considerations from priorities research to a wider audience.</p>
<h4 id="parking-money"><a class="toclink" href="#parking-money">Parking Money</a></h4>
<p>Large funders currently park their money in for-profit businesses until the growth rate from the growing wisdom of the funder (plus the low average growth rate of the businesses) drops below the growth rate of the public goods. That can take a long time and is therefore very wasteful, but it’s a necessary evil because investments into public goods are currently illiquid, so if you get them wrong, you can’t withdraw the money again. Impact markets would change that, and large funders could park their money in ETFs that comprise many big, well-established and somewhat impactful charities until they find better investment vehicles.</p>
<h3 id="benefits-for-charities"><a class="toclink" href="#benefits-for-charities">Benefits for Charities</a></h3>
<h4 id="cheaper-liquidity"><a class="toclink" href="#cheaper-liquidity">Cheaper Liquidity</a></h4>
<p>Charity startups and individual altruists will receive seed funding from the investors who are already in their networks and who already trust them. These may be business angels, <span class="caps">VC</span>, or simply room mates who have the sort of income that they can advance the rent for a year.</p>
<h4 id="aligning-incentives"><a class="toclink" href="#aligning-incentives">Aligning Incentives</a></h4>
<p>Charities typically like to hire very closely value-aligned people. That means that they can pay lower salaries because everyone cares about the mission. But that doesn’t seem like the optimal state. There are people who care about the mission but also have a family to feed or a debt to repay. There are also extremely capable people who care a bit less about the mission. The charity will lose out on them.</p>
<p>If we now assume that the employee that the charity wants to hire has a bit of runway and can run about as much risk as an early startup employee. If they think that a retroactive funder will like the charity enough to buy its impact certificates, that employee may agree to a deal where the regular salary plus the expected value of 1% of all impact certificates adds up to more than a regular salary. If the employee cared a bit about the charity’s mission before, now they care a lot.</p>
<h4 id="incentivizing-excellence"><a class="toclink" href="#incentivizing-excellence">Incentivizing Excellence</a></h4>
<p>Many funders today aim to fill or to partially fill the funding gaps of any project that clears some bar in terms of expected cost-effectiveness. So highly capable startup founders have a tradeoff to make whether they want to start a for-profit business and donate billions or whether they want to start a charity and get at most exactly as much as they need. They wouldn’t want more unless they have enough spare time to become better grantmakers than those who make grants to them. But if charities can fundraise from for-profit venture capitalists, the gulf between charities and businesses shrinks. This means entrepreneurs have a less difficult tradeoff to consider when choosing between launching a charity or a for-profit.</p>
<p>The quick, quantitative feedback could also by itself be motivational for founders and employees.</p>
<h2 id="risks"><a class="toclink" href="#risks">Risks</a></h2>
<p>(We covered this section in <a href="https://www.youtube.com/watch?v=NTIAdn0Oms8&t=2m53s">our talk at Funding the Commons <span class="caps">II</span></a>.)</p>
<p><a href="https://docs.google.com/document/d/1fQIbl6vi8rs68uj96Zg0zdcwMmx4IdPdrA_ClfmxydI/edit">Have a look at this summary table for a quick overview.</a></p>
<h3 id="impact-and-profit-distribution-mismatch"><a class="toclink" href="#impact-and-profit-distribution-mismatch">Impact and Profit Distribution Mismatch</a></h3>
<p>Investors should have strong reasons to expect that the prices of certificates will, in the limit, be proportional to the value that a <a href="https://longtermrisk.org/msr">Pareto-optimal compromise axiology</a> would assign to them – that is the moral standard that is reached only when no gains from moral trade are left on the table.</p>
<p>But we think that is unlikely to happen by default. There is a mismatch between the probability distribution of investor profits and that of impact. Impact can go vastly negative while investor profits are capped at only losing the investment. We therefore risk that our market exacerbates negative externalities.</p>
<p><img alt="Impact distribution" src="/images/toward-impact-markets/impact-distribution.png">
<img alt="Profit distribution" src="/images/toward-impact-markets/profit-distribution.png"></p>
<p><strong>Standard distribution mismatch.</strong> Standard investment vehicles work the way that if you invest into a project and it fails, you lose 1 x your investment; but if you invest into a project and it’s a great success, you may make back 1,000 x your investment. So investors want to invest into many (say, 100) moonshot projects hoping that one will succeed.</p>
<p>When it comes to for-profits, governments are to some extent trying to limit or tax externalities, and one could also argue that if one company didn’t cause them, then another would’ve done so only briefly later. That’s cold comfort to most people, but it’s the status quo, so we would like to at least not make it worse.</p>
<p>Charities are more (even more) of a minefield because there is less competition, so it’s harder to argue that anything anyone does would’ve been done anyway. But at least they don’t have as much capital at their disposal. They have other motives than profit, so the externalities are not quite the same ones, but they too increase incarceration rates (Scared Straight), increase poverty (preventing contraception), reduce access to safe water (some Playpumps), maybe even exacerbate s-risks from multipolar <span class="caps">AGI</span> takeoffs (some <span class="caps">AI</span> labs), etc. These externalities will only get worse if we make them more profitable for venture capitalists to invest in.</p>
<p>We’re most worried about charities that have extreme upsides and extreme downsides (say, intergalactic utopia vs. suffering catastrophe). Those are the ones that will be very interesting for profit-oriented investors because of their upsides and because they don’t pay for the at least equally extreme downsides.</p>
<p>Most profit-oriented investors also care about countless things besides profit, but we think it makes sense to think about these risks with a security mindset and assume that there are purely profit-oriented investors out there who can move a lot of capital. Besides, most investors will not be aware of the interests of future generations millions of years from now, of invertebrates, or of beings close to our acausal trade partners.</p>
<p><img alt="Silly quiz" src="/images/toward-impact-markets/quiz.png"></p>
<p>A simplified analogy is pictured above. This first aid quiz counts correct answers only – just like the market – so if you select all answers, you get the full points even though most people would not survive such treatment for nosebleed.</p>
<p>This risk applies to prize contests in general, especially relatively long-running ones. We find it reassuring that prize contests have not yet ended the world, but we would like to see more historical analysis of the actual hidden (or maybe obvious) harms that have come off well-intentioned prize contests in the past.</p>
<p><strong>Anthropic distribution mismatch.</strong> Moreover, if the downsides are strictly about extinction, then the investors will lose their bets in worlds in which they wouldn’t have been able to spend the money anyway.</p>
<p>They might regret this if they thought that their bet itself had increased the probability of losing the bet, but they’ll probably assign a low probability to that because they may, for example, reason that they merely defected in an already hopeless collective prisoners’ dilemma. Some people reason along the same lines when they argue that divestment is ineffective (<a href="https://sideways-view.com/2019/05/25/analyzing-divestment/">something that Paul Christiano critiques</a>): If the market is sufficiently efficient, any divestment will result in an inefficiency that will quickly be compensated by equally sized investments of others. These factors cause us to worry that investors will be likely to defect against values that are concerned with x-risks (including s-risks).</p>
<p>Both of these can be framed as problems of lacking moral trade because the investors defect against some futures for the benefit of others when they should’ve invested in ways that respect the interests of all futures.</p>
<p>Below we present a definition of the “Impact Attribution Norm” that combines earnest intention with outcome and thereby addresses the problems related to moral trade and extinction.</p>
<h3 id="moral-cooperation-failure"><a class="toclink" href="#moral-cooperation-failure">Moral Cooperation Failure</a></h3>
<p>It is uncommon for two people to each have only one interest, and for these interests to be exact opposites. So in most cases it should be possible to find a compromise that improves the aggregate good that is achieved and, ideally, one that is as good or better than the cooperation failure for both, a Pareto-improvement.</p>
<p>Here are some pilfered images by Brian Tomasik from the <a href="https://longtermrisk.org/gains-from-trade-through-compromise/">Center on Long-Term Risk</a>:</p>
<p><img alt="Impact market diagram" src="/images/toward-impact-markets/policy-space.png"></p>
<p>The caption is: “Pareto improvements for competing value systems. The two axiologies are opposed on the x-axis dimension but agree on the y-axis dimension. Axiology #2 cares more about the y-axis dimension and so is willing to accept some loss on the x-axis dimension to compensate Axiology #1.”</p>
<p><img alt="Impact market diagram" src="/images/toward-impact-markets/pareto-compromise.png"></p>
<p>Caption: “Figure 2: Imputations for compromise between deep ecologists and animal welfarists, with pi = 0.5 for both sides. By “Pareto frontier” in this context, I mean the set of possible Pareto-optimal Pareto improvements relative to the (50, 50) disagreement point.”</p>
<p>The line labeled as “Fight” is the conflict that we would see on our markets if both parties were to fundraise for their conflicting goals and consequently burn most of their resources in a zero-sum conflict. But if they talk to each other and compromise, then they can realize outcomes that are better in aggregate (almost inevitably), and in particular outcomes on the Pareto frontier, which are directly desirable (or neutral) for both parties!</p>
<p>Had impact markets existed a few decades ago, there might’ve been projects that pushed for coal energy over nuclear energy. These might then have gotten a lot of funding from investors and retroactive funders who didn’t think that climate change was a significant worry. Our and future generations would then have been defected against. (Our interests would’ve been ignored because we didn’t participate in the market through retro funders.) Instead it would’ve been possible for these people to instead invest into R <span class="amp">&</span> D for the next generation of safer nuclear reactors, which would’ve allayed their safety concerns without exacerbating climate change.</p>
<p>Similarly, an animal conservation organization might want to use impact markets to fundraise for the protection of certain predator animals. They might fundraise large sums from people who are interested in the protection of that species, but they would thereby defect against the interests of the prey animals who are much more numerous and probably similarly capable of suffering. Instead they could’ve fundraised for the protection of a herbivore species that mostly eats fruit and whose members are unlikely to experience great amounts of suffering themselves throughout their lives.</p>
<p>Worse, the conservation charity that increases the number of predators at the expense of the prey animals may sell impact at a positive price and another charity that protects the prey animals against the predators by (say) making them infertile could also sell their impact at a positive price. Both impact certificates may have a large positive valuation when really their impact more or less cancels out! So this egregious waste of resources can even happen when both parties participate in the market in some way.</p>
<p>Another egregious example is terrorism. Terrorists may want to use impact markets to fundraise for their attacks. Maybe they are religiously motivated and are attacking people as a form of proselytizing with extreme prejudice. Meanwhile other groups may use the same markets to fundraise for terrorism prevention. That is again wasteful since they could instead pool their resources in open-access adversarial collaborations on questions of religion.</p>
<p>We think this lack of any incentive for individuals to engage in moral trade is a major risk. The distribution mismatch is, strictly speaking, a sub-risk, but that’s a bit unintuitive, and we think that both are so important that they deserve their own sections be it only for emphasis.</p>
<p>Finally note the difference between <em>revealed</em> and <em>idealized</em> preferences. <a href="https://impartial-priorities.org/values-spreading-taxonomy.html">The first step</a> when attempting a compromise should probably not be to trade right away but, if possible, <a href="https://www.lesswrong.com/tag/double-crux">to discuss the object-level disagreement</a>. It may be that at least one of you is simply wrong, will realize that, and will no longer be interested in the competing intervention at all. (But others might be.)</p>
<h3 id="manipulation-of-the-default"><a class="toclink" href="#manipulation-of-the-default">Manipulation of the Default</a></h3>
<p>An attacker might first build a reputation of regularly doing exactly the sorts of things that large funders hate, such as running flash loan attacks against crypto markets or denial of service attacks against blockchains, and then skip some of these regular attacks in order to sell their impact from <em>not</em> attacking. This exploit is based on the ability of the attacker to change the default state of the world from one in which they don’t attack to one in which they do attack, and regularly.</p>
<p>This failure mode would cause impact markets to exacerbate <em>exactly</em> the sorts of problems we want to solve. The “Impact Attribution Norm” definition below addresses it.</p>
<p>One real example that sort of fits is the following anecdote that I’ve heard from a reliable source but haven’t fact-checked: When the cap <span class="amp">&</span> trade carbon trading system was first introduced in the <span class="caps">EU</span>, many polluting companies either first polluted even more or exaggerated their level of pollution so that their default level of pollution would be set higher. With the exaggerated default it was cheap for them to (seemingly) reduce their carbon emissions again, and they didn’t have to buy carbon certificates.</p>
<p>Similarly, one <a href="https://www.reddit.com/r/UIUC/comments/ev9l5g/careful_shopping_at_county_market_these_are_some/">sometimes hears of stores</a> that momentarily increase the price for a product (or just lie about the old price) and then advertise the same or a higher price as a discounted sale price.</p>
<p>We’re not as worried about this as we are about moral trade and, in particular, the distribution mismatch, because attackers could, in many cases, already attack funders this way. The funders just have less of a target painted onto them, figuratively speaking. Yet funders should make it clear that they will not be fooled this way, because even a failed attempt at such a ruse can do harm, as in the case of the increased pollution.</p>
<h3 id="drawbacks-of-verifiability"><a class="toclink" href="#drawbacks-of-verifiability">Drawbacks of Verifiability</a></h3>
<p>Impact markets will probably tend toward high verifiability requirements all by themselves, but only after many buyers have been burned by investments into impact certificates that later turned out to overlap with other impact certificates and whose price dropped as a result. Prescribing a high degree of clarity from the start will hopefully avoid this friction and keep more buyers interested in impact markets. Keeping impact certificates more clearly delineated also helps to separate them from impact stock.</p>
<p>But there are two possible drawbacks to this. The first is related to Goodharting. Charities that have a very clear output – e.g., papers or bed nets – will have an easy time fundraising on impact markets. But charities with more fuzzy outputs – e.g., community cohesion or plan changes – may be relatively disadvantaged by the format. That could lead to a relative underinvestment of money and effort into these more fuzzy interventions. Worse, most interventions can probably easily be measured by a few proxy measures that, when optimized for, lead to terrible outcomes. That’s a case of a failure of moral trade. The Impact Attribution Norm discourages it, but overconfident issuers may still seek to sneak in the bad proxy measures that they perform excellently on.</p>
<p>The other drawback is that explicit contracts with collaborators may not be enough. We’re currently collaborating with many others in a range of ways. Some examples in rough order of how explicit they appear to us: collaborations between coauthors, collaborations with reviewers, collaborations with casual conversation partners, collaborations with providers of infrastructure, and collaborations with those in the past who have made it possible that we are alive today.</p>
<p>If the sale of an impact certificate for, say, a paper rewards only the authors, there’s a risk that other collaborators may decide that they will be rewarded better if they don’t engage with other authors anymore to fully focus on their own papers and not give away any ideas before they have published them.</p>
<p>This could also happen between organizations. Each may think that it has a uniquely important mission and a responsibility toward its employees, and that it may have to shut down without funding from the sale of impact certificates. Two seemingly altruistic reasons will push these organizations to withhold all sorts of resources from the other.</p>
<p>The solution is to talk to all collaboration partners that form a coalition and to negotiate an allocation of the returns from impact certificates that has the core property, that is that no collaborator thinks that it is in their interest to split off from the coalition.</p>
<p>That’s easily done among explicit collaborators such as coauthors, cofounders, or employees. Hence why the first rule requires explicit contracts between these. But in many other cases it will require acausal trade because it would be infeasible or impossible to (causally) negotiate with the collaboration partners – they may be anonymous, dead, distant, too many, or too busy. So the first rule falls short of solving this problem.</p>
<p>On the other hand, impact markets are likely to reduce financial scarcity, and people are less likely to behave uncooperatively when they have plentiful resources.</p>
<h3 id="noise"><a class="toclink" href="#noise">Noise</a></h3>
<p>The market itself will also serve as a source of information for investors to understand what impact likely maximizes the compromise morality. Investors who correctly identified highly uncontroversially valuable impact early on may sell to take profit, manage their risk, or pay their rent – reasons other than thinking that the impact is overvalued – but this sell may be misinterpreted by others as a signal that the impact was more controversial than they had thought. All of this introduces a lot of noise that distracts from the price discovery. This is something that happens on financial markets all the time, and public companies have to be sophisticated enough in their budgeting not to be ruined by fluctuations that have nothing to do with their fundamentals.</p>
<p>Sophisticated funders who are highly confident in their judgment, or more so than other market participants, can stabilize the prices by buying, holding, and dedicating impact certificates, but they can do so only to the extent that their budgets allow.</p>
<p>This is more of an inefficiency than a risk of net harm, so we’re less worried about it than about the above.</p>
<h2 id="solutions"><a class="toclink" href="#solutions">Solutions</a></h2>
<p>(We briefly covered this section in <a href="https://www.youtube.com/watch?v=NTIAdn0Oms8&t=7m57s">our talk at Funding the Commons <span class="caps">II</span></a>.)</p>
<h3 id="curation"><a class="toclink" href="#curation">Curation</a></h3>
<p>There are three forms of curation:</p>
<ol>
<li>Curation of each impact certificate that is submitted to the marketplace.</li>
<li>Curation of the issuers who are allowed to submit to the marketplace.</li>
<li>Curation of the investors and retro funders who are permitted to use the marketplace.</li>
</ol>
<p>These can be performed by a team of expert curators, by some sort of oracle, or by an auditor network (see below). Exposé certificates and shorting (see below) can help the curation team not miss any bad certificates.</p>
<p>The first method can be aided by limiting the market to a particular topic. If only <span class="caps">AI</span> safety research is allowed to be sold on the market at all, then it already excludes a lot of dangerous categories of things, such as terrorist attacks, and the curation team does not need to contain experts from more than one field.</p>
<p>The second and third can be based on the implicit web of trust that emerges when people are only allow to apply for access to the platform if they were recommended by existing users of the platform. The application process could include courses and quizzes if the platform is valuable enough for the users.</p>
<h3 id="impact-attribution-norm-formerly-attributed-impact"><a class="toclink" href="#impact-attribution-norm-formerly-attributed-impact">Impact Attribution Norm (formerly “Attributed Impact”)</a></h3>
<p>We think it will be key to find a definition of impact that has four properties:</p>
<ol>
<li>It leads investors to value impact in proportion to the moral gains from trade that it has generated,</li>
<li>It leads investors to value impact in accordance with some form of idealized rather than revealed preferences of all affected groups,</li>
<li>It forms an attractor state so that over time more investors tend to adopt the definition (be it unconsciously) rather than fewer, and</li>
<li>It tracks but makes more precise the existing shared intuition of what “someone’s impact” is.</li>
</ol>
<p>The following definition is an attempt at that. It relies on a lot of culturally shared understanding, which is hopefully also shared to a sufficient degree between investors. Enforcing honesty or detecting lying of market participants is outside the scope of the definition.</p>
<p>It boils down to something like, “Your impact must be widely regarded as morally good, positive-sum, and non-risky before and after you undertake your action for it to be valued at a positive price.”</p>
<h4 id="definition"><a class="toclink" href="#definition">Definition</a></h4>
<p>Impact Attribution Norm (v0.3):</p>
<ol>
<li>We define <em>impact</em> as the <em>gains from moral trade</em> achieved by a set of <em>actions</em>.<ol>
<li>We take into account the interests of <em>all moral patients</em>, regardless of time, location, species, substrate, gender, phenotype, orientation, market participation, etc.</li>
<li>Specifically, we value actions positively only if they were robustly positive in anticipation.</li>
</ol>
</li>
<li>We <em>attribute impact</em> to all those actors whose actions make them responsible for the impact.<ol>
<li>We draw on widely accepted cultural norms guiding responsibility.</li>
</ol>
</li>
<li>We assess impact as compared to counterfactual world histories that seem ordinary by broad societal standards.<ol>
<li>In particular we do not rely on counterfactuals that are easily influenced or merely claimed by the issuer.</li>
</ol>
</li>
<li>We leave it to the actors to make a case for the value of their impact.</li>
<li>Where we have to make judgment calls on behalf of others we will assume that they favor noncausal decision theories and the Kalai bargaining solution.</li>
</ol>
<h4 id="in-plain-terms"><a class="toclink" href="#in-plain-terms">In Plain Terms</a></h4>
<p>As issuer of an impact certificate:</p>
<ol>
<li>You need to make a case that the actions you’re planning to take will generate robustly positive impact. (You should reference the particular version of the definition that you’re using.)</li>
<li>You cannot just pick one metric or moral dimension along which you want your impact to be measured. You need to make the argument that your action is good or neutral across all moral views and interests, that harms are offset in ways the participants would accept, or something along those lines. In short, <em>an’ it harm none, do what ye will</em>. (Slight oversimplification.)</li>
<li>In particular, you need to argue that at this time (when you haven’t taken the actions yet) it is reasonable for the median market participant to think that your action will generate impact in expectation. This should be positive impact unless you’re happy that your impact certificate will have negative value and no one will buy it.</li>
<li>Note that you can’t inflate your impact by claiming that something unusually terrible would’ve happened if you hadn’t performed your action. If such a claim is not plausible to someone like Erin McKean, or anyone else who doesn’t know and knows nothing about you, then it doesn’t count.</li>
<li>Note also that you can’t lie about your actions. If you write an impact certificate for one action and then then perform a different action, the impact certificate is for an action that didn’t happen and there is no impact certificate for the action that you did do.</li>
<li>Finally, talk to others who you are bargaining with if at all possible. If not, please don’t automatically assume that they make decisions according to <span class="caps">CDT</span>, and if you’re unsure about what bargaining solution they’ll favor, go for Kalai.</li>
</ol>
<h4 id="commentary"><a class="toclink" href="#commentary">Commentary</a></h4>
<p><strong>Gains from moral trade.</strong> The reference to the “gains from moral trade” codifies that what makes an action valuable are not its effects along some arbitrary moral dimension but all moral dimensions, and in particular the improvement that the action achieves over a hypothetical completely disjoined state of the world where everyone is on their own.</p>
<p>There is the degenerate case of an action that benefits one moral view and is irrelevant for all others. Such actions might as well be valued positively, because why not, and we would welcome an elegant way of fitting them into this framework, but they are probably quite rare or small scale, so that they don’t matter much in the scheme of things. Maybe we can imagine that in such cases there is a being that gives you something that neither of you value or disvalue in return for you looking after your own moral interests.</p>
<p>This exclusive focus on the gains from moral trade has a host of related advantages: Investments into zero-sum games will have no value on a certificate market, except insofar as they generate net positive externalities; projects that defect against some future beings for the benefit of others are penalized; progress on some values that is offset by harm to other values is without value; projects are incentivized to seek out new moral dimensions orthogonal to all known ones; and projects are encouraged to do cross-cutting work that benefits a variety of moral systems.</p>
<p>The subitems clarify some assumptions. In particular, the category of moral patients is supposed to be interpreted maximally widely. You can’t just avoid defecting against someone by declaring them not a moral patient. When it doubt, these rules should be interpreted so to err on the side of caution and inclusion in the set of moral patients.</p>
<p><strong>Responsibility for the impact.</strong> Especially the term “responsibility” is one that relies on a culturally shared understanding, ideally one codified in law or a contract. We see no objective way to attribute impact other than by convention. Here the burden of proof is upon the person who makes the claim.</p>
<p><strong>Counterfactual world histories.</strong> This clause prevents issuers from manipulating the default as described above.</p>
<p>Changes in the supplies of goods need to be assessed against some counterfactual, and in this case we want to use a counterfactual that everyone can know and that is very hard to influence by any one actor. We might call this perspective the <em>systemic</em> stance, in reference to Dennett’s <a href="https://en.wikipedia.org/wiki/Intentional_stance">intentional stance</a> and system science, which is a perspective that abstracts from agents and intentions to recognize systemic or emergent mechanisms. It’s a form of underfitting or dimensionality reduction. As a result, an attacker would now have to change large swaths of society into one that regularly conducts flash loan attacks (per the example above) before their omissions to do so becomes different from the default.</p>
<p>Retro funders need to proactively signal their adherence to the norm to make it clear to any potential attackers that they are not vulnerable to this exploit.</p>
<p>A disclaimer, in case you now think we hate historians: We compare full world histories (unless the impact period is explicitly curtailed in the certificate), so not only the time between the decision to implement the project and the present moment or any other such period. But this is done in the spirit of <a href="https://forum.effectivealtruism.org/posts/XwYptiJQEaYZ72Zij/summary-of-evidence-decision-and-causality#Tickle_Defense">a belief that the Tickle Defense is probably true</a>. So a person who surprisingly kills someone produces evidence that people sufficiently like them have killed people before and thus narrows the set of subjectively plausible pasts to a set of worse pasts, but a historian who researches past killings has not the same evidential effect because the killings they research screen off any evidential effect the historian has. The spirit of our definition is that if the Tickle Defense fails, we want to correct our definition and not put blame on historians!</p>
<p>The sets of world histories all contain several pasts, presents, and futures, because the issuer had uncertainty over past, present, and future at the time of the decision. (In view of the many-worlds interpretation, we intend “world histories” to include all Everett branches weighed by their measures.)</p>
<p><strong>Make a case.</strong> Every issuer has to make the case that their action conforms to the norms. This should not be left only to the retro funders to figure out. The retro funders’ job is rather to find weaknesses in the case. If every impact certificate thus recapitulates the definition, it will gradually become reified culturally as the consensus definition of the Impact Attribution Norm. We call this a “driver of adoption.” The Impact Attribution Norm contains several for redundancy.</p>
<p>The norm follows (in spirit) the <a href="https://plato.stanford.edu/entries/knowledge-analysis/#KnowJustTrueBeli"><span class="caps">JTB</span> definition of knowledge</a> as a <em>justified, true belief</em>. In the impact certificate or some ancillary document, the issuers justify their claim to the Impact Attribution Norm by laying out how they think the reader ought to arrive at the conclusion that they were, at the time of the decision, justified to believe that what they did would generate positive impact. Time and the market then arbitrate the question of truth. Only if <em>ex ante</em> and <em>ex post</em> impact are positive is the aggregate impact (the minimum) positive.</p>
<p>This avoids the pathological case of <em>x-/s-risk gambles</em> where someone runs a large number of very risky projects – projects that can turn out vastly net positive or net negative – and then makes money by selling only the impact from those that happened to turn out positive even if they are in the minority. (Or likewise for funders.) Those that happened to turn out positive would still not be worth anything because their 0 or negative <em>ex ante</em> impact will be less than the positive <em>ex post</em> impact, so that the positive impact will have no influence on the eventual value. This mirrors the intuition between expectational consequentialism where, for example, a doctor who treats a mild headache with a medicine that is 99% likely to kill the patient acts immorally even if the patient actually happens to survive and be cured of the mild headache.</p>
<p>Note that this applies in the same way to the anthropic formulation of the x-/s-risk gamble where issuers or investors believe that (1) they hardly affect the odds of extinction, for example because of market efficiency, and (2) they are neutral about x-/s-risk futures because they can’t spend money anyway if they’re dead or otherwise incapacitated but they do care about the okay outcomes where they can spend money.</p>
<table>
<tr>
<td>
<strong>Aside.</strong> The similarity to the <span class="caps">JTB</span> definition suggests that the <a href="https://en.wikipedia.org/wiki/Gettier_problem">Gettier problem</a> may be a problem. But we haven’t managed to construct an actual pathological or exploitable version of it. (You can skip to the next subsection if you’re in a hurry.)
<p>
An example: Aanya Altruist reviews some 450+ studies and concludes that there is strong evidence that the serotonin transporter gene 5-<span class="caps">HTTLPR</span> is associated with depression. She develops and promotes a school-based screening procedure that provides people with the particular versions of 5-<span class="caps">HTTLPR</span> with resources on how to access therapy. Biology teachers love the program for how helpful and scientific it is. Aanya makes a case for her impact and sells it. Then it turns out that (1) <a href="https://slatestarcodex.com/2019/05/07/5-httlpr-a-pointed-review/">5-<span class="caps">HTTLPR</span> has nothing to do with depression</a>, and (2) many of the children who had received the resources had become depressed at the population base rate but started therapy at a much higher rate.
</p>
<p>
Aanya Altruist’s defense of her impact is state of the art by the standards of her field at the time, and she is honest about believing her conclusions. Yet they turn out to be false. So her impact is valuable by dint of her earnest, justified intentions and by dint of her quite unrelated impact. This is probably not going to be a common case, because we expect there to be few impactful interventions so that few people will find them by accident, but it doesn’t seem harmful either.
</p>
<p>
But the system is hypothetically exploitable in another way: Someone may find a new, previously unknown <span class="caps">CO</span><sub>2</sub> sink that they predict to start absorbing <span class="caps">CO</span><sub>2</sub> within a few years. They build a complex contraption that no one understands but that is just plausible enough that they can fool investors with an amphigorical justification. Then they start it right when the natural <span class="caps">CO</span><sub>2</sub> sink goes into action. Finally they claim credit for the impact.
</p>
<p>
Another variation of the theme: Someone builds in secret a lot of machines with fairly general seeming purpose but without actual function. Then they wait for a near catastrophes. Maybe there’s a hurricane that changed course just before it hit the shore. They reveal one of the machines that happened to be nearby and claim to have used the hurricane as a beta test for the hurricane-averting machine and sell the impact from it.
</p>
<p>
Maybe people will come up with more realistic variations on this theme. When that happens, the problem can probably be addressed through preregistration of impact certificates and buyers penalizing any lack of preregistration. (Preregistration may be as simple as issuing an impact certificate before the impact has happened, which should be the norm anyway.)
</p>
</td>
</tr>
</table>
<p><strong>Noncausal decision theory and Kalai bargaining solution.</strong> There is a risk that people will default to assuming that other market participants are causal decision theorists. This would preclude gain from acausal moral trade. We’re unsure what decision theory is best to recommend here, but it is probably not causal decision theory. The bargaining solution is also something we’re unsure about. We don’t know which bargaining solution to pick here, <a href="https://longtermrisk.org/coordination-challenges-for-preventing-ai-conflict/">but not specifying any seems worse than specifying a random one</a>. Resource monotonicity seems to us like an important criterion for the general acceptability of a bargaining solution in a rapidly growing market (the Nash bargaining solution doesn’t ensure it), maybe more so than scale invariance (the weak point of the Kalai one). Independence of irrelevant alternatives seems superficially important to us (the Kalai-Smorodinsky solution doesn’t ensure it), but we haven’t thought about this in detail and would greatly appreciate feedback on this decision.</p>
<p><strong>Summary.</strong> This definition of the Impact Attribution Norm addresses (hopefully successfully) the following problems:</p>
<ol>
<li><strong>Impact is not naturally uniquely owned.</strong> It addresses this by relying on cultural norms and laws around responsibility, such as legal ownership, authorship, etc.</li>
<li><strong>Issuers may generate one good at the expense of another.</strong> This is addressed by tying the evaluation to an aggregate of <em>all</em> public goods.</li>
<li><strong>Markets incentivize risky gambles.</strong> It addresses this by defining the Impact Attribution Norm as the minimum of <em>ex ante</em> and <em>ex post</em> expected impact.</li>
<li><strong>Counterfactuals are impossible and arbitrary.</strong> This is addressed by stipulating a particular way of thinking about the counterfactuals that relies on the subjective expectation at two points in time of an observer that knows no specifics about the case.</li>
<li><strong>Buyers are vulnerable to extortion.</strong> This is addressed by stipulating a counterfactual (see above) that is very hard for an individual to influence.</li>
<li><strong>This definition may be ignored.</strong> It addresses this by asking issuers to defend their impact in terms of the definition and by protecting buyers from extortion if they proactively signal adherence to the definition.</li>
</ol>
<h4 id="timeline"><a class="toclink" href="#timeline">Timeline</a></h4>
<p><img alt="Impact Attribution Norm timeline" src="/images/toward-impact-markets/attributed-impact-timeline.png"></p>
<p>This is a sample timeline of how an issuer may produce impact. Period 1 is when the impact is issued and generated, in whatever order. Period 1 separates Period 0, where the real and the counterfactual worlds haven’t diverged, from Period 2, where they have diverged and continue to diverge.</p>
<p>On both sides of Period 1 there is one systemic observer, who is imagined/simulated by the investors. In Period 2 there is also the actual investor. (We’ll use <em>simulate</em> over <em>imagine</em> to convey that the goal of the mental simulation is accurate prediction and not, say, entertainment.)</p>
<p>So this is what happens in temporal order:</p>
<ol>
<li>The prospective issuers consider that a systemic observer would typically expect people like them to do <em>stuff</em>. So they surmise that future buyers would also agree that a systemic observer would expect that. But if they did a <em>thing</em> instead, a systemic observer would be surprised.</li>
<li>The prospective issuers iterate through a few options for <em>things</em> to do, and eventually settle on one. That <em>thing</em> is a great <em>thing</em> because it generates morally relevant preference satisfaction for some beings without undermining the preference satisfaction of others. They implement it and document it in an impact certificate alongside a lengthy defense.</li>
<li>A month has passed and a few buyers get interested in the impact certificate.<ol>
<li>They by and large agree with the issuers on the expectation of the systemic observer in Period 0. This systemic observer is virtually the same as the one in Period 2. They imagine themselves in the shoes of the issuers in Period 0, and find that they might have predicted a few additional benefits and drawbacks but nothing substantial, and agree with those that the issuers have documented in the impact certificate, and with their conclusion that the <em>thing</em> appears like it would be net positive in expectation.</li>
<li>Then they return into their own shoes where they have the invaluable benefit of hindsight. From this vantage point they realize that a few of the drawbacks they were retroactively worried about had not manifested but that some of the best-case scenarios have also become less likely. All in all, the thing still appears to be net positive in expectation, now with slightly reduced variance. They can’t quite tell which expectation, <em>ex ante</em> or <em>ex post</em>, is lower, but they think that both are positive and fairly close together, so they don’t care much and just bid on the impact certificate according to their budgets.</li>
</ol>
</li>
</ol>
<p>Note that strictly speaking the buyers also simulate the issuers because the certificate will be phrased in natural language which is highly ambiguous without knowledge of the (likely) intent of the author. We’ll omit this simulation in the graphs since it’s quite intuitive.</p>
<p><img alt="Impact Attribution Norm counterfactuals" src="/images/toward-impact-markets/attributed-impact-counterfactuals.png"></p>
<p>That’s all a bit of a simplification because many prospective buyers (let’s call them investors) want to make a profit. So they’ll perform further evaluations to understand whether they may know more or have more accurate information about the real expected impact than other investors and whether the other investors will attain that knowledge too, be it automatically over time or through a publication by other investors. Occasionally they may also seek to prove that the issuer had private information already in Period 0, which they pretended not to have, or conversely, whether something that is widely assumed to have been known at the time actually wasn’t knowable.</p>
<h4 id="example"><a class="toclink" href="#example">Example</a></h4>
<table>
<tr>
<td>
<strong>Example.</strong> The issuer has sold 500 pins with cute designs and animal rights messages on them at a convention. They’ve donated the proceeds to an animal rights charity. They now sell the impact from this action.
<ol>
<li>The issuer thinks that the donation of the proceeds is the main driver of the positive impact, so the certificate text focuses on the money.</li>
<li>But one investor thinks that most investors are unexcited by the donation because they think it just displaces donations from corporate partners of the charity who don’t have set <span class="caps">CSR</span> budgets.</li>
<li>This one investor thinks that most investors think that reaching 500 people with the animal rights messaging is the main driver of the impact, in particular because the designs were actually created by well-known members of the community that the issuer sold the pins to, something that is widely known but that the issuer didn’t mention.</li>
<li>But this one investor also freshly found photos from the convention and noticed that the issuer had sold five different designs, and that, in group shots, a lot of people could be seen wearing multiple or even all five pins. The investor shorts the certificate before announcing their finding because they expect that the issuer had reached much fewer than 500 people.</li>
</ol>
</td>
</tr>
</table>
<p>This example aims to clarify that:</p>
<ol>
<li>Issuers and investors can disagree on the path to impact, yet it can still be valuable <em>ex ante</em> and <em>ex post</em>.</li>
<li>The certificate description is never going to describe reality fully (and critical omissions can be unintentional), so it should contain as much information as possible that will help buyers do their own research to fill in gaps – such as finding photos from the convention.</li>
<li>The issuers and all investors can have different ideas of the actual and the counterfactual world histories from the perspective of the systemic observer.</li>
<li>Only when an investor can anticipate a price-relevant update of some of the other investors can they make money. If an investor knows more than the others but has no way of convincing them, they can’t scoop them. (But in this case they could because they had photographic evidence.)</li>
</ol>
<h4 id="pricing-impact"><a class="toclink" href="#pricing-impact">Pricing Impact</a></h4>
<p>Attribute Impact itself has no particular unit. One could pick a benchmark project – such as $1 distributed through GiveDirectly – and measure projects in relation to it, but we have not attempted this.</p>
<p>But that is not important for the market to function. The fair market price of an impact certificate can be determined through an auction. Then afterwards everyone is free to have their opinion on whether the certificate is a bargain or is overpriced, but either way, there will be a price.</p>
<p>The way the auction could work (so that it requires minimal time investment from the retro funders) is as follows:</p>
<ol>
<li>A retro funder wants to incentivize work on Vingean reflection.</li>
<li>The retro funder announces a prize pool of $1k for anyone who makes progress on the problem.</li>
<li>They receive no serious or interesting submissions.</li>
<li>The retro funder announces an increase of the prize pool to $10k.</li>
<li>They receive no serious or interesting submissions.</li>
<li>The retro funder announces an increase of the prize pool to $100k.</li>
<li>The retro funder receives several interesting submissions and buys shares in some of them for most of the $100k budget.</li>
</ol>
<p>This way the retro funder has found out that it took a prize pool of $100k to incentivize the submissions. Maybe each submission received $30k, and years later experts will conclude that they created $30b in social benefits. But that is not important for the functioning of the market if the researchers are ready to accept $30k.</p>
<h4 id="limitations"><a class="toclink" href="#limitations">Limitations</a></h4>
<p>Trivially, the Impact Attribution Norm is only powerful to the extent that it is adopted by the market, in particular by the retro funders. That means that it cannot prevent people who have a bit of money from going on Twitter and announcing “I will give you $1m if you realize this very risky project.” They can do that already. We will need to rely on curation rather than the Impact Attribution Norm to prevent them from doing it on our marketplaces.</p>
<p>Second, retro funders need to be transparent about their use of the Impact Attribution Norm and they need to understand it well enough that investors trust them to apply it well. If they fail to signal that sophistication, investor may still end up investing into overly risky projects under the assumption that they can fool the retro funders. This need not happen on purpose. If risks are sufficiently well hidden, they may remain hidden from investors and retro funders for the same reasons. (But in that case it is likely that prospective funding wouldn’t fare any better.)</p>
<p>Finally, something that can exacerbate the second problem is that investors may believe that retro funders will be systematically biased to overestimate the <em>ex ante</em> expected value of projects that have already turned out well, be it through sheer luck. Phil Trammel has proposed that issuers who preregister their impact certificates can start prediction markets for their impact in accordance with the Impact Attribution Norm alongside them. Then retro funders can debias themselves by looking at the history of the predictions, in particular the predictions from before any actions were preformed.</p>
<h3 id="responsible-retroactive-funders"><a class="toclink" href="#responsible-retroactive-funders">Responsible Retroactive Funders</a></h3>
<p>The main responsibility for the health of the market is upon the retro funders. Hence they need to be sufficiently wise to steer the market and sufficiently involved to notice when something is going wrong in the market.</p>
<p>First, a retro funder needs to ask themselves:</p>
<ol>
<li>How much time are we spending on finding giving opportunities whose interventions are effective and that are implemented by capable teams? How much is it worth to us to not have to evaluate the teams? How much are we willing to offer to investors to incentivize them to do the work for us?</li>
<li>Do we think that in the future people will value the impact that we’re interested in even higher, so that we can one day resell it?</li>
<li>Do we want to see more highly effective charities, and how much are we willing to pay to incentivize their founding?</li>
</ol>
<p>If these questions generally turn out that the retro funding will be worth it, there are further consideration:</p>
<ol>
<li>Investors and founders will rely on our future funding (not that they’ll get it necessarily but that it’ll still be there). Can we commit a certain budget firmly, and if we need to discontinue it, can we afford a few years of grace period between the announcement and the discontinuation?</li>
<li>Can we make it sufficiently clear to founders and investors what the impact is that we want to see, so as to prevent people from expending great effort on potentially harmful projects that we would never pay for?</li>
<li>Are there maybe even metrics that can be used as rough guides by founders to check whether they’re on the right path without great risks of Goodharting or moral defection?</li>
<li>Do we have the resources for occasional calls with founders or investors who want to be sure that they’re starting the right sort of project?</li>
<li>Do we have resources to occasionally write a blog post to clarify our interests when we see investments being poured into uninteresting projects?</li>
</ol>
<p>Finally the hardest question:</p>
<ol>
<li>Are we ready to commit to being as impartial as possible about our impact evaluations? Especially in cases where we have strong moral feelings in one direction or the other, can we be trusted to still reward only that project that successfully compromises between the competing interests? Or put differently: Can we be trusted to price projects in proportion to the moral gains from trade that they generate?</li>
</ol>
<p>If someone is only interested in providing retro funding but not in doing all the work associated with being a responsible retro funder, then the next section will be interesting, as they can simply contribute to the “Pot of Money” and leave the work to the jury of the pot.</p>
<h3 id="pot-of-money"><a class="toclink" href="#pot-of-money">Pot of Money</a></h3>
<p>If there are sophisticated, responsible retro funders on a market and if they have a lot of money to give away compared to the volume that is transacted on the whole market, then everything is well and good. But even if either (1) there is no retro funder, (2) there are sophisticated, responsible retro funders but they are overpowered by rogue ones, or (3) there are only sophisticated, responsible retro funders but speculation on the market has developed its own rules, then the influence of the good retro funders will be insufficient to keep the market healthy. This is where the pot comes in.</p>
<p>The pot is a reference to the pot in poker, a vehicle for investors to “gamble” – except that no intentional randomness is involved, only the imperfections of impact evaluations and of predicting the future. For simplicity, we’ll copy the example from earlier but rewrite it to the case where the only retro funder on the market is the pot.</p>
<p>There are now four components:</p>
<ol>
<li><strong>Projects:</strong> The charitable projects that are waiting to be realized.</li>
<li><strong>Supporters:</strong> These include founders, investors, advisors, et al.</li>
<li><strong>Jury:</strong> An expert panel of sophisticated altruists, who become the source of truth when it comes to what has been impactful.</li>
<li><strong>Pot of money:</strong> When the sophisticated altruists publish their decree, this pot of money is dispensed to the supporters.</li>
</ol>
<p>At first, the pot of money likely has to be managed separately. The charities are not a good fit for managing it, and the funds may not be sold on the whole system from the start, or not sufficiently for all the overhead it would involve for them.</p>
<p><img alt="Impact market with pot" src="/images/toward-impact-markets/impact-market-with-pot.png"></p>
<p>Let’s suppose again that the Against Malaria Foundation (<span class="caps">AMF</span>) wants to do a distribution of long-lasting insecticide-treated bednets with its old distribution partner Concern Universal (<span class="caps">CU</span>):</p>
<ol>
<li><span class="caps">AMF</span> ascertains that the jury of the pot is (and maybe other retroactive funders are) still interested in more net distributions.</li>
<li>The <span class="caps">AMF</span> staff decide how they want to split the Impact Attribution Norm from the distribution among themselves, the legal entity <span class="caps">AMF</span>, all their partners, contributors, and advisors, and their funders. These decisions can be adjusted in later negotiations. (we’ll treat <span class="caps">AMF</span>, the legal entity, and all its staff as just “<span class="caps">AMF</span>” in the following to save space.)</li>
<li><span class="caps">AMF</span> decides on the details of the auction process, that is, what fraction of the profits from each sale should go to <span class="caps">AMF</span> and what the minimum percentage raise is between bids.</li>
<li><span class="caps">AMF</span> creates an impact certificate.</li>
<li><span class="caps">AMF</span> contacts <span class="caps">CU</span>, negotiates their split with them, and sends them their fraction of the impact certificate. (we’ll assume here that the distribution partner has its own source of seed funding just to make its case interestingly different from <span class="caps">AMF</span>’s.)</li>
<li><span class="caps">AMF</span> contacts a venture capital firm. The <span class="caps">VC</span> agrees to fund the distribution. They negotiate the split, and <span class="caps">AMF</span> sends them their fraction of the impact certificate.</li>
<li>Everyone contributes to the pot as they see fit (or doesn’t), and the pot records who contributed how much to bet on which certificate.</li>
<li>A year later the distribution gets completed successfully.</li>
<li>The jury notices the successful distribution, is ecstatic, and buys fractions of the certificate from all holders who have contributed to it. <span class="caps">AMF</span> has no work with this: Any holder just has to give their certificate to the pot (or a fraction thereof), which ascertains that they’ve previously bet on it, and the pot buys it from them at a higher price.</li>
</ol>
<p>The core of the system should, in our opinion, not depend on any additional retroactive funders because (1) they may not be interested in the market until they see it work; (2) their funding is finite and independent of the impact market; and (3) they may not have enough in-house experts so that they want to defer to our jury anyway, effectively adding to the pot.</p>
<p>That said, external retroactive funders would make the market attractive for many more participants!</p>
<p>When it comes to the size of the reward, we currently favor the following system:</p>
<ol>
<li>Everyone contributes whatever they want but has to tie this “bet” to a particular impact certificate.<ol>
<li>This removes incentives that investors who don’t like the pot might otherwise have to circumvent the market.</li>
<li>They need to bet on a particular certificate rather than just any that they own so that they can’t just buy minimal fractions of countless certificates at random, which would add little to no wisdom to the market and wouldn’t differentially benefit projects with real promise.</li>
</ol>
</li>
<li>When the jury approves of a certificate, the pot sets a budget for the purchase and buys fractions of the certificate from each buyer according to the product of (how much of the certificate they hold) * (their contribution to the pot for that certificate). 3. That way people are rewarded for contributing to the pot and the project, and those who fail to do either can’t sell to the pot. 4. They may of course do so on purpose because they want to keep the certificate as a long-term investment or to sell it to another retroactive funder.</li>
</ol>
<p>This still glosses over a lot of details. They are easier to explain in the context of the auction platform.</p>
<h4 id="auction-platform_1"><a class="toclink" href="#auction-platform_1">Auction Platform</a></h4>
<p>We want to create an auction platform to streamline the whole process. (<a href="https://www.impactcerts.com/">You can keep up-to-date on the progress here.</a>) The current model is a type of auction where the impact certificate is always owned by the highest bidder and so constantly changes hands. Meanwhile the profits from each sale are split between the seller and other parties, typically the issuer and the Pot. We’ve covered it earlier in the subsection Auction Platform.</p>
<p>In this model the issuer determines what fraction of the profits of each sale should go to them. Here’s the rough timeline of the trades.</p>
<p><img alt="Pot timeline" src="/images/toward-impact-markets/pot-timeline.png"></p>
<p>Here an overview of the accounts of each participant at each step in the process (<a href="https://docs.google.com/spreadsheets/d/1Rnexsgn337zBSryNVn9k3IvUilApBEEevFd6T42-z8I/edit#gid=0">you can find the spreadsheet here</a>):</p>
<p><img alt="Pot table" src="/images/toward-impact-markets/pot-table.png"></p>
<p>Various parameters:</p>
<p><img alt="Pot parameters" src="/images/toward-impact-markets/pot-parameters.png"></p>
<p>In this scenario <span class="caps">AMF</span> made its profit share relatively small (1%) to encourage investment. The minimum raise (the step between bids) is at an arbitrary value of 10%. The minimum bid is set such that <span class="caps">AMF</span> won’t raise amounts of money that are insufficient to fund the net distribution. The nets and salaries on <span class="caps">AMF</span>’s side will be $500k, and the salaries and miscellaneous costs on <span class="caps">CU</span>’s side will be $500k as well. (Traditionally, <span class="caps">AMF</span>’s strategy has been to fundraise for the nets but to leave their distribution to other organizations.)</p>
<p>The deal with <span class="caps">CU</span> is described above. It’s now that 30% of the certificate are issued. The deal with <span class="caps">VC</span> 1 is also described above. This is the first bid on the market, so <span class="caps">VC</span> 1 gets to buy them at the minimum bid price. <span class="caps">VC</span> 2 buys the shares from <span class="caps">VC</span> 1 right away, but this time the minimum raise and the profit share apply, that is, the valuation increases by 10%, and 1% of <span class="caps">VC</span> 1’s profit goes to <span class="caps">AMF</span> instead.</p>
<p>Next, all three certificate shareholders contribute to the pot. <span class="caps">AMF</span> and <span class="caps">CU</span> add an arbitrary $50k to the pot whereas <span class="caps">VC</span> 2 tries hard to maximize the weight it’ll have when the pot buys the certificate from them again, but without selling more of the certificate than it owns. The prices in the ratchet auction can only go up, so even if it could borrow certificate shares from (say) <span class="caps">AMF</span>, that would not be profitable.</p>
<p>We assume here that (1) <span class="caps">VC</span> 2 is certain that the pot will later buy the certificate, (2) can wait for and then observe the pot contributions of all other shareholders, and (3) knows how much money is going to be in the pot and how much it will budget for this purchase. These are all slightly tricky assumptions: Assumption 1 is never going to be completely true; assumption 2 can lead everyone to wait until the last possible second to make their contributions; and assumption 3 is probably again somewhat difficult in practice.</p>
<p>If the jury were to deem the net distribution to have been ineffective, the scenario would end here with an extra $174k in the pot and all organizations having lost money.</p>
<p><img alt="Pot allocation" src="/images/toward-impact-markets/pot-allocation.png"></p>
<p>Finally, though, the pot decides on a budget of 20% and a bid 20% above the last one. The resulting valuation is more than the budget, so that the budget doesn’t need to be capped. The weight is proportional to the value that the shareholder has in the certificate times the size of the pot contribution. (<a href="https://docs.google.com/spreadsheets/d/1Rnexsgn337zBSryNVn9k3IvUilApBEEevFd6T42-z8I/edit#gid=0">You can find the spreadsheet here.</a>)</p>
<p>The column “Share” is the share in the impact certificate that the respective party owns. The “Value” is the value of the share according to the latest valuation of the certificate. The “Contrib. [-ution]” is the amount that the party has freely chosen to contribute to the pot. The “Product” is the product of the value of the share and the contribution – it doesn’t map to anything in particular in the real world, but the weight is proportional to it. The aforementioned “Weight” is the fraction of the budget of the pot that the pot uses to buy certificates from the respective party. The “Cost” is the resulting total dollar volume of the purchase. The “Size” is the fraction of the certificate that changes hands.</p>
<p>The final budget for each individual purchase may not be fully used up if any of the shareholders doesn’t own enough shares (though this case doesn’t happen here and is not part of the calculations in the spreadsheet because it made it harder for us to find errors in it).</p>
<p>The result is that the pot spent about $1.86 million and every organization profited, either monetarily like the <span class="caps">VC</span>, altruistically like <span class="caps">CU</span>, or both like <span class="caps">AMF</span>.</p>
<p>Notes:</p>
<ol>
<li>It took us a while to come up with a scenario that “works,” in the sense that in particular the VCs make a profit (or else they wouldn’t invest) without making the pot or the fraction that it invests unrealistically big or its bid unrealistically high. This indicates that the parameters of the pot budget will need a lot of tuning and that issuers and investors will be well-advised to carefully think through their funding scenarios to make sure that they’re realistic.</li>
<li>This difficulty is exacerbated since the pot will probably require holders to have held the certificates about a year earlier already to prevent frontrunning of pot purchases. Investors who hold certificates for a year will need to make much greater profits off them to make up for the counterfactual uses of their funds. Unless another investor buys it from them, and they buy it back much later.<ol>
<li>The current auction system doesn’t support or would make it at all lucrative, but under different conditions, holders could lend their certificates to short-sellers to earn interest on them.</li>
</ol>
</li>
<li>But this may be a feature rather than a bug in the short run: The distribution mismatch problem below is alleviated if (1) the Impact Attribution Norm has time to be adopted as the Schelling point of impact evaluation on the market, and (2) there are few purely profit-oriented investors. We may grow a safer, more altruistic community if the market is not of interest to anyone unless the social bottom line also counts into their personal profits.</li>
<li>There’s a questionable dynamic where some investors benefit if other investors into the same certificate pay less into the pot. We don’t yet know whether there are setups where it’s profitable for all investors if one of them pays other investors outside the market in exchange for them withholding money from the pot. But if there are such scenarios, then that may be bad for the pot.</li>
<li>There’s another bad dynamic, maybe a negative externality, where sophisticated for-profit investors are incentivized to advertise the system to unsophisticated investors and potentially mislead them. They may promote the impact certificates of highly ineffective charities to unsuspecting people as investment vehicles to use in this system. These people would enlarge the pot without having any chance of ever winning it. Meanwhile all the misinformation may make it harder to find high-quality charity reviews for people who don’t already know where to find them.</li>
<li>Finally, we’re unsure how to prevent insider-trading by jury members, but there are probably already mechanisms for that, such as that jury members need to be many and not know each other. There are probably time-tested solutions to this problem that we just need to find out about.</li>
</ol>
<table>
<tr>
<td>
<strong>Prediction markets.</strong> Something we want to think about more is whether the Pot should instead be a scalar prediction market predicting something like the percentage of endorsement of the project from the jury. Investors may be hesitant to pay into the Pot if it’s unclear for how long their money will be locked up in there even if they win their bet – a prediction market would address that.
</td>
</tr>
</table>
<h3 id="auditors"><a class="toclink" href="#auditors">Auditors</a></h3>
<p>Our ideas for how to structure the ecosystem of auditors are still inchoate.</p>
<p>Auditors will need to audit impact at least twice. First there needs to be an audit that issuers need to get before they fundraise from investors. Instead of having to plow through all the impact certificates on the market, investors can then just consider ones that have been audited in this way. Such an audit needs to consist of a few basic checks, such as whether the argument for the Impact Attribution Norm makes structural sense (e.g., is not autogenerated gobbledy-gook), is concrete enough that it can be audited at all, and crucially that the impact is not being double-issued.</p>
<p>The second audit needs to also ascertain that the impact has happened. A lot of forms of impact are easily verifiable – an article author, for example, can just link to the article they’ve written.</p>
<p>Checking that impact has not been double-issued is hard in full generality, but it can be mostly solved if all auditors publish all their audits, and all auditors know about all other auditors and check their lists of published audits. Then they can’t prove that impact has not been double-issued at all, but if it has, at least the other issue does not have an audit. This can still break down in cases where the double-issue is sufficiently subtle, but it should catch a lot of more obvious cases.</p>
<p>One thing that would be helpful for this job is a standardized format that all auditors use to publish their audits. Maybe a <span class="caps">JSON</span> blob on their websites. Then a search engine can extract structured data from there, and the search will be much easier for all of them.</p>
<p>One idea how to incentivize auditors and ameliorate risks from harmful projects too is to require issuers to be insured when they issue on the platform. If it is later decided that the project was <em>ex ante</em> harmful and it is removed from the platform, the insurer needs to pay a fine on behalf of the issuer. If a project actually predictably backfires, a larger fine is due. This incentivizes insurers to deny insurance for projects that are not safe enough. A problem is hidden in the passive-voice “If it is later decided,” because whatever process decides this has to be trusted by the insurers. That promises to be a difficult problem to solve.</p>
<h3 id="shorting"><a class="toclink" href="#shorting">Shorting</a></h3>
<p>All of our solutions try to make it uninteresting to issue or to invest into impact that is likely net negative, but another option is to require investors to deposit collateral as insurance against the case where the impact turns out negative. The catastrophes that we’re worried about are of the civilization-ending type, so no amount of collateral can realistically be enough, but it may still increase the friction a bit when it comes to investing into potentially net negative impact.</p>
<p>One idea is that we expect that in the long run most people will be agnostic about which impact certificate (e.g., which net distribution of <span class="caps">AMF</span>) they invest in because they’re not close enough to the specifics of the intervention to see the differences between them. They’ll just want to invest in “<span class="caps">AMF</span>.” So I find it plausible that there will be demand for <a href="https://www.bitmex.com/app/perpetualContractsGuide">perpetual futures</a> (see also “<a href="https://www.coindesk.com/learn/what-is-a-perpetual-swap-contract/">What are perpetual swaps?</a>”) that track something like the market capitalization of all impact certificates issued by <span class="caps">AMF</span>. (That’s not an interesting market in the case of the ratchet auction where prices can only go up, but it may be with other auction mechanisms.)</p>
<p>To get exposure to this perpetual future, investors could deposit collateral and open long positions on it. If they use too much leverage and the price decreases too much, they can lose their collateral.</p>
<p>But the other benefit – hinted at in the subsection title – is shorting. Shorting can help to price in the expectation of some investors that retro funders will not be interested in the impact or not at that price. It doesn’t by itself have the same upside potential as longing, but if a short-seller regularly reinvests their collateral when it increases due to decreasing prices, they can maintain a constant leverage and gain a similar exposure to nonanthropic downsides – at least before adjusting for the risks.</p>
<p>This sounds complicated, but there are already “hedge” tokens that automate the process. These could be applied to the perpetual future and thus give investors a simple, safe, and low-effort way to get negative exposure to overpriced impact.</p>
<p>However, we’ve seen these tokens fail especially during crashes, when you want to profit from them most. My hypothesis is that there is so little buy-side liquidity during crashes that when the token smart contract tries to reinvest its PnL to increase the leverage, it has very few buyers to sell to and so has to sell at low prices, prices close to where the market is crashing to anyway. They probably work better in markets where the prices decrease gradually.</p>
<p>All in all, it’s much too early for this system. There’ll first need to be highly liquid markets and popular issuers of many impact certificates before it makes sense to think about this system more.</p>
<h3 id="expose-certificates"><a class="toclink" href="#expose-certificates">Exposé Certificates</a></h3>
<p>Shorting becomes particularly powerful in combination with curation: Shorting incentivizes market actors to open their short positions and then to publish exposés to get others to divest (too). But these exposés are invaluable to the curation team in and of themselves because they help them spot their own oversights.</p>
<p>This particular function of shorting can be achieved more simply: The action of helping the curation team spot net negative projects is one that is usually quite robustly positive, so that exposés can become quite valuable if turned into impact certificates. That way the impact market creates its own “bug” bounty system. Retro-funding of these certificates has to be funded by a retro funder that is interested in the success of the market.</p>
<h3 id="targeted-marketing"><a class="toclink" href="#targeted-marketing">Targeted Marketing</a></h3>
<p>Especially in the early days when the market is still young it’ll be important who to market it to. The above section on responsible retroactive funders has already clarified why it is critical to recruit the sorts of funders who will let the markets thrive rather than ones that’ll lead them astray. That is one problem that can be partially addressed by reaching out specifically to the retro funders who we think will do a good job of it.</p>
<p>But the same is true of charities and investors, just to a lesser extent. The charities on the market need to be the ones that produce good/s that is/are interesting to the sorts of retro funders we want to have on our markets. The reason is different in the case of the investors: If we recruit investors whose motivations are mostly altruistic, and we then notice that our markets create bad incentives for them, then we can warn them of these incentives, and it’ll be in their altruistic interest to resist them. But if we run into bad incentives and have unaltruistic investors, we don’t have this option.</p>
<h2 id="current-work"><a class="toclink" href="#current-work">Current Work</a></h2>
<p>Protocol Labs has organized two iterations of the conference <a href="https://www.youtube.com/playlist?list=PLhuBigpl7lqtMdPkejuo3mHdLFX53ftXJ">Funding the Commons</a>. “Funding Public Goods – Algorithms and Mechanisms” by Vitalik Buterin and “S-Process Funding” by Andrew Critch are particularly interesting because they highlight important mechanisms, and “Quadratic Funding on Gitcoin” by Kevin Owocki, “Retroactive Public Goods Funding Experiment 1” by Karl Floersch, and “Impact Evaluators” by Evan Miyazono are particularly interesting because they give insight into the work of the actors who are currently active in the space. Finally, “Impact Certificates and Impact Markets” by Owen Cotton-Barratt is the talk that we collaborated on and that we already linked above. <a href="https://www.youtube.com/watch?v=NTIAdn0Oms8&list=PLhuBigpl7lqvngC9oNecjfWMqFucr5GvG&index=16">We also had a talk at the second Funding the Commons conference.</a> We’re planning to summarize some of the talks in a separate article.</p>
<p>These talks highlight our work, the work of the Ethereum Foundation, the Survival and Flourishing Fund, Gitcoin, Optimism, Protocol Labs, and many more. Beyond that, Sentience Research, the <span class="caps">EA</span> Hotel, and Giveth have been following the field with interest.</p>
<p>We’re also in touch with <a href="https://npxadvisors.com/"><span class="caps">NPX</span> Advisors</a> who’ve been running a quasi–impact market with greater amounts of funding but a very small set of funders, investors, and charities in the <span class="caps">US</span>, and who conceive of “impact certificates” as debt securities rather than impact equity securities.</p>
<p>We think these are viable approaches because they test impact markets in fields that are safer than some of the main <span class="caps">EA</span> cause areas. That way they either work with participants who are highly informed when it comes to financial markets or within fields where it is hard for us to see how impact markets may backfire catastrophically.</p>
<p>Note that we – Kenny Bambridge, Matt Brooks, Dony Christie, and I – can use funding: $100k would enable us to invest more time into this project (about 70% more in my case) and any amount around $10k to $1m would would allow us to conduct retro-funding experiments. This money could be merely committed to these experiments and wouldn’t need to be transferred in advance. We would then make grant recommendations according to guidelines that we would like to publish in advance in order to make the recommendations as predictable as possible. Please get in touch if you’d like to help with funding, advice, or otherwise.</p>
<h2 id="acknowledgements"><a class="toclink" href="#acknowledgements">Acknowledgements</a></h2>
<p>Many thanks to my collaborators Kenny Bambridge, Matt Brooks, and Dony Christie! Thanks to Katja Grace and Paul Christiano for the initial inspiration. Thanks also to Owen Cotton-Barratt, Ofer Givoli, Jay, mqp, bryjnar, Mako Yass, and Justin Shovelain for discussions and ideas. You’ll each receive 1% of the impact of this article. Thanks to Protocol Labs for the Funding the Commons conferences.</p>SquigglyPy: Alpha Version of Squiggle for Python2021-06-15T00:00:00+00:002021-06-15T00:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2021-06-15:/squigglypy-alpha.html<p><a href="https://github.com/Telofy/SquigglyPy">SquigglyPy</a> uses sampling to allow you to do math with arbitrary probability distributions in two dimensions – e.g., you can plot a thousand traces of how economic growth might pan out over the next couple of millions of years.</p><h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<p>One of my projects in 2020 was a software for doing math with probability distributions – SquigglyPy. Sort of like <a href="https://www.getguesstimate.com/">Guesstimate</a> but with the ability to estimate probabilistic trajectories over time or some other independent variable.</p>
<p>I had seen a number of researchers either write code for Monte Carlo simulations from scratch or create models based on mere point estimates. My software would save the first group some time and help the second group get a better sense of the degree of uncertainty in their model.</p>
<p>My vision for it was of course greater than that. I hoped to inspire people to use the software to model various interesting effects in the world in the various ways that seem most useful or parsimonious to them, to publish these various models online, and then to recombine existing models to create ever more comprehensive models. Thus people would loosely collaborate to create a faster and faster growing ecosystem of Baysian models of the world – along the lines of how open source software allows developers to develop new software faster and faster by recombining increasingly powerful chunks of existing software.</p>
<p>My hope was that this would be done primarily by researchers in global priorities research and would lead, over the course of decades, to an improved allocation of resources.</p>
<p>I’m putting this project on hold for now because (1) I updated downward on the time we have left till transformative <span class="caps">AI</span> emerges, which had complex effects on my prioritization and (2) because I didn’t feel like it had as much traction as other projects of mine. That may change of course, and I may pick it up again at a more opportune time.</p>
<p>Meanwhile I’m happy to add other developers to the repository or to accept pull requests. In particular, it would be great to improve:</p>
<ol>
<li>the tests, to catch mathematical errors,</li>
<li>the sensitivity analysis, which is currently a hacky sketch on Starboard, and</li>
<li>the documentation, later, when the code is better tested.</li>
</ol>
<p><strong>Warning:</strong> When using SquigglyPy, please take great care to sanity-check all your results! Hardly anyone has used the software yet and the tests are currently better at catching broken code than subtly wrong math.</p>
<h2 id="squiggle"><a class="toclink" href="#squiggle">Squiggle</a></h2>
<p>My vision was similar to Ozzie Gooen’s (<a href="https://quantifieduncertainty.org/software">Quantified Uncertainty Research Institute</a>), which is part of the reason I chose the name SquigglyPy. Ozzie developed a similar software in ReasonML, which also runs in the browser. It’s the ReasonML implementation of his <a href="https://www.lesswrong.com/posts/i5BWqSzuLbpTSoTc4/squiggle-an-overview">Squiggle language</a> whereas SquigglyPy is the Python implementation.</p>
<p>I makes senses to me to implement Squiggle in a few different languages:</p>
<ol>
<li>Any language that compiles to JavaScript or WebAssembly has the advantage that it runs in a browser. The right <span class="caps">UI</span> and an efficient implementation can give people the ability to experiment and explore rapidly. Models might be shared with others through repositories on the same server to save the time it would take to download and install a software from a popular existing platform like <span class="caps">NPM</span> or PyPI.</li>
<li>Languages like R, Julia, and Python have the advantage that the main target demographic (as I currently imagine it) uses them and has built a wealth of statistical tooling for them already.</li>
<li>Finally, Python has the advantage that there is already a project that has made <a href="https://pyodide.org/en/stable/">Python run in the browser</a>, combining all advantages. (It also has the advantage that I’m very familiar with Python.)</li>
</ol>
<p>But there are also drawbacks to using Python in the browser:</p>
<ol>
<li>Pyodide is a very small, young project and may not survive.<ol>
<li>But chances are another project will pick up the slack because Python is popular and WebAssembly seems to me like the obvious next step in the technological development of websites.</li>
</ol>
</li>
<li>Many statistical libraries use some C or Cython extensions, and so can’t just be installed through Micropip. Instead the Pyodide team (or whoever wants to use them) has to create a WebAssembly version of them. The Pyodide team at least is too small to support a lot of WebAssembly builds.<ol>
<li>But over the years it may become more common for library authors to produce WebAssembly builds themselves or for companies to publish any such builds that they’re using for their products.</li>
</ol>
</li>
<li>The Python interpreter takes a moment to load, which may degrade the user experience compared to a JavaScript implementation.<ol>
<li>But it loads only once per page load and then stays in memory.</li>
</ol>
</li>
</ol>
<p>I feel like the advantages outweigh the disadvantages here, but I could easily be wrong in any of various way:</p>
<ol>
<li>The world of stats may transition toward Julia or some other more specialized language than Python,</li>
<li>Python may lose market share to Rust or some other similar language,</li>
<li>More statistical tooling may be built for JavaScript or even WebAssembly directly, or</li>
<li>Any of surely countless unforeseen vicissitudes may eventuate.</li>
</ol>
<h2 id="risks"><a class="toclink" href="#risks">Risks</a></h2>
<p><a href="https://impartial-priorities.org/epistemic-hazards.html">I’ve detailed a lot of risks in “How Might Better Collective Decision-Making Backfire?”</a> They fall into the general buckets of risks that need to be addressed immediately and risks that can be addressed later or reacted to only if they manifest.</p>
<h3 id="immediate"><a class="toclink" href="#immediate">Immediate</a></h3>
<ol>
<li>Programming errors.<ol>
<li>This is the primary reason I’ve put off writing this post and why now I only publish it on my blog where few people will see it. It’s very likely that there are mathematical errors in some parts of the code that will skew any results. This might lead to disastrous misprioritization.</li>
<li>The main way to address this issue is to improve the tests and to get people to use the software for many typical but not absolutely crucial purposes. These people need to be closely in touch with the developers and take their roles as beta testers seriously.</li>
</ol>
</li>
<li>Version conflicts and path dependent legacy software.<ol>
<li>The Squiggle language will need a clear, versioned language definition so that implementations of models in different languages are interoperable, either to the point where you can copy-paste the code or at least to the point where the real Squiggle code can be turned into a language specific version automatically. (Along the lines of how the 2to3 tool turns Python 2 code into Python 3 code.) Otherwise some collections of models may end up siloed by their dialect or it’ll become hard to innovate for fear of losing backward compatibility.</li>
</ol>
</li>
<li>Failure to catch on among the right people.<ol>
<li>Squiggle may fail to get used or it might be used by people with nonaltruistic motives or uncooperative methods.</li>
<li>I’ve changed the name of the library from Swungdash to SquigglyPy to make it easier to set up Google Alerts to track who is using it.</li>
<li>It may also be valuable to cultivate a community around it that has a lot of knowledge and resources so that people who use it have an incentive to join the community, so that the developers can stay in touch with them.</li>
</ol>
</li>
<li>Conflict through coordination failures.<ol>
<li>If a community forms around the creation and composition of Squiggle models, then there will be power in the repository of knowledge of that community. It’ll be important to instill the right ethos in the community early on to avoid the sorts of <a href="https://longtermrisk.org/coordination-challenges-for-preventing-ai-conflict/">coordination failures that can arise even among well-intentioned actors</a>. It is not clear what ethos that is, so a general readiness to adopt it and an interest in cooperative conflict resolution may be a good start.</li>
</ol>
</li>
<li>Discrimination against illegible values.<ol>
<li>There is a risk that Squiggle will differentially enhance the thinking of people with broadly utilitarian values because they’re easier to quantify than other values. Other values may be displaced or abandoned.</li>
<li>I consider it an open question which procedures to idealized preferences are cooperative, but to the extent that the influence of Squiggle will be uncooperative, it’ll be a problem. We may choose to reject such uncooperative behavior terminally or consider that it can backfire against us in an acausal bargain. (See <a href="https://longtermrisk.org/multiverse-wide-cooperation-via-correlated-decision-making/">Oesterheld 2017, especially section 3.3.1</a>.)</li>
<li>Either way, the cooperative ethos of the hypothetical Squiggle community should include respect for other value systems and protection of strong interests of moral minorities.</li>
</ol>
</li>
<li>Missing of failure modes when they happen.<ol>
<li>On a higher level, we can try to prevent or address all these risks, but we may also fail to notice them when they happen, which seems like a risk in itself.</li>
</ol>
</li>
<li>Benefiting malevolent actors<ol>
<li>Finally, there is the risk that we’ll build up a vast body of knowledge that can then be abused by people with malevolent or otherwise uncooperative intentions.</li>
<li>It may pay off to reduce the generality of the knowledge we build by focusing on research questions that are particularly interesting to altruists only.</li>
<li>Making the community inhospitable to such people may not be a good approach, at least not generally, because that may not slow them down but it will deprive us of a chance to talk to them and make them see the future from our perspective.</li>
</ol>
</li>
</ol>
<h3 id="delayed"><a class="toclink" href="#delayed">Delayed</a></h3>
<p>These will either take a while to manifest, will be obvious, will be easy to control, or will be problems no matter what. They are explained in <a href="https://impartial-priorities.org/epistemic-hazards.html">“How Might Better Collective Decision-Making Backfire?”</a> This subsection is mostly notes I wrote to myself a while back.</p>
<ol>
<li>Decision theoretic catch-22s.<ol>
<li>I expect that military intelligence makes a big difference here, but prediction tools may still improve upon this significantly. (I’m mostly thinking of conflicts between countries here since behaviors of other people in interpersonal conflicts can be more easily intuited by many/most people.)</li>
<li>This will be easier to address with better decision theory, and it’s a big problem anyway, so people (e.g., at <span class="caps">MIRI</span>) are probably long working on said decision theory.</li>
<li>The example I cite in the blog post has the feature that the payoff chosen by the dumb dictator is not a typical Schelling point. If this is a general feature of such situations, and if I’m not missing more commonplace examples of this exploit, it’ll likely take a long time before we become sophisticated enough to become vulnerable to this. I feel like I’m probably wrong about this, but I can’t tell in what way.</li>
</ol>
</li>
<li>More power to the group.<ol>
<li>I think this can be observed and reacted to.</li>
</ol>
</li>
<li>Overconfidence may be necessary for motivation.<ol>
<li>There is already a strong focus in the community on individual rationality. It seems like a stretch to think that a collaborative system would influence this potential issue at the margin.</li>
</ol>
</li>
<li>The sunk cost fallacy may be necessary for sustained motivation.<ol>
<li>Same as with overconfidence.</li>
</ol>
</li>
<li>Silencing of lone dissenters.<ol>
<li>This only works if the collaborative group has a lot of social power, so at this point, it’s probably not something that needs to be addressed.</li>
<li>It also seems like something that can be reacted to, i.e. that doesn’t need to be prevented at all costs.</li>
</ol>
</li>
<li>Loss of valuable ambiguity.<ol>
<li>This will probably take a long time to manifest if it does.</li>
</ol>
</li>
<li>Social ramifications.<ol>
<li>This can be evaluated at the time.</li>
</ol>
</li>
<li>Legibility and centralized surveillance.<ol>
<li>This depends heavily on who will feel threatened by what research. So in the end, not using the tools for potential sensitive research is an option.</li>
</ol>
</li>
<li>Cooperation-enhancing effects.<ol>
<li>Probably good on balance,</li>
<li>Unlikely to be affected by the tools,</li>
<li>Will take a long time.</li>
</ol>
</li>
</ol>
<h2 id="demo"><a class="toclink" href="#demo">Demo</a></h2>
<p>Christian Tarsney writes in his paper <a href="https://globalprioritiesinstitute.org/christian-tarsney-the-epistemic-challenge-to-longtermism/">The Epistemic Challenge to Longtermism</a>:</p>
<blockquote>
<p>The ideal Bayesian approach would be to treat all the model parameters as random variables rather than point estimates, choose a probability distribution that represents our uncertainty about each parameter, and compute <span class="caps">EV</span>(L) on that basis. But for our purposes, this approach has significant drawbacks: <span class="caps">EV</span>(L) would be extremely sensitive to the tails of the distributions for parameters like r, s, and vs. And specifying full distributions for these parameters—in particular, specifying the size and shape of the tails—would require a great deal of subjective and questionable guesswork, especially since we have nothing like observed, empirical distributions to rely on. Even if we aim to adopt distributions that are conservative (i.e., unfavorable to longtermism), it would be hard to be confident that the tails of our chosen distributions are genuinely as conservative as we intended.</p>
</blockquote>
<p>That seems fair. My software, SquigglyPy, addresses the concern about subjective guesswork insofar as it’s not a printed and published paper but an interactive software tool that anyone can play around with and run with their own subjective guesses for parameter values.</p>
<p>SquigglyPy does not yet address the related concern about sensitivity to tiny differences in the shapes of the tails of distributions. This concern takes two forms: (1) We may have no reason to prefer one tail shape over another and (2) the Monte Carlo–based sampling methods that SquigglyPy uses to approximate the tails may never use enough samples to approximate them sufficiently no matter how many samples we tell SquigglyPy to use.</p>
<p>I consider the first problem to be outside of the scope of SquigglyPy. You can try to address it by experimenting with many very different tail shapes or, if you’re dealing with a univariate distribution, to plot the distribution for a range of tail shapes. But those are, admittedly, not solutions that are universally applicable or sure to cover all possible shapes.</p>
<p>The second problem is one that it would be very interesting to address in a future version of SquigglyPy. The ReasonML implementation already tries to manipulate various well-known distributions analytically for as long as possible before resorting to sampling. That seems like a very promising 80/20 solution to this problem.</p>
<p>Technical note: Ideally, I’d like to implement a class <code>Distribution</code> that always has a representation that is a mapping from x to y coordinates but also, whenever possible, has a representation that is a formula and one that is a set of samples. All sorts of mathematical operations could then be defined on distribution samples (the status quo) and formulas, all fully encapsulated in the Distribution class. Any code that uses the distributions could be completely oblivious to the inner workings of the distributions, just mathing them like any other variable, while the distributions would take care to keep their formula representations for as long as possible. Plots could always use the x/y coordinate representation, again not caring how the distribution generated it.</p>
<h3 id="starboard"><a class="toclink" href="#starboard">Starboard</a></h3>
<p><a href="https://starboard.gg/Telofy/squigglypy-demo-nQBTns8">The following is best viewed on Starboard.</a></p>
<p>Starboard is like a Jupyter Notebook or like Google Colab except that it runs in your browser! The nice folks of <a href="https://pyodide.org/en/stable/">Pyodide</a> have compiled the CPython interpreter to WebAssembly so you don’t have to connect to a server to run your Python scripts. Starboard then built a Jupyter Notebook–like environment around it.</p>
<p>To use SquigglyPy in Starboard, you can simply import it with Micropip, because it’s pure Python. Of course you don’t <em>have</em> to use Starboard. You can also install it with pip, e.g., in your <a href="https://python-poetry.org/">Poetry</a> environment.</p>
<div class="highlight"><pre><span></span><code><span class="kn">import</span> <span class="nn">micropip</span>
<span class="n">micropip</span><span class="o">.</span><span class="n">install</span><span class="p">(</span><span class="s1">'https://packages.claviger.net/public/squigglypy-0.2.2-py3-none-any.whl'</span><span class="p">)</span>
</code></pre></div>
<p>Next, a bunch of imports. From SquigglyPy in particular we import the context, which stores some more or less global state and settings, various distributions, and miscellaneous utility functions.</p>
<div class="highlight"><pre><span></span><code><span class="kn">import</span> <span class="nn">math</span>
<span class="kn">import</span> <span class="nn">scipy</span>
<span class="kn">import</span> <span class="nn">numpy</span> <span class="k">as</span> <span class="nn">np</span>
<span class="kn">import</span> <span class="nn">matplotlib</span> <span class="k">as</span> <span class="nn">mpl</span>
<span class="kn">import</span> <span class="nn">matplotlib.pyplot</span> <span class="k">as</span> <span class="nn">plt</span>
<span class="kn">import</span> <span class="nn">matplotlib.transforms</span> <span class="k">as</span> <span class="nn">transforms</span>
<span class="kn">import</span> <span class="nn">pandas</span> <span class="k">as</span> <span class="nn">pd</span>
<span class="kn">from</span> <span class="nn">collections.abc</span> <span class="kn">import</span> <span class="n">Iterable</span>
<span class="kn">from</span> <span class="nn">squigglypy.context</span> <span class="kn">import</span> <span class="n">Context</span>
<span class="kn">from</span> <span class="nn">squigglypy.dsl</span> <span class="kn">import</span> <span class="n">normal</span><span class="p">,</span> <span class="n">uniform</span><span class="p">,</span> <span class="n">mixture</span><span class="p">,</span> <span class="n">pareto</span><span class="p">,</span> <span class="n">lognormal</span>
<span class="kn">from</span> <span class="nn">squigglypy.utils</span> <span class="kn">import</span> <span class="n">bfs</span><span class="p">,</span> <span class="n">as_model</span>
<span class="n">plt</span><span class="o">.</span><span class="n">style</span><span class="o">.</span><span class="n">use</span><span class="p">(</span><span class="s1">'seaborn'</span><span class="p">)</span>
</code></pre></div>
<p>The plotting function is a mess and quite specific to our case here, so I didn’t want to include it in the library. Ideally, I’d like to have plotting methods with sensible defaults integrated into the model <span class="caps">API</span> so that you can produce a plot with one command. But that would require (1) knowing what these sensible defaults are, (2) coming up with an <span class="caps">API</span> that allows you to override these defaults at will when they don’t work for you without copy-pasting a big chunk of code, and (3) tweaking the result based on the actual needs of actual users. So I’d be happy to accept pull requests that solve that for me. ^.^</p>
<div class="highlight"><pre><span></span><code><span class="k">def</span> <span class="nf">plot</span><span class="p">(</span><span class="n">model_</span><span class="p">,</span> <span class="n">points</span><span class="p">,</span> <span class="n">sample_count</span><span class="o">=</span><span class="mi">100</span><span class="p">,</span> <span class="n">sample_fmt</span><span class="o">=</span><span class="s1">'b-'</span><span class="p">,</span> <span class="n">quantiles</span><span class="o">=</span><span class="p">(</span><span class="mf">0.1</span><span class="p">,</span> <span class="mf">0.5</span><span class="p">,</span> <span class="mf">0.9</span><span class="p">)):</span>
<span class="k">with</span> <span class="n">Context</span><span class="p">(</span><span class="n">sample_count</span><span class="o">=</span><span class="n">sample_count</span><span class="p">):</span>
<span class="n">samples</span> <span class="o">=</span> <span class="n">np</span><span class="o">.</span><span class="n">stack</span><span class="p">([</span><span class="o">~</span><span class="n">model_</span><span class="p">(</span><span class="n">point</span><span class="p">)</span> <span class="k">for</span> <span class="n">point</span> <span class="ow">in</span> <span class="n">points</span><span class="p">])</span>
<span class="n">p10</span><span class="p">,</span> <span class="n">p50</span><span class="p">,</span> <span class="n">p90</span> <span class="o">=</span> <span class="n">np</span><span class="o">.</span><span class="n">quantile</span><span class="p">(</span><span class="n">samples</span><span class="p">,</span> <span class="n">quantiles</span><span class="p">,</span> <span class="n">axis</span><span class="o">=</span><span class="mi">1</span><span class="p">)</span>
<span class="n">plt</span><span class="o">.</span><span class="n">figure</span><span class="p">(</span><span class="n">figsize</span><span class="o">=</span><span class="p">(</span><span class="mi">8</span><span class="p">,</span> <span class="mi">6</span><span class="p">))</span>
<span class="n">plt</span><span class="o">.</span><span class="n">yscale</span><span class="p">(</span><span class="s1">'log'</span><span class="p">)</span>
<span class="n">plt</span><span class="o">.</span><span class="n">plot</span><span class="p">(</span><span class="n">points</span><span class="p">,</span> <span class="n">p50</span><span class="p">,</span> <span class="s1">'r-'</span><span class="p">,</span> <span class="n">label</span><span class="o">=</span><span class="s1">'mean'</span><span class="p">)</span>
<span class="n">plt</span><span class="o">.</span><span class="n">fill_between</span><span class="p">(</span><span class="n">points</span><span class="p">,</span> <span class="n">p10</span><span class="p">,</span> <span class="n">p90</span><span class="p">,</span> <span class="n">color</span><span class="o">=</span><span class="s2">"red"</span><span class="p">,</span> <span class="n">alpha</span><span class="o">=</span><span class="mf">0.1</span><span class="p">)</span>
<span class="n">plt</span><span class="o">.</span><span class="n">plot</span><span class="p">(</span><span class="n">points</span><span class="p">,</span> <span class="n">samples</span><span class="p">,</span> <span class="n">sample_fmt</span><span class="p">,</span> <span class="n">alpha</span><span class="o">=</span><span class="mf">0.1</span><span class="p">)</span>
<span class="n">plt</span><span class="o">.</span><span class="n">xlim</span><span class="p">(</span><span class="mi">0</span><span class="p">,</span> <span class="kc">None</span><span class="p">)</span>
</code></pre></div>
<p><a href="https://globalprioritiesinstitute.org/christian-tarsney-the-epistemic-challenge-to-longtermism/">Christian Tarsney’s cubic growth model</a> (section 4) describes the scenario that seems most sensible to me. I’ll focus on it for the purposes of this demo. My original plan was to implement it here faithfully, but I don’t have time for that now, so it has more the character of a proof of concept.</p>
<p>I’ll keep the explanations brief here because Tarsney has already done a great job at that.</p>
<div class="highlight"><pre><span></span><code><span class="n">value_star</span> <span class="o">=</span> <span class="n">pareto</span><span class="p">(</span><span class="mf">.4</span><span class="p">)</span> <span class="o">*</span> <span class="mf">3.333e5</span>
<span class="n">value_star</span><span class="o">.</span><span class="n">name</span> <span class="o">=</span> <span class="s1">'Value delta per star (v_s)'</span>
<span class="n">value_earth</span> <span class="o">=</span> <span class="n">lognormal</span><span class="p">(</span><span class="mi">0</span><span class="p">,</span> <span class="mi">1</span><span class="p">)</span> <span class="o">*</span> <span class="mf">2e6</span>
<span class="n">value_earth</span><span class="o">.</span><span class="n">name</span> <span class="o">=</span> <span class="s1">'Value delta on Earth (v_e)'</span>
<span class="n">rate</span> <span class="o">=</span> <span class="n">uniform</span><span class="p">(</span><span class="mi">10</span><span class="p">,</span> <span class="mf">1e8</span><span class="p">)</span><span class="o">**-</span><span class="mi">1</span>
<span class="n">rate</span><span class="o">.</span><span class="n">name</span> <span class="o">=</span> <span class="s1">'Rate of exogenous nullifying events (r)'</span>
<span class="c1"># This doesn’t seem to work for some reason?</span>
<span class="n">speed</span> <span class="o">=</span> <span class="mf">0.1</span> <span class="c1"># lognormal(1, .3) / 10</span>
<span class="c1">#speed.name = 'Speed of settlement in c (s)'</span>
<span class="n">year_start</span> <span class="o">=</span> <span class="mi">0</span> <span class="c1"># = t_l, year of start of settlement</span>
<span class="n">year_end</span> <span class="o">=</span> <span class="mf">1e14</span> <span class="c1"># = t_f, eschatological bound</span>
<span class="n">probability</span> <span class="o">=</span> <span class="mf">2e-14</span> <span class="c1"># = p, probability delta</span>
<span class="n">density_galaxy</span> <span class="o">=</span> <span class="mf">2.2e-5</span> <span class="c1"># = d_g, star density within a radius of 130k ly</span>
<span class="n">density_supercluster</span> <span class="o">=</span> <span class="mf">2.9e-9</span> <span class="c1"># = d_s, star density within Virgo Supercluster</span>
<span class="k">def</span> <span class="nf">number_stars</span><span class="p">(</span><span class="n">radius</span><span class="p">):</span> <span class="c1"># In light years</span>
<span class="k">if</span> <span class="n">radius</span> <span class="o"><=</span> <span class="mi">0</span><span class="p">:</span>
<span class="k">return</span> <span class="mi">0</span>
<span class="k">elif</span> <span class="n">radius</span> <span class="o"><=</span> <span class="mf">1.3e5</span><span class="p">:</span>
<span class="k">return</span> <span class="p">(</span><span class="mi">4</span><span class="o">/</span><span class="mi">3</span><span class="p">)</span> <span class="o">*</span> <span class="n">math</span><span class="o">.</span><span class="n">pi</span> <span class="o">*</span> <span class="n">radius</span><span class="o">**</span><span class="mi">3</span> <span class="o">*</span> <span class="n">density_galaxy</span>
<span class="k">return</span> <span class="p">(</span><span class="mi">4</span><span class="o">/</span><span class="mi">3</span><span class="p">)</span> <span class="o">*</span> <span class="n">math</span><span class="o">.</span><span class="n">pi</span> <span class="o">*</span> <span class="p">(</span>
<span class="n">radius</span><span class="o">**</span><span class="mi">3</span> <span class="o">*</span> <span class="n">density_supercluster</span> <span class="o">+</span>
<span class="mf">1.3e5</span><span class="o">**</span><span class="mi">3</span> <span class="o">*</span> <span class="p">(</span><span class="n">density_galaxy</span> <span class="o">-</span> <span class="n">density_supercluster</span><span class="p">)</span>
<span class="p">)</span>
<span class="k">def</span> <span class="nf">cubic_growth</span><span class="p">(</span><span class="n">year</span><span class="p">):</span>
<span class="k">return</span> <span class="n">value_earth</span> <span class="o">+</span> <span class="p">(</span>
<span class="n">value_star</span> <span class="o">*</span>
<span class="n">number_stars</span><span class="p">((</span><span class="n">year</span> <span class="o">-</span> <span class="n">year_start</span><span class="p">)</span> <span class="o">*</span> <span class="n">speed</span><span class="p">)</span> <span class="o">*</span>
<span class="n">math</span><span class="o">.</span><span class="n">e</span><span class="o">**</span><span class="p">(</span><span class="n">rate</span> <span class="o">*</span> <span class="n">year</span> <span class="o">*</span> <span class="o">-</span><span class="mi">1</span><span class="p">)</span>
<span class="p">)</span>
<span class="n">years</span> <span class="o">=</span> <span class="nb">range</span><span class="p">(</span><span class="mi">0</span><span class="p">,</span> <span class="nb">int</span><span class="p">(</span><span class="mf">1e6</span><span class="p">),</span> <span class="nb">int</span><span class="p">(</span><span class="mf">1e3</span><span class="p">))</span>
</code></pre></div>
<p>This is the value over time. There is an <code>Integral</code> resolver in SquigglyPy, but I haven’t gotten around to using it, so this is just the part inside the integral. (Also ignoring the probability delta, which is outside the integral.)</p>
<p>You can see that all the semitransparent lines overlap and so naturally indicate where the probability density is highest. In this case, this is similar to the 80% <span class="caps">HDI</span> (reddish) and sort of clustered around the mean, but you can just as soon imagine a more multimodal distribution where the <span class="caps">HDI</span> would hide the multimodal nature.</p>
<div class="highlight"><pre><span></span><code><span class="n">plot</span><span class="p">(</span><span class="n">cubic_growth</span><span class="p">,</span> <span class="n">years</span><span class="p">,</span> <span class="n">sample_count</span><span class="o">=</span><span class="mi">100</span><span class="p">)</span>
<span class="n">plt</span><span class="o">.</span><span class="n">show</span><span class="p">()</span>
</code></pre></div>
<p><img alt="Graph of the value distribution over time" src="/images/squigglypy-alpha/value.svg"></p>
<p>Consider, for example, the fabulous <code>foobar</code> model. The reddish <span class="caps">HDI</span> and the mean are rather uninformative compared to the traces.</p>
<div class="highlight"><pre><span></span><code><span class="k">def</span> <span class="nf">foobar</span><span class="p">(</span><span class="n">x</span><span class="p">):</span>
<span class="n">weight</span> <span class="o">=</span> <span class="n">mixture</span><span class="p">([</span><span class="n">normal</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mf">.1</span><span class="p">),</span> <span class="n">normal</span><span class="p">(</span><span class="mi">0</span><span class="p">,</span> <span class="mf">.1</span><span class="p">)])</span>
<span class="n">bias</span> <span class="o">=</span> <span class="n">uniform</span><span class="p">(</span><span class="mi">100</span><span class="p">,</span> <span class="mi">200</span><span class="p">)</span>
<span class="k">return</span> <span class="n">weight</span> <span class="o">*</span> <span class="n">x</span><span class="o">**</span><span class="mi">2</span> <span class="o">+</span> <span class="n">bias</span> <span class="o">/</span> <span class="mi">5</span>
<span class="n">plot</span><span class="p">(</span><span class="n">foobar</span><span class="p">,</span> <span class="nb">range</span><span class="p">(</span><span class="mi">100</span><span class="p">),</span> <span class="n">sample_count</span><span class="o">=</span><span class="mi">100</span><span class="p">)</span>
<span class="n">plt</span><span class="o">.</span><span class="n">yscale</span><span class="p">(</span><span class="s1">'linear'</span><span class="p">)</span>
<span class="n">plt</span><span class="o">.</span><span class="n">show</span><span class="p">()</span>
</code></pre></div>
<p><img alt="Example graph" src="/images/squigglypy-alpha/foobar.svg"></p>
<p>From here it’s also conceptually easy to develop a sensitivity analysis that tells us what influence a given constant (by which I mean a univariate random variable, one that, in our example, is independent of the time) has on the model value for a selection of points in time.</p>
<p>The main difficulties here are:</p>
<ol>
<li>What methods of visualization are there that are more intuitive and/or informative than the one I’ve chosen here?</li>
<li>What graphing library do we use to make this pretty and interactive?</li>
<li>How do we integrate this into SquigglyPy so that the APIs are intuitive but also powerful enough to allow for a few other comparisons?</li>
</ol>
<p>The goal would be to come up with an <span class="caps">API</span> in SquigglyPy that allows for a range of comparisons of parts of the model to other parts regardless of whether they are dependent or independent of <code>x</code>. (<code>x</code> is the time in our example.) The visualizations would have to be different ones for these combinations.</p>
<div class="highlight"><pre><span></span><code><span class="k">def</span> <span class="nf">sensitivity</span><span class="p">(</span><span class="n">model</span><span class="p">,</span> <span class="n">xs</span><span class="o">=</span><span class="kc">None</span><span class="p">,</span> <span class="n">sample_counts</span><span class="o">=</span><span class="p">(</span><span class="mi">10</span><span class="p">,</span> <span class="mi">10</span><span class="p">)):</span>
<span class="n">x_sample_count</span><span class="p">,</span> <span class="n">dist_sample_count</span> <span class="o">=</span> <span class="n">sample_counts</span>
<span class="n">xs</span> <span class="o">=</span> <span class="n">xs</span> <span class="ow">or</span> <span class="nb">range</span><span class="p">(</span><span class="mi">30</span><span class="p">)</span>
<span class="n">constants</span><span class="p">,</span> <span class="n">variables</span><span class="p">,</span> <span class="n">tracer</span> <span class="o">=</span> <span class="n">bfs</span><span class="p">(</span><span class="n">model</span><span class="p">)</span>
<span class="n">constants</span> <span class="o">=</span> <span class="p">[</span><span class="n">constant</span> <span class="k">for</span> <span class="n">constant</span> <span class="ow">in</span> <span class="n">constants</span> <span class="k">if</span> <span class="nb">getattr</span><span class="p">(</span><span class="n">constant</span><span class="p">,</span> <span class="s1">'name'</span><span class="p">,</span> <span class="kc">False</span><span class="p">)]</span>
<span class="n">fig</span><span class="p">,</span> <span class="n">axes</span> <span class="o">=</span> <span class="n">plt</span><span class="o">.</span><span class="n">subplots</span><span class="p">(</span><span class="nb">len</span><span class="p">(</span><span class="n">constants</span><span class="p">),</span> <span class="n">figsize</span><span class="o">=</span><span class="p">(</span><span class="mi">10</span><span class="p">,</span> <span class="nb">len</span><span class="p">(</span><span class="n">constants</span><span class="p">)</span> <span class="o">*</span> <span class="mi">5</span><span class="p">))</span>
<span class="n">plt</span><span class="o">.</span><span class="n">subplots_adjust</span><span class="p">(</span><span class="n">hspace</span><span class="o">=</span><span class="mf">0.5</span><span class="p">)</span>
<span class="k">for</span> <span class="n">ci</span><span class="p">,</span> <span class="n">constant</span> <span class="ow">in</span> <span class="nb">enumerate</span><span class="p">(</span><span class="n">constants</span><span class="p">):</span>
<span class="n">axis</span> <span class="o">=</span> <span class="n">axes</span><span class="p">[</span><span class="n">ci</span><span class="p">]</span>
<span class="k">with</span> <span class="n">Context</span><span class="p">(</span><span class="n">sample_count</span><span class="o">=</span><span class="n">dist_sample_count</span><span class="p">):</span>
<span class="n">constant_samples</span> <span class="o">=</span> <span class="o">~</span><span class="n">constant</span>
<span class="n">variable_samples</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">DataFrame</span><span class="p">([</span><span class="o">~</span><span class="n">model</span><span class="p">(</span><span class="n">x</span><span class="p">)</span> <span class="k">for</span> <span class="n">x</span> <span class="ow">in</span> <span class="n">xs</span><span class="p">],</span> <span class="n">index</span><span class="o">=</span><span class="n">xs</span><span class="p">)</span>
<span class="k">if</span> <span class="ow">not</span> <span class="p">(</span>
<span class="nb">hasattr</span><span class="p">(</span><span class="n">constant_samples</span><span class="p">,</span> <span class="s1">'shape'</span><span class="p">)</span>
<span class="ow">and</span> <span class="n">constant_samples</span><span class="o">.</span><span class="n">shape</span> <span class="o">==</span> <span class="p">(</span><span class="n">dist_sample_count</span><span class="p">,)</span>
<span class="ow">and</span> <span class="n">variable_samples</span><span class="o">.</span><span class="n">shape</span> <span class="o">==</span> <span class="p">(</span><span class="nb">len</span><span class="p">(</span><span class="n">xs</span><span class="p">),</span> <span class="n">dist_sample_count</span><span class="p">)</span>
<span class="p">):</span>
<span class="k">continue</span>
<span class="n">normalize_index</span> <span class="o">=</span> <span class="n">mpl</span><span class="o">.</span><span class="n">colors</span><span class="o">.</span><span class="n">Normalize</span><span class="p">(</span><span class="n">vmin</span><span class="o">=</span><span class="mf">0.0</span><span class="p">,</span> <span class="n">vmax</span><span class="o">=</span><span class="mf">10.0</span><span class="p">)</span>
<span class="n">x_sample_indices</span> <span class="o">=</span> <span class="n">np</span><span class="o">.</span><span class="n">linspace</span><span class="p">(</span><span class="mi">0</span><span class="p">,</span> <span class="nb">len</span><span class="p">(</span><span class="n">xs</span><span class="p">)</span> <span class="o">-</span> <span class="mi">1</span><span class="p">,</span> <span class="n">num</span><span class="o">=</span><span class="mi">10</span><span class="p">,</span> <span class="n">dtype</span><span class="o">=</span><span class="nb">int</span><span class="p">)</span>
<span class="n">axis</span><span class="o">.</span><span class="n">set_xlabel</span><span class="p">(</span><span class="nb">str</span><span class="p">(</span><span class="n">constant</span><span class="p">))</span>
<span class="n">axis</span><span class="o">.</span><span class="n">set_ylabel</span><span class="p">(</span><span class="s1">'model value'</span><span class="p">)</span>
<span class="n">rows</span> <span class="o">=</span> <span class="n">variable_samples</span><span class="o">.</span><span class="n">iloc</span><span class="p">[</span><span class="n">x_sample_indices</span><span class="p">]</span><span class="o">.</span><span class="n">iterrows</span><span class="p">()</span>
<span class="k">for</span> <span class="n">i</span><span class="p">,</span> <span class="p">(</span><span class="n">x</span><span class="p">,</span> <span class="n">variable_samples_at_x</span><span class="p">)</span> <span class="ow">in</span> <span class="nb">enumerate</span><span class="p">(</span><span class="n">rows</span><span class="p">):</span>
<span class="n">combined_samples</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">DataFrame</span><span class="p">({</span><span class="s1">'model'</span><span class="p">:</span> <span class="n">variable_samples_at_x</span><span class="p">,</span> <span class="s1">'node'</span><span class="p">:</span> <span class="n">constant_samples</span><span class="p">})</span>
<span class="n">combined_samples</span> <span class="o">=</span> <span class="n">combined_samples</span><span class="o">.</span><span class="n">sort_values</span><span class="p">(</span><span class="s1">'node'</span><span class="p">)</span>
<span class="n">cmap</span> <span class="o">=</span> <span class="n">plt</span><span class="o">.</span><span class="n">get_cmap</span><span class="p">(</span><span class="s1">'coolwarm'</span><span class="p">,</span> <span class="mi">256</span><span class="p">)</span>
<span class="n">color</span> <span class="o">=</span> <span class="n">mpl</span><span class="o">.</span><span class="n">colors</span><span class="o">.</span><span class="n">to_rgb</span><span class="p">(</span><span class="n">cmap</span><span class="p">(</span><span class="n">normalize_index</span><span class="p">(</span><span class="n">i</span><span class="p">)))</span>
<span class="n">axis</span><span class="o">.</span><span class="n">plot</span><span class="p">(</span>
<span class="n">combined_samples</span><span class="o">.</span><span class="n">node</span><span class="p">,</span>
<span class="n">combined_samples</span><span class="o">.</span><span class="n">model</span><span class="p">,</span>
<span class="n">c</span><span class="o">=</span><span class="n">color</span><span class="p">,</span>
<span class="n">linewidth</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span>
<span class="p">)</span>
<span class="n">axis</span><span class="o">.</span><span class="n">legend</span><span class="p">([</span><span class="sa">f</span><span class="s1">'x = </span><span class="si">{</span><span class="n">x</span><span class="si">}</span><span class="s1">'</span> <span class="k">for</span> <span class="n">x</span> <span class="ow">in</span> <span class="n">np</span><span class="o">.</span><span class="n">array</span><span class="p">(</span><span class="n">xs</span><span class="p">)[</span><span class="n">x_sample_indices</span><span class="p">]],</span> <span class="n">loc</span><span class="o">=</span><span class="s1">'upper left'</span><span class="p">)</span>
<span class="n">sensitivity</span><span class="p">(</span><span class="n">cubic_growth</span><span class="p">,</span> <span class="n">xs</span><span class="o">=</span><span class="n">years</span><span class="p">)</span>
<span class="n">plt</span><span class="o">.</span><span class="n">show</span><span class="p">()</span>
</code></pre></div>
<p><img alt="Sensitivity plot" src="/images/squigglypy-alpha/sensitivity.svg"></p>Researchers Answering Questions 20202021-01-17T00:00:00+00:002021-01-17T00:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2021-01-17:/researchers-answering-questions-2020.html<p>I used a recent Ask-Me-Anything (<span class="caps">AMA</span>) of Rethink Priorities to ask a few researchers a series of questions about research in general. They are questions whose answers are of general interest to anyone doing research and maybe even to even wider groups of people, so researchers outside of <span class="caps">RP</span> have answered them too, and they may be interesting to you too!</p><p>You can find <a href="https://forum.effectivealtruism.org/posts/WJ3rBBkawoTJcY642/ask-rethink-priorities-anything-ama?commentId=ZYd8p8sJwRmuSKmaS">the whole question thread in the <span class="caps">EA</span> Forum</a>. This is a lightly edited transposition of the thread, that is, this article puts each question first and then lists all answers. This way, the table of contents allows you to jump to whichever question interests you most.</p>
<div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#thinking-vs-reading">Thinking vs. reading</a><ul>
<li><a href="#jason-schukraft">Jason Schukraft</a></li>
<li><a href="#david-bernard">David Bernard</a></li>
<li><a href="#michael-aird">Michael Aird</a></li>
<li><a href="#alex-lintz">Alex Lintz</a></li>
</ul>
</li>
<li><a href="#self-consciousness">Self-Consciousness</a><ul>
<li><a href="#jason-schukraft_1">Jason Schukraft</a></li>
<li><a href="#holly-elmore">Holly Elmore</a></li>
<li><a href="#alex-lintz_1">Alex Lintz</a></li>
<li><a href="#michael-aird_1">Michael Aird</a></li>
</ul>
</li>
<li><a href="#is-there-something-interesting-here">Is there something interesting here?</a><ul>
<li><a href="#jason-schukraft_2">Jason Schukraft</a></li>
<li><a href="#david-bernard_1">David Bernard</a></li>
<li><a href="#michael-aird_2">Michael Aird</a></li>
</ul>
</li>
<li><a href="#survival-vs-exploratory-mindset">Survival vs. exploratory mindset</a><ul>
<li><a href="#jason-schukraft_3">Jason Schukraft</a></li>
<li><a href="#david-bernard_2">David Bernard</a></li>
</ul>
</li>
<li><a href="#optimal-hours-of-work-per-day">Optimal hours of work per day</a><ul>
<li><a href="#alex-lintz_2">Alex Lintz</a></li>
<li><a href="#david-bernard_3">David Bernard</a></li>
<li><a href="#jason-schukraft_4">Jason Schukraft</a></li>
</ul>
</li>
<li><a href="#learning-a-new-field">Learning a new field</a><ul>
<li><a href="#holly-elmore_1">Holly Elmore</a></li>
<li><a href="#alex-lintz_3">Alex Lintz</a></li>
<li><a href="#michael-aird_3">Michael Aird</a></li>
<li><a href="#david-bernard_4">David Bernard</a></li>
<li><a href="#jason-schukraft_5">Jason Schukraft</a></li>
</ul>
</li>
<li><a href="#hard-problems">Hard problems</a><ul>
<li><a href="#jason-schukraft_6">Jason Schukraft</a></li>
<li><a href="#alex-lintz_4">Alex Lintz</a></li>
<li><a href="#michael-aird_4">Michael Aird</a></li>
</ul>
</li>
<li><a href="#emotional-motivators">Emotional motivators</a><ul>
<li><a href="#jason-schukraft_7">Jason Schukraft</a></li>
<li><a href="#david-bernard_5">David Bernard</a></li>
<li><a href="#alex-lintz_5">Alex Lintz</a></li>
<li><a href="#michael-aird_5">Michael Aird</a></li>
</ul>
</li>
<li><a href="#typing-speed">Typing speed</a><ul>
<li><a href="#jason-schukraft_8">Jason Schukraft</a></li>
<li><a href="#david-bernard_6">David Bernard</a></li>
<li><a href="#linchuan-zhang">Linchuan Zhang</a></li>
<li><a href="#alex-lintz_6">Alex Lintz</a></li>
<li><a href="#michael-aird_6">Michael Aird</a></li>
</ul>
</li>
<li><a href="#obvious-questions">Obvious questions</a><ul>
<li><a href="#jason-schukraft_9">Jason Schukraft</a></li>
<li><a href="#linchuan-zhang_1">Linchuan Zhang</a></li>
<li><a href="#michael-aird_7">Michael Aird</a></li>
</ul>
</li>
<li><a href="#tiredness-focus-etc">Tiredness, focus, etc.</a><ul>
<li><a href="#jason-schukraft_10">Jason Schukraft</a></li>
<li><a href="#david-bernard_7">David Bernard</a></li>
<li><a href="#linchuan-zhang_2">Linchuan Zhang</a></li>
<li><a href="#alex-lintz_7">Alex Lintz</a></li>
<li><a href="#michael-aird_8">Michael Aird</a></li>
</ul>
</li>
</ul>
</div>
<h2 id="thinking-vs-reading"><a class="toclink" href="#thinking-vs-reading">Thinking vs. reading</a></h2>
<p>If you want to research a particular topic, how do you balance reading the relevant literature against thinking yourself and recording your thoughts? I’ve heard second-hand that Hilary Greaves recommends thinking first so to be unanchored by the existing literature and the existing approaches to the problem. Another benefit may be that you start out reading the literature with a clearer mental model of the problem, which might make it easier to stay motivated and to remain critical/vigilant while reading. (<a href="https://impartial-priorities.org/harnessing-cognitive-dissonance.html">See this theory of mine.</a>) Would you agree or do you have a different approach?</p>
<h3 id="jason-schukraft"><a class="toclink" href="#jason-schukraft">Jason Schukraft</a></h3>
<blockquote>
<p>I think it depends on the context. Sometimes it makes sense to lean toward thinking more and sometimes it makes sense to lean toward reading more. (I wouldn’t advise focusing exclusively on one or the other.) Unjustified anchoring is certainly a worry, but I think reinventing the wheel is also a worry. One could waste two weeks groping toward a solution to a problem that could have been solved in afternoon just by reading the right review article.</p>
</blockquote>
<h3 id="david-bernard"><a class="toclink" href="#david-bernard">David Bernard</a></h3>
<blockquote>
<p>Another benefit of thinking before reading is that it can help you develop your research skills. Noticing some phenomena and then developing a model to explain it is a super valuable exercise. If it turns out you reproduce something that someone else has already done and published, then great, you’ve gotten experience solving some problem and you’ve shown that you can think through it at least as well as some expert in the field. If it turns out that you have produced something novel then it’s time to see how it compares to existing results in the literature and get feedback on how useful it is.</p>
<p>This said, I think this is more true for theoretical work than applied work, e.g. the value of doing this in philosophy > in theoretical economics > in applied economics. A fair amount of <span class="caps">EA</span>-relevant research is summarising and synthesising what the academic literature on some topic finds and it seems pretty difficult to do that by just thinking to yourself!</p>
</blockquote>
<h3 id="michael-aird"><a class="toclink" href="#michael-aird">Michael Aird</a></h3>
<blockquote>
<p>I don’t think I really have explicit policies regarding balancing reading against thinking myself and recording my thoughts. Maybe I should.</p>
<p>I’m somewhat inclined to think that, on the margin and on average (so not in every case), <span class="caps">EA</span> would benefit from a bit more reading of relevant literatures (or talking to more experienced people in an area, watching of relevant lectures, etc.), even at the expense of having a bit less time for coming up with novel ideas.</p>
<p>I feel like <span class="caps">EA</span> might have a bit too much a tendency towards “think really hard by oneself for a while, then kind-of reinvent the wheel but using new terms for it.” It might be that, often, people could get to similar ideas faster and in a way that connects to existing work better (making it easier for others to find, build on, etc.) by doing some extra reading first.</p>
<p>Note that this is not me suggesting EAs should increase how much they defer to experts/others/existing work. Instead, I’m tentatively suggesting spending more time learning what experts/others/existing work has to say, which could be followed by agreeing, disagreeing, critiquing, building on, proposing alternatives, striking out in a totally different direction, etc.</p>
<p>(On this general topic, I liked the post <a href="https://www.lesswrong.com/posts/64FdKLwmea8MCLWkE/the-neglected-virtue-of-scholarship">The Neglected Virtue of Scholarship</a>.)</p>
<p>Less important personal ramble:</p>
<p>I often feel like I might be spending more time reading up-front than is worthwhile, as a way of procrastinating, or maybe out of a sort-of perfectionism (the more I read, the lower the chance that, once I start writing, what I write is mistaken or redundant). And I sort-of scold myself for that.</p>
<p>But then I’ve repeatedly heard people remark that I have an unusually large amount of output. (I sort-of felt like the opposite was true, until people told me this, which is weird since it’s such an easily checkable thing!) And I’ve also got some feedback that suggested I should move more in the direction of depth and expertise, even at the cost of breadth and quantity of output.</p>
<p>So maybe that feeling that I’m spending too much time reading up-front is just mistaken. And as mentioned, that feeling seems to conflict with what I’d (tentatively) tend to advise others, which should probably make me more suspicious of the feeling. (This reminds me of asking “Is this how I’d treat a friend?” in response to negative self-talk [<a href="https://self-compassion.org/exercise-1-treat-friend/">source with related ideas</a>].)</p>
</blockquote>
<h3 id="alex-lintz"><a class="toclink" href="#alex-lintz">Alex Lintz</a></h3>
<blockquote>
<p>I’ve been playing around with spending 15–60 min. sketching out a quick model of what I think of something before starting in on the literature (by no means a consistent thing I do though). I find it can be quite nice and help me ask the right questions early on.</p>
</blockquote>
<h2 id="self-consciousness"><a class="toclink" href="#self-consciousness">Self-Consciousness</a></h2>
<p>I imagine that virtually any research project, successful and unsuccessful, starts with some inchoate thoughts and notes. These will usually seem hopelessly inadequate but they’ll sometimes mature into something amazingly insightful. Have you ever struggled with mental blocks when you felt self-conscious about these beginnings, and have you found ways to (reliably) overcome them?</p>
<h3 id="jason-schukraft_1"><a class="toclink" href="#jason-schukraft_1">Jason Schukraft</a></h3>
<blockquote>
<p>Yep, I am intimately familiar with hopelessly inchoate thoughts and notes. (I’m not sure I’ve ever completed a project without passing through that stage.) For me at least, the best way to overcome this state is to talk to lots of people. One piece of advice I have for young researchers is to come to terms with sharing your work with people you respect before it’s polished. I’m very grateful to have a large network of collaborators willing to listen to and read my confused ramblings. Feedback at an early stage of a project is often much more valuable than feedback at a later stage.</p>
</blockquote>
<h3 id="holly-elmore"><a class="toclink" href="#holly-elmore">Holly Elmore</a></h3>
<blockquote>
<p>Personally, I’m very self-conscious about my work and tend to wait to long to share it. But the culture of <span class="caps">RP</span> seems to fight that tendency – which I think is very productive!</p>
</blockquote>
<h3 id="alex-lintz_1"><a class="toclink" href="#alex-lintz_1">Alex Lintz</a></h3>
<blockquote>
<p>Idk if this fits exactly but when I started my research position I tried to have the mindset of, “I’ll be pretty bad at this for quite a while.” Then when I made mistakes I could just think, “right, as expected. Now let’s figure out how to not do that again.” Not sure how sustainable this is but it felt good to start! In general it seems good to have a mindset of research being nearly impossibly hard. Humans are just barely able to do this thing in a useful way and even at the highest levels academics still make mistakes (most papers have at least some flaws).</p>
</blockquote>
<h3 id="michael-aird_1"><a class="toclink" href="#michael-aird_1">Michael Aird</a></h3>
<p>This is his answer to the questions about self-consciousness and “Is there something interesting here?”</p>
<blockquote>
<p>These questions definitely resonate with me, and I imagine they’d resonate with most/all researchers.</p>
<p>I have a tendency to continually wonder if what I’m doing is what I should be doing, or if I should change my priorities. I think this is good in some ways. But sometimes I’d make better decisions faster if I just actually pursued an idea more “confidently” for a bit, to get more info on whether it’s worth pursuing, rather than just “wondering” about it repeatedly and going back and forth without much new info to work with. Basically, I might do too much self-doubt-style armchair reasoning, with too little actual empirical info.</p>
<p>Also, pursuing an idea more “confidently” for a bit will not only inform me about whether to continue pursuing it further, but also might result in outputs that are useful for others. So I try to sometimes switch into “just commit and focus mode” for a given time period, or until I hit a given milestone, and mostly minimise reflection on what I should prioritise during that time. But so far this has been like a grab bag of heuristics and habits I use, rather than a more precise guideline for myself.</p>
<p>See also <a href="https://forum.effectivealtruism.org/posts/e2heLEnbeqTHaJqBf/when-to-focus-and-when-to-re-evaluate">When to focus and when to re-evaluate</a>.</p>
<p>Things that help me with this include, and/or some scattered related thoughts, include:</p>
<ul>
<li>Talking to others and getting feedback, including on early-stage ideas</li>
<li>I liked David and Jason’s remarks on this in their comments</li>
<li>A sort-of minimum viable product and quick feedback loop approach has often seemed useful for me – something like:<ul>
<li>First getting verbal feedback from a couple people on a messy, verbal description of an idea</li>
<li>Then writing up a rough draft about the idea and circulating it to a couple more people for a bit more feedback</li>
<li>Then polishing and fleshing out that draft and circulating it to a few more people for more feedback</li>
<li>Then posting publicly</li>
<li>(But only proceeding to the next step if evidence from the prior one – plus one’s own intuitions – suggested this would be worthwhile)</li>
</ul>
</li>
<li>Feedback has often helped me determine whether an idea is worth pursuing further, feel more comfortable/motivated with pursuing an idea further (rather than being mired in unproductive self-doubt), develop the idea, work out which angles of it are most worth pursuing, and work out how to express it more clearly</li>
<li>Reminding myself that I haven’t really gathered any new info since the last time I thought “Should this really be what I spend my time on?,” so thinking about that again is unlikely to reveal new insights, and is probably just a stupid part of my psychology rather than something I’d endorse.</li>
<li>I might think to myself something like “If a friend was doing this, you’d think it’s irrational, and gently advise them to just actually commit for a bit and get new info, right? So shouldn’t you do the same yourself?”</li>
<li>Remembering <em>Algorithms to Live By</em> drawing an analogy to a failure mode in which computers continually reprioritise tasks and the reprioritisation takes up just enough processing power to mean no actual progress on any of the tasks occurs, and this can just cycle forever. And the way to get out of this is to at some point just do tasks, even without having confidence that these “should” be top priority.</li>
<li>This is just my half-remembered version of that part of the book, and might be wrong somehow.</li>
<li>Remembering that I’d be deeply uncertain about the “actual” value of any project I could pursue, because the world is very complicated and my ambitions (contribute to improving the long-term future) are pretty lofty. The best I can do is something that seems good in expected value but with large error bars. So the fact I feel some uncertainty and doubt provides basically no evidence that this project isn’t worth pursuing. (Though feeling an unusually large amount of uncertainty and doubt might.)</li>
<li>Remembering that, if the idea ends up seeming to have not been important <em>but there was a reasonable ex ante case that it <strong>might’ve</strong> been important</em>, there’s a decent chance someone else would end up pursuing it if I don’t. So if I pursue it, then find out it seems to not be important, then write about what I found, that <em>might</em> still have the effect of causing an important project to get done, because it might cause someone else to do that important project rather than doing something similar to what I did.</li>
</ul>
<p><em>Examples to somewhat illustrate the last two points:</em></p>
<p>This year, in some so-far-unpublished work, I wrote about some ideas that:</p>
<ul>
<li>I initially wasn’t confident about the importance of</li>
<li>Seemed like they should’ve been obvious to relevant groups, but seemed not to have been discussed by them. And that generally seems like (at least) weak evidence that an idea either (a) actually <em>isn’t</em> important or (b) has been in essence discussed in some other form or place that I just am not familiar with.</li>
</ul>
<p>So when I had the initial forms of these ideas and wasn’t sure how much time (if any) to spend on them, I took roughly the following approach:</p>
<p>I developed some thoughts on some of the ideas. Then I shared those thoughts verbally or as very rough drafts with a small set of people who seemed like they’d have decent intuitions on whether the ideas were important vs unimportant, somewhat novel vs already covered, etc.</p>
<p>In most cases, this early feedback indicated that it was at least plausible that the ideas were somewhat important and somewhat novel. This – combined with my independent impression that these ideas might be somewhat important and novel – seemed to provide sufficient reason to flesh those ideas out further, as well as to flesh out related ideas (which seemed like they’d probably also be important and novel if the other ideas were, and vice versa).</p>
<p>So I did so, then shared that slightly more widely. Then I got more positive feedback, so I bothered to invest the time to polish the writings up a bit more.</p>
<p>Meanwhile, when I fleshed one of the ideas out a little, it seemed like that one turned out to probably <em>not</em> be very important at all. So with that one, I just made sure that my write-up made it clear early on that my current view was that this idea probably didn’t matter, and I neatened up the write-up just a bit, because I still thought the write-up might be a bit useful either to:</p>
<ul>
<li>Explain to others why they shouldn’t bother exploring the same thing</li>
<li>Make it easy for others to see if they disagreed with my reasoning for why this probably didn’t matter, because I might be wrong about that, and it might be good for others to quickly check that reasoning</li>
</ul>
<p>Having spent time on that idea sort-of felt in hindsight silly or like a mistake. But I think I probably shouldn’t see that as having been a bad decision ex ante, given that:</p>
<ul>
<li>It seems plausible that, if not for my write-up, someone else would’ve eventually “wasted” time on a similar idea</li>
<li>This was just one out of a set of ideas that I tried to flesh out and write up, many/most of which still (in hindsight) seem like they were worth spending time on</li>
<li>So maybe it’s very roughly like I gave 60% predictions for each of 10 things, and decided that that’d mean the expected value of betting on those 10 things was good, and then 6 of those things happened, suggesting I was well-calibrated and was right to bet on those things<ul>
<li>(I didn’t actually make quantitative predictions)</li>
</ul>
</li>
</ul>
<p>And some of the other ideas were in between – no strong reason to believe they were important or that they weren’t – so I just fleshed them out a bit and left it there, pending further feedback. (I also had other things to work on.)</p>
</blockquote>
<p>In a reply, I referred to <a href="https://impartial-priorities.org/the-bulk-of-the-impact-iceberg.html">this related blog post of mine</a>. Michael replied:</p>
<blockquote>
<blockquote>
<p>It’s also important to be transparent about one’s rigor and to make the negative results findable for others. The second is obvious. The first is because the dead end may not actually be a dead end but only looked that way given the particular way in which you had resolved the optimal stopping problem of investigating it (even) further.</p>
</blockquote>
<p>I agree with these points, and think that they might sometimes be under-appreciated (both in and outside of <span class="caps">EA</span>).</p>
<p>To sort-of restate your points:</p>
<ul>
<li>I think it’s common for people to not publish explorations that turned out to seem to “not reveal anything important” (except of course that this direction of exploration might be worth skipping).</li>
<li>Much has been written about this sort of issue, and there can be valid reasons for that behaviour, but sometimes it seems unfortunate.</li>
<li>I think another failure mode is to provide some sort of public info of your belief that this direction of exploration seems worth skipping, but without sufficient <a href="https://www.openphilanthropy.org/reasoning-transparency">reasoning transparency</a>, which could make people rule this out too much/too early.</li>
<li>Again, there can be valid reasons for this (if you’re sufficiently confident that it’s worth ruling out this direction and you have sufficiently high-value other things to do, it might not be worth spending time on a write-up with high reasoning transparency), but sometimes it seems unfortunate.</li>
</ul>
</blockquote>
<p>In the same context, I also brought up a bit of <span class="caps">CFAR</span> lore:</p>
<p>Part of this reminds me a lot of <a href="https://www.rationality.org/resources/updates/2015/qa-isnt-self-deception-sometimes-productive"><span class="caps">CFAR</span>’s approach</a> here (I can’t quite tell whether Julia Galef is interviewer, interviewee, or both):</p>
<blockquote>
<p>For example, when I’ve decided to take a calculated risk, knowing that I might well fail but that it’s still worth it to try, I often find myself worrying about failure even after having made the decision to try. And I might be tempted to lie to myself and say, “Don’t worry! This is going to work!” so that I can be relaxed and motivated enough to push forward.</p>
<p>But instead, in those situations I like to use a framework <span class="caps">CFAR</span> sometimes calls “Worker-me versus <span class="caps">CEO</span>-me.” I remind myself that <span class="caps">CEO</span>-me has thought carefully about this decision, and for now I’m in worker mode, with the goal of executing <span class="caps">CEO</span>-me’s decision. Now is not the time to second-guess the <span class="caps">CEO</span> or worry about failure.</p>
</blockquote>
<p>Your approach to gathering feedback and iterating on the output, making it more and more refined with every iteration but also deciding whether it’s worth another iteration, that process sounds great!</p>
<p>I think a lot of people aim for such a process or want to after reading your comment, but will held back from showing their first draft to their first round of reviewers because they worry the reviewers will think badly of them for addressing a topic of this particular level of perceived difficulty or relevance (maybe it’s too difficult or too irrelevant in the reviewer’s opinion), or think badly of them for a particular wording, or think badly of them because they think you should’ve anticipated a negative effect of writing about the topic and not done so (e.g., some complex acausal trade or social dynamics thing that didn’t occur to you), or just generally have diffuse fears holding them back. Such worries are probably disproportionate, but still, overcoming them will probably require particular tricks or training.</p>
<p>Michael replied:</p>
<blockquote>
<p>I like that “Worker-me versus <span class="caps">CEO</span>-me” framing, and hadn’t heard of it or seen that page, so thanks for sharing that. It does seem related to what I said in the parent comment.</p>
<p>I share the view that it’ll be decently common for a range of disproportionate worries to hold people back from striking out into areas that seem good in expected value but very uncertain and with real counterarguments, and from sharing early-stage results from such pursuits. I also think there can be a range of <em>good</em> reasons to hold back from those things, and that it can be hard to tell when the worries are disproportionate!</p>
<p>I imagine it’d be hard (though not impossible) to generate advice on this that’s quite generally useful without being vague/littered with caveats. People will probably have to experiment to some extent, get advice from trusted people on their general approach, and continue reflecting, or something like that.</p>
</blockquote>
<h2 id="is-there-something-interesting-here"><a class="toclink" href="#is-there-something-interesting-here">Is there something interesting here?</a></h2>
<p>I often have some (for me) novel ideas, but then it turns out that whether true or false, the idea doesn’t seem to have any important implications. Conversely, I’ve dismissed ideas as unimportant, and years later someone developed them – through a lot of work I didn’t do because I thought it wasn’t important – into something that did connect to important topics in unanticipated ways. Do you have rules of thumb that help you assess early on whether a particular idea is worth pursuing?</p>
<h3 id="jason-schukraft_2"><a class="toclink" href="#jason-schukraft_2">Jason Schukraft</a></h3>
<blockquote>
<p>Yep, this also happens to me. Unfortunately, I don’t have any particular insight. Oftentimes the only way to know whether an idea is interesting is to put in the hard exploratory work. Of course, one shouldn’t be afraid to abandon an idea if it looks increasingly unpromising.</p>
</blockquote>
<h3 id="david-bernard_1"><a class="toclink" href="#david-bernard_1">David Bernard</a></h3>
<blockquote>
<p>I mostly try to work out how excited I am by this idea and whether I could see myself still being excited in 6 months, since for me having internal motivation to work on a project is pretty important. I also try to chat about this idea with various other people and see how excited they are by it.</p>
</blockquote>
<h3 id="michael-aird_2"><a class="toclink" href="#michael-aird_2">Michael Aird</a></h3>
<p>See section <a href="#self-consciousness">Self-Consciousness</a>.</p>
<h2 id="survival-vs-exploratory-mindset"><a class="toclink" href="#survival-vs-exploratory-mindset">Survival vs. exploratory mindset</a></h2>
<p>I’ve heard of the distinction between survival mindset and exploratory mindset, which makes intuitive sense to me. (I don’t remember where I learned of these terms, but I tried to clarify how I use them in a comment below.) I imagine that for most novel research, exploratory mindset is the more useful one. (Or would you disagree?) If it doesn’t come naturally to you, how do you cultivate it?</p>
<p>By survival mindset I mean: extreme risk aversion, fear, distrust toward strangers, little collaboration, isolation, guarded interaction with others, hoarding of money and other things, seeking close bonds with family and partners, etc., but I suppose it also comes with modesty and contentment, equanimity in the face of external catastrophes, vigilance, preparedness, etc.</p>
<p>By exploratory mindset I mean: risk neutrality, curiosity, trust toward strangers, collaboration, outgoing social behavior, making oneself vulnerable, trusting partners and family without much need for ritual, quick reinvestment of profits, etc., but I suppose also a bit lower conscientiousness, lacking preparedness for catastrophes, gullibility, overestimating how much others trust you, etc.</p>
<p>Those categories have been very useful for me, but maybe they’re a lot less useful for most other people? You can just ignore that question if the distinction makes no intuitive sense this way or doesn’t quite fit your world models.</p>
<h3 id="jason-schukraft_3"><a class="toclink" href="#jason-schukraft_3">Jason Schukraft</a></h3>
<blockquote>
<p>Insofar as I understand the terms, an exploratory mindset is an absolute must. Not sure how to cultivate it, though.</p>
</blockquote>
<h3 id="david-bernard_2"><a class="toclink" href="#david-bernard_2">David Bernard</a></h3>
<blockquote>
<p>I also haven’t heard these terms before, but from your description (which frames a survival mindset pretty negatively), an exploratory mindset comes fairly naturally to me and therefore I haven’t ever actively cultivated it. Lots of research projects fail so extreme risk aversion in particular seems like it would be bad for researchers.</p>
</blockquote>
<h2 id="optimal-hours-of-work-per-day"><a class="toclink" href="#optimal-hours-of-work-per-day">Optimal hours of work per day</a></h2>
<p>Have you found that a particular number of hours of concentrated work per day works best for you? By this I mean time you spend focused on your research project, excluding time spent answering emails, AMAs, and such. (If hours per day doesn’t seem like an informative unit to you, imagine I asked “hours per week” or whatever seems best to you.)</p>
<h3 id="alex-lintz_2"><a class="toclink" href="#alex-lintz_2">Alex Lintz</a></h3>
<blockquote>
<p>I tend to work about 4–7 hours per day including meetings and everything. Including only mentally intensive tasks I probably get around 4–5 a day. Sometimes I’m able to get more if I fall into a good rhythm with something. Looking around at estimates (Rescuetime says just ~ 3 hours per day average of productive work) it seems clear I’m hitting a pretty solid average. I still can’t shake the feeling that everyone else is doing more work. Part of this is because people claim they do much more work. I assume this is mostly exaggeration though because hours worked is used as a signal of status and being a hard worker. But still, it’s hard to shake the feeling.</p>
</blockquote>
<h3 id="david-bernard_3"><a class="toclink" href="#david-bernard_3">David Bernard</a></h3>
<blockquote>
<p>I typically aim for 6–7 hours of deep work a day and a couple of dedicated hours for miscellaneous tasks and meetings. Since starting part-time at <span class="caps">RP</span> I’ve been doing 6 days a week (2 <span class="caps">RP</span>, 4 PhD), but before that I did 5. I find <span class="caps">RP</span> deep work less taxing than PhD work. 6 days a week is at the upper limit of manageable for me at the moment, so I plan to experiment with different schedules in the new year.</p>
</blockquote>
<h3 id="jason-schukraft_4"><a class="toclink" href="#jason-schukraft_4">Jason Schukraft</a></h3>
<blockquote>
<p>I work between 4 and 8 hours a day. I don’t find any difference in my productivity within that range, though I imagine if I pushed myself to work more than 8, I would pretty quickly hit diminishing returns.</p>
</blockquote>
<h2 id="learning-a-new-field"><a class="toclink" href="#learning-a-new-field">Learning a new field</a></h2>
<p>I don’t know what I mean by “field,” but probably something smaller than “biology” and bigger than “how to use Pipedrive.” If you need to get up to speed on such a field for research that you’re doing, how do you approach it? Do you read textbooks (if so, linearly or more creatively?) or pay grad students to answer your questions? Does your approach vary depending on whether it’s a subfield of your field of expertise or something completely new?</p>
<h3 id="holly-elmore_1"><a class="toclink" href="#holly-elmore_1">Holly Elmore</a></h3>
<blockquote>
<p>I can answer [this], as I’ve been doing it for Wild Animal Welfare since I was hired in September. <span class="caps">WAW</span> is a new and small field, so it is relatively easy to learn the field, but there’s still so much! I started by going backwards (into the Welfare Biology movement of the 80s and 90s) and forwards (into the <span class="caps">WAW</span> <span class="caps">EA</span> orgs we know today) from Brain Tomasik, consulting the primary literature over various specific matters of fact. A great thing about <span class="caps">WAW</span> being such a young field (and so concentrated in <span class="caps">EA</span>) is that I can reach out to basically anyone who’s published on it and have a real conversation. It’s a big shortcut!</p>
<p>I should note that my background is in Evolutionary Biology and Ecology, so someone else might need a lot more background in those basics if they were to learn <span class="caps">WAW</span>.</p>
</blockquote>
<h3 id="alex-lintz_3"><a class="toclink" href="#alex-lintz_3">Alex Lintz</a></h3>
<blockquote>
<p>I just do a lot of literature review. I tend to search for the big papers and meta-analyses, skim lot’s of them and try to make a map of what the key questions are and what the answers proposed by different authors are for each question (noting citations for each answer). This helps to distill the field I think and serves as something relatively easy to reference. Generally there’s a lot of restructuring that needs to happen as you learn more about a topic area and see that some questions you used were ill-posed or some papers answer somewhat different questions. In short this gets messy, but it seems like a good way to start and sometimes it works quite well for me.</p>
</blockquote>
<h3 id="michael-aird_3"><a class="toclink" href="#michael-aird_3">Michael Aird</a></h3>
<blockquote>
<p>I don’t know if I have a great, well-chosen, or transferable method here, so I think people should pay more attention to my colleagues’ answers than mine. But <span class="caps">FWIW</span>, I tend to do a mixture of:</p>
<ul>
<li>reading Wikipedia articles</li>
<li>reading journal article abstracts</li>
<li>reading a small set of journal articles more thoroughly</li>
<li>listening to podcasts</li>
<li>listening to audiobooks</li>
<li>watching videos (e.g., a Yale lecture series on game theory)</li>
<li>talking to people who are already at least sort-of in my network (usually more to get a sounding board or “generalist feedback,” rather than to leverage specific expertise of theirs)</li>
</ul>
<p>I’ve also occasionally used free online courses, e.g. the Udacity Intro to <span class="caps">AI</span> course. (See also <a href="https://forum.effectivealtruism.org/posts/u2DM4Xfbj2CwhWcC5/what-are-some-good-online-courses-relevant-to-ea">What are some good online courses relevant to <span class="caps">EA</span>?</a>)</p>
<p>Whether I take many notes depends on whether I’m just learning about a field because I think it might be useful in some way in future for me to know about that field, or because I have at least a vague idea of a project I might work on within that field (e.g., “how bad would various possible types of nuclear wars be, from a longtermist perspective?”). In the latter case, I’ll take a lot of notes as I go in Roam, beginning to structure things into relevant sub-questions, things to learn more about, etc.</p>
<p>Since leaving university, I haven’t really made much use of textbooks, flashcards, or reaching out to experts who aren’t already in my network. It’s not really that I actively chose to not make much use of these things (it’s just that I never actively chose <em>to</em> make much use of these things), and think it’s plausible that I should make more use of these things. I’ll very likely talk to a bunch of experts for my current or upcoming research projects.</p>
</blockquote>
<h3 id="david-bernard_4"><a class="toclink" href="#david-bernard_4">David Bernard</a></h3>
<blockquote>
<p>I’m a big fan of textbooks and schedule time to read a couple of textbook chapters each week. LessWrong’s <a href="https://www.lesswrong.com/posts/xg3hXCYQPJkwHyik2/the-best-textbooks-on-every-subject#comments">best textbooks on every subject thread is pretty good for finding them</a>. I usually make Anki flashcards to help me remember the key facts, but I’ve recently started experimenting with Roam Research to take notes which I’m also enjoying so my “learning flow” is in flux at the moment.</p>
</blockquote>
<h3 id="jason-schukraft_5"><a class="toclink" href="#jason-schukraft_5">Jason Schukraft</a></h3>
<blockquote>
<p>I can’t emphasize enough the value of just talking to existing experts. For me at least, it’s by far the most efficient way to get up-to-speed quickly. For that reason, I really value having a large network of diverse people I can contact with questions. I put a fair amount of effort into cultivating such a network.</p>
</blockquote>
<h2 id="hard-problems"><a class="toclink" href="#hard-problems">Hard problems</a></h2>
<p>I imagine that you’ll sometimes have to grapple with problems that are sufficiently hard that it feels like you didn’t make any tangible progress on them (or on how to approach them) for a week or more. How do you stay optimistic and motivated? How and when do you “escalate” in some fashion – say, discuss hiring a freelance expert on some other field?</p>
<h3 id="jason-schukraft_6"><a class="toclink" href="#jason-schukraft_6">Jason Schukraft</a></h3>
<blockquote>
<p>I’m fortunate that my work is almost always intrinsically interesting. So even if I don’t make progress on a problem, I continue to be motivated to work on it because the work itself is so very pleasant. That said, as I’ve emphasized above, when I’m stuck, I find it most helpful to talk to lots of people about the problem.</p>
</blockquote>
<h3 id="alex-lintz_4"><a class="toclink" href="#alex-lintz_4">Alex Lintz</a></h3>
<blockquote>
<p>I have a maybe-controversial take that research (even in <span class="caps">LT</span> space) is motivated largely by signalling and status games. From this view the advice many gave about talking to people about it sounds good. Then you generate some excitement as you’re able to show someone else you’re smart enough to solve it, or they get excited to share what they know, etc. I think if you had a nice working group on any topic, no matter how boring, everyone would get super excited about it. In general, connecting the solution to a hard problem to social reward is probably going to work well as a motivator by this logic.</p>
</blockquote>
<h3 id="michael-aird_4"><a class="toclink" href="#michael-aird_4">Michael Aird</a></h3>
<blockquote>
<p>I’m not actually sure if the <em>precise</em> problem you’re describing resonates with me. I definitely often feel very uncertain about:</p>
<ul>
<li>whether the goal I’m striving towards really matters at all</li>
<li>even if so, whether it’s a goal worth <em>prioritising</em></li>
<li>whether <em>I</em> should prioritise it (is it my comparative advantage?)</li>
<li>whether anything I produce in pursuing this goal will be of any use to anyone</li>
</ul>
<p>But I’m not sure there have been cases where, for a week or more, I didn’t feel like I was at least progressing towards:</p>
<ul>
<li>having the sort of output I had planned or now planned to produce(setting aside the question of whether that output will be useful to anyone), and/or</li>
<li>deciding (for good reason) to not bother trying to create that sort of output</li>
</ul>
<p>Note that I’d count as “progress” cases where I explored some solutions/options that I thought might work/be useful for X, and all turned out to be miserable wastes of time, so I can at least rule those out and try something else next week. I’d also count cases where I learned other potentially useful things in the process of pursuing dead ends, and that knowledge seems likely to somehow benefit this or other projects.</p>
<p>It <em>is</em> often the case that my estimate of how many remaining days something will take is longer at the end of the week than it was at the beginning of the week. But this is usually coupled with me thinking that I <em>have</em> made some sort of progress – I just <em>also</em> realised that some parts will be harder than I thought, or that I should do a more thorough job than I’d planned, or something like that.</p>
<p>(But I feel like maybe I’m just interpreting your question differently to what you intended.)</p>
</blockquote>
<h2 id="emotional-motivators"><a class="toclink" href="#emotional-motivators">Emotional motivators</a></h2>
<p>It’s easy to be motivated on a System 2 basis by the importance of the work, but sometimes that fails to carry over to System 1 when dealing with some very removed or specific work – say, understanding some obscure proof that is relevant to <span class="caps">AI</span> safety along a long chain of tenuous probabilistic implications. Do you have tricks for how to stay System 1 motivated in such cases – or when do you decide that a lack of motivation may actually mean that something is wrong with the topic and you should question whether it is sufficiently important?</p>
<h3 id="jason-schukraft_7"><a class="toclink" href="#jason-schukraft_7">Jason Schukraft</a></h3>
<blockquote>
<p>When I reflect on my life as a whole, I’m happy that I’m in a career that aims to improve the world. But in terms of what gets me out of bed in the morning and excited to work, it’s almost never the impact I might have. It’s the intrinsically interesting nature of my work. I almost certainly would not be successful if I did not find my research to be so fascinating.</p>
</blockquote>
<h3 id="david-bernard_5"><a class="toclink" href="#david-bernard_5">David Bernard</a></h3>
<blockquote>
<p>My main trick for dealing with this is to always plan my day the night before. I let System 2 Dave work out what is important and needs to be done and put blocks in the calendar for these things. When System 1 Dave is working the next day, his motivation doesn’t end up mattering so much because he can easily defer to what System 2 Dave said he should do. I don’t read too much into lack of System 1 motivation, it happens and I haven’t noticed that it is particularly correlated with how important the work is, it’s more correlated with things like how scary it is to start some new task and irrelevant things like how much sunlight I’ve been getting.</p>
</blockquote>
<h3 id="alex-lintz_5"><a class="toclink" href="#alex-lintz_5">Alex Lintz</a></h3>
<blockquote>
<p>I’ve been thinking a lot recently about what I’m calling “incentive landscaping.” The basic idea is that your system 2 has a bunch of things it wants to do (e.g. have impact). Then you can shape your incentive landscape such that your system 1 is also motivated to do the highest impact things. Working for someone who shares your values is the easiest way to do this as then your employer and peers will reward you (either socially or with promotions) for doing things which are impact-oriented. This still won’t be perfectly optimized for impact but it gets you close. Then you can add in some extra motivators like a small group you meet with to talk about progress on some thing which seems badly motivated, or ask others to make your reward conditional on you completing something your system 2 thinks is important. Still early days for me on this though and I think it’s a really hard thing to get right.</p>
</blockquote>
<h3 id="michael-aird_5"><a class="toclink" href="#michael-aird_5">Michael Aird</a></h3>
<blockquote>
<p>(Disclaimer: I’m just reporting on my own experience, and think people will vary a lot in this sort of area, so none of the following is even slightly a <em>recommendation</em>.)</p>
<p>In general:</p>
<ul>
<li>Personally, I seem to just find it pretty natural to spend a lot of hours per week doing work-ish things</li>
<li>I tend to be naturally driven to “work hard” (without it necessarily feeling much like working) by intellectual curiosity, by a desire to produce things I’m proud of, and by a desire for positive attention (especially but not only from people whose judgement I particularly respect)</li>
<li>That third desire in particular can definitely become a problem, but I try to keep a close eye on it and ensure that I’m channeling that desire towards actions I actually endorse on reflection</li>
<li>I <em>do</em> get run down sometimes, and sometimes this has to do with too many hours per week for too many weeks in a row. But the things that seem more liable to run me down are feeling that I lack sufficient autonomy in what I do, how, and when; or feeling that what I’m doing isn’t valuable; or feeling that I’m not developing skills and knowledge I’ll use in future</li>
<li>That last point means that one type of case in which I <em>do</em> struggle to be motivated is cases where I know I’m going to switch away from a broad area after finishing some project, and that I’m unlikely to use the skills involved in that project again.<ul>
<li>In these cases, even if I know that finishing that project to a high standard would still be valuable and is worth spending time on, it can be hard for me to be internally motivated to do so, because it no longer feels like doing so would “level me up” in ways I care about.</li>
</ul>
</li>
<li>I seem to often become intensely focused on a general area in an ongoing way (until something switches my focus to another area), and just continually think about it, in a way that feels positive or natural or <a href="https://en.wikipedia.org/wiki/Flow_(psychology)">flow</a>-like or something</li>
<li>This happened for stand-up comedy, then for psychology research, then for teaching, then for <span class="caps">EA</span> stuff (once I learned about <span class="caps">EA</span>)<ul>
<li>(The other points above likewise applied during each of those four “phases” of my adult life)</li>
</ul>
</li>
</ul>
<p>Luckily, the sort of work I do now:</p>
<ul>
<li>is very intellectually stimulating</li>
<li>involves producing things I’m (at least often!) proud of</li>
<li>can bring me positive attention</li>
<li>allows me a sufficient degree of autonomy</li>
<li>seems to me to be probably the most valuable thing I could realistically be doing at the moment (in expectation, and with vast uncertainty, of course)</li>
<li>involves developing skills and knowledge I expect I might use in future</li>
</ul>
<p>That means it’s typically been relatively easy for me to stay motivated. I feel very fortunate both to have the sort of job and “the sort of psychology” I’ve got. I think many people might, through no fault of their own, find it harder to be emotionally motivated to spend lots of hours doing valuable work, even when they know that that work would be valuable and they have the skills to do it. Unfortunately, we can’t entirely choose what drives us, when, and how.</p>
<p>(There’s also a scary possibility that my tendency so far to be easily motivated to work on things I think are valuable is just the product of me being relatively young and relatively new to <span class="caps">EA</span> and the areas I’m working in, and that that tendency will fade over time. I’d bet against that, but could be wrong.)</p>
</blockquote>
<h2 id="typing-speed"><a class="toclink" href="#typing-speed">Typing speed</a></h2>
<p>I have this pet theory that a high typing speed is important for some forms of research that involves a lot of verbal thinking (e.g., maybe not maths). The idea is that our memory is limited, so we want to take notes of our thoughts. But handwriting is slow, and typing is only mildly faster, so unless one thinks slowly or types very fast, there is a disconnect that causes continual stalling, impatience, forgotten ideas, and prevents the process from flowing. Does that make any intuitive sense to you? Do you have any tricks (e.g., dictation software)?</p>
<h3 id="jason-schukraft_8"><a class="toclink" href="#jason-schukraft_8">Jason Schukraft</a></h3>
<blockquote>
<p>No idea what my typing speed is, but it doesn’t feel particularly fast, and that doesn’t seem to handicap me. I’ve always considered myself a slow thinker, though.</p>
</blockquote>
<h3 id="david-bernard_6"><a class="toclink" href="#david-bernard_6">David Bernard</a></h3>
<blockquote>
<p>I struggle to imagine typing speed being a binding constraint on research productivity since I’ve never found typing speed to be a problem for getting into flow, but when I just checked my wpm was 85 so maybe I’d feel different if it was slower. When I’m coding the vast majority of my time is spent thinking about how to solve the problem I’m facing, not typing the code that solves the problem. When I’m writing first drafts, I think typing speed is a bit more helpful for the reasons you mention, but again more time goes into planning the structure of what I want to say and polishing, than the first pass at writing where speed might help.</p>
</blockquote>
<h3 id="linchuan-zhang"><a class="toclink" href="#linchuan-zhang">Linchuan Zhang</a></h3>
<blockquote>
<p>I think my own belief is that typing speed is probably less important than you appear to believe, but I care enough about it that I logged 53 minutes of typing practice on keybr this year (usually during moments where I’m otherwise not productive and just want to get “in flow” doing something repetitive), and I suspect I still can productively use another 3–5 hours of typing practice next year even if it trades off against deep work time (and presumably many more hours than that if it does not).</p>
</blockquote>
<h3 id="alex-lintz_6"><a class="toclink" href="#alex-lintz_6">Alex Lintz</a></h3>
<blockquote>
<p>At least when I’m doing reflections or broad thinking I often circumvent this by doing a lot of voice notes with Dragon. That way I can type at the speed of thought. It’s never perfect but ~ 97% of it is readable so it’s good enough. Then if you want to actually have good notes you go through and summarize your long jumble of semi-coherent thoughts into something decent sounding. This has the side of effect of some spaced repetition learning as well!</p>
</blockquote>
<h3 id="michael-aird_6"><a class="toclink" href="#michael-aird_6">Michael Aird</a></h3>
<blockquote>
<p>I’d be surprised if typing speed was a big factor explaining differences in how much different researchers produce, or in their ability to produce certain types of output. (But of course, that claim is pretty vague – how surprised would I be? What do I mean by “big factor?”)</p>
<p>But I just did a <a href="https://www.typingtest.com/">typing test</a>, and got 92 <span class="caps">WPM</span> (with “medium” words, and 1 typo), which is apparently high. So perhaps I’m just taking that for granted and not recognising how a slower typing speed could’ve limited me. Hard to say.</p>
</blockquote>
<h2 id="obvious-questions"><a class="toclink" href="#obvious-questions">Obvious questions</a></h2>
<p><a href="http://mindingourway.com/obvious-advice/">Nate Soares</a> has an essay on “obvious advice.” Michael Aird mentioned that in many cases he just wanted to follow up on some obvious ideas. They were obvious in hindsight, but evidently they hadn’t been obvious to anyone else for years. Is there a distinct skill of “noticing the obvious ideas” or “noticing the obvious open questions”? And can it be trained or turned into a repeatable process?</p>
<h3 id="jason-schukraft_9"><a class="toclink" href="#jason-schukraft_9">Jason Schukraft</a></h3>
<blockquote>
<p>Yeah, I think there is a general skill of “noticing the obvious.” I don’t think I’m great at it, but one thing I do pretty often is reflect on the sorts of things that appear obvious now that weren’t obvious to smart people ~200 years ago.</p>
</blockquote>
<h3 id="linchuan-zhang_1"><a class="toclink" href="#linchuan-zhang_1">Linchuan Zhang</a></h3>
<blockquote>
<p>I suspect that while sometimes ignoring/not noticing “obvious questions/advice” etc is coincidental unforced errors, more often than not there is some form of motivated reasoning going on behind the scenes (e.g., because this story will invalidate a hypothesis I’m wedded to, because it involves unpleasant tradeoffs, because some beliefs are lower prestige, because it makes the work I do seem less important, etc). I think training myself carefully to notice these things has been helpful, though I suspect I still miss a lot of obvious stuff.</p>
</blockquote>
<h3 id="michael-aird_7"><a class="toclink" href="#michael-aird_7">Michael Aird</a></h3>
<blockquote>
<p><em>(Just my personal, current, <a href="https://forum.effectivealtruism.org/posts/WJ3rBBkawoTJcY642/ask-rethink-priorities-anything-ama?commentId=uQx7HNJdx76x4aRPR">non-expert</a> thoughts, as always. Also, I’m not sure I’m addressing precisely the question you had in mind.)</em></p>
<p>A summary of my recommendations in this vicinity:</p>
<ol>
<li>If people want to do research and want a menu of ideas/questions to work on, including ideas/questions that seem like they obviously should have a bunch of work on them but don’t yet, they could check out this <a href="https://forum.effectivealtruism.org/posts/MsNpJBzv5YhdfNHc9/a-central-directory-for-open-research-questions">central directory for open research questions</a>, and/or an overlapping <a href="https://80000hours.org/articles/research-questions-by-discipline/">80,000 Hours post</a>.</li>
<li>
<p>If people want to discover “new” instances of such ideas/questions, one option might be to just try to notice ideas/variables/assumptions that seem important to some people’s beliefs, but that seem debatable and vague, have been contested by others, and/or haven’t been stated explicitly and fleshed out.</p>
<ul>
<li>One way to do this might be to have a go at rigorously, precisely writing out the arguments that people seem to be acting as if they believe, in order to spot the assumptions that seem required but that those people haven’t stated/emphasised.</li>
<li>One could then try to explore those assumptions in detail, either just through more fleshed-out “armchair reasoning,” or through looking at relevant empirical evidence and academic work, or through some mixture of those things.</li>
<li>I think this is a big part of what I’ve done this year.</li>
<li><a href="https://docs.google.com/document/d/10uudOQx19NCLrnrySfozRbOIDVpEQxeI-d8zpZi1qhw/edit">Here’s</a> one example of a piece of my own work which came from roughly that sort of process.</li>
</ul>
</li>
</ol>
<p>I’ll add more detailed thoughts below.</p>
<hr>
<p>I interpret this question as being focused on cases in which an idea/open question seems like it should’ve been obvious, or seems obvious in retrospect, yet it has been neglected so far. (Or the many cases we should assume still exist in which the idea/question is <em>still</em> neglected, but <em>would</em> – if and when finally tackled – seem obvious.)</p>
<p>It seems to me that there are two major types of such cases:</p>
<ul>
<li><strong>Unnoticed:</strong> Cases in which the ideas/open questions haven’t even been noticed by almost anyone</li>
<li>Or at least, almost anyone in the relevant community/field.<ul>
<li>So I’d still say an idea counts as “unnoticed” for these purposes even if, for example, a very similar ideas has been explored thoroughly in sociology, but no one in longtermism has noticed that that idea is relevant to some longtermist issue, nor independently arrived at a similar idea.</li>
</ul>
</li>
<li><strong>Noticed yet neglected:</strong> Cases in which the ideas/open questions <em>have</em> been noticed, but no one has really fleshed them out or tackled them much</li>
<li>E.g., a fair number of longtermists have <em>noticed</em> that the question of how likely various types of recovery are from various types of civilizational collapse. But as far as I’m aware, there was nothing even approaching a thorough analysis of the question until some recent still-in-progress work, and there’s still room for much more work here.<ul>
<li>More thoughts and notes on this <a href="https://forum.effectivealtruism.org/posts/qY5q2QTG44avBbNKn/modelling-the-odds-of-recovery-from-civilizational-collapse">here</a> and <a href="https://www.youtube.com/watch?v=SbhDfo2vt_g">here</a>.</li>
</ul>
</li>
<li>Another example is questions related to how likely global, stable totalitarianism is; what factors could increase or decrease the odds of that; and what to do about this. Some people have highlighted such questions (<a href="https://www.youtube.com/watch?v=JfpTywk6chw&feature=youtu.be">including</a> but not only in the context of advanced <span class="caps">AI</span>), but I’m not aware of any detailed work on them.</li>
</ul>
<p>This is really more a continuum than a binary distinction. In almost all cases, there’s probably been <em>someone</em> in a relevant community who’s at least briefly noticed <em>something</em> relevant. But sometimes it’ll just be that something kind-of relevant has been discussed verbally a few times and then forgotten, while other times it’ll be that people have prominently highlighted pretty precisely the relevant open question, yet no one has actually worked on it. (And of course there’ll be many cases in between.)</p>
<hr>
<p>For “noticed yet neglected” ideas/questions, recommendation 1 from above will be more relevant: people could find many ideas/questions of this type in this <a href="https://forum.effectivealtruism.org/posts/MsNpJBzv5YhdfNHc9/a-central-directory-for-open-research-questions">central directory for open research questions</a>, and just get cracking on them.</p>
<p>That directory is like a map pointing the way to many trees that might be full of low-hanging fruit <em>that would’ve been plucked by now in a better world</em>. And I really would predict that a lot of EAs could do valuable work by just having a go at those questions. (I’m less confident that this is the <em>most</em> valuable thing lots of EAs could be doing, and each person would have to think that through for themselves, in light of their specific circumstances. <a href="https://www.youtube.com/watch?v=FdrTMUQdBSg">See also</a>.)</p>
<p>So we don’t necessarily need <em>all</em> <span class="caps">EA</span>-aligned researchers to try to cultivate a skill of “noticing the ideas that should’ve been tackled/fleshed out already” (though I’m sure <em>some</em> should). Some could just focus on actually exploring the ideas that <em>have</em> been noticed but <em>still</em> haven’t been tackled/fleshed out.</p>
<hr>
<p>For “unnoticed” ideas/questions, recommendation 2 from above will be more relevant.</p>
<p>I think this dovetails somewhat with <a href="https://80000hours.org/podcast/episodes/ben-garfinkel-classic-ai-risk-arguments/">Ben Garfinkel</a> calling for [1] more people to just try to rigorously write up more detailed versions of arguments about <span class="caps">AI</span> risk that often float around in sketchier or briefer form. (Obviously brevity is better than length, all else held equal, but often a few pages isn’t enough to give an idea proper treatment.)</p>
<hr>
<p>There are at least two other approaches for finding “unnoticed” ideas/questions which seem to have sometimes worked for me, but which I’m less sure would often be useful for many people, and less sure I’ll describe clearly. These are:</p>
<ul>
<li>Trying to sketch out causal diagrams of the pathway to something (e.g., an existential catastrophe) happening</li>
<li>I think that doing something like this has sometimes helped me notice there there are:<ul>
<li>assumptions or steps missing in the standard/fleshed-out stories of how something might happen,</li>
<li>alternative pathways by which something could happen, and/or</li>
<li>alternative/additional outcomes that may occur</li>
</ul>
</li>
<li><a href="https://forum.effectivealtruism.org/posts/TfRexamDYBqSwg7er/causal-diagrams-of-the-paths-to-existential-catastrophe">See also</a></li>
<li>Trying to define things precisely, and/or to precisely distinguish concepts from each other, and seeing if anything interesting falls out</li>
<li>Here’s an abstract example, but one which matches various real examples that have happened for me:<ul>
<li>I try to define X, but then notice that that definition would fail to cover some cases of what I’d usually think of as X, and/or that it <em>would</em> cover some cases of what I’d usually think of as Y (which is a distinct concept).</li>
<li>This makes me realise that X and/or Y might be able to take somewhat different forms or occur via different pathways to what was typically considered, or that there’s actually an extra requirement for X or Y to happen that was typically ignored.</li>
</ul>
</li>
<li>I feel like it’d be easy to misinterpret my stance here.<ul>
<li>I actually think that definitions will never or almost never really be “perfect,” and I agree with the ideas in <a href="https://www.lesswrong.com/posts/WBw8dDkAWohFjWQSk/the-cluster-structure-of-thingspace">this post</a> (see also <a href="https://en.wikipedia.org/wiki/Family_resemblance">family resemblance</a>). And I think that many debates over definitions are largely nitpicking and wasting time.</li>
<li>But I <em>also</em> think that, in many case, being clearer about definitions can substantially benefit both thought and communication.</li>
</ul>
</li>
</ul>
<hr>
<p>I should again mention that I’m only ~ 1.5 years into my research career, so maybe I’ll later change my mind about a bunch of those points, and there are probably a lot of useful things that could be said on this that I haven’t said.</p>
<p>[1] See the parts of <a href="https://80000hours.org/podcast/episodes/ben-garfinkel-classic-ai-risk-arguments/#transcript">the transcript</a> after Howie asks “Do you know what it would mean for the arguments to be more sussed out?”</p>
</blockquote>
<h2 id="tiredness-focus-etc"><a class="toclink" href="#tiredness-focus-etc">Tiredness, focus, etc.</a></h2>
<p>We sometimes get tired or have trouble focusing. Sometimes this happens even when we’ve had enough sleep (just to get an obvious solution out of the way: sleep/napping). What are your favorite things to do when focusing seems hard or you feel tired? Do you use any particular nootropics, supplements, air quality monitor, music, or exercise routine?</p>
<h3 id="jason-schukraft_10"><a class="toclink" href="#jason-schukraft_10">Jason Schukraft</a></h3>
<blockquote>
<p>Regular exercise certainly helps. Haven’t tried anything else. Mostly I’ve just acclimated to getting work done even though I’m tired. (Not sure I would recommend that “solution,” though!)</p>
</blockquote>
<h3 id="david-bernard_7"><a class="toclink" href="#david-bernard_7">David Bernard</a></h3>
<blockquote>
<p>My favourite thing to do is to stop working! Not all days can be good days and I became a lot happier and more productive when I stopped beating myself up for having bad days and allowed myself to take the rest of the afternoon off.</p>
</blockquote>
<h3 id="linchuan-zhang_2"><a class="toclink" href="#linchuan-zhang_2">Linchuan Zhang</a></h3>
<blockquote>
<p>I haven’t figured this out yet and am keen to learn from my coworkers and others! Right now I take a lot of caffeine and I suspect if I were more careful about optimization I should be cycling drugs over a weekly basis rather than taking the same one every day (especially a drug like caffeine that has tolerance and withdrawal symptoms).</p>
</blockquote>
<h3 id="alex-lintz_7"><a class="toclink" href="#alex-lintz_7">Alex Lintz</a></h3>
<blockquote>
<p>I’ve had lot’s of ongoing and serious problems with fatigue and have tried many interventions. Certainly caffeine (ideally with l-theanine) is a nice thing to have but tolerance is an issue. Right now what seems to work for me (no idea why) is a greens powder called Athletic Greens. I’m also trying pro/prebiotics which might be helping. Magnesium supplementation also might have helped. A medication I was taking was causing some problems as well and causing me to have some really intense fatigue on occasion (again, probably…). It’s super hard to isolate cause and effect in this area as there are so many potential causes. I’d say it’s worth dropping a lot of money on different supplements and interventions and seeing what helps. If you can consistently increase energy by 5–10% (something I think is definitely on the table for most people), that adds up really quickly in terms of the amount of work you can get done, happiness, etc. Ideally you’d do this by introducing one intervention at a time for 2-4 weeks each. I haven’t had patience for that and am currently just trying a few things at once, then I figure I can cut out one at a time and see what helped. Things I would loosely recommend trying (aside from exercise, sleep, etc): Prebiotics, good multivitamins, checking for food intolerances, checking if any pills you take are having adverse effects.</p>
<p>I do also work through tiredness sometimes and find it helpful to do some light exercise (for me, games in <span class="caps">VR</span>) to get back some energy. That also works as a decent gauge for whether I’ll be able to push past the tiredness. If playing 10 min of Beatsaber feels like a chore, I probably won’t be able to work.</p>
<p>How you rest might also be important. E.g. might need time with little input so your default mode network can do it’s thing. No idea how big of a deal this is but I’ve found going for more walks with just music (or silence) to maybe be helpful, especially in that I get more time for reflection.</p>
<p>I’ve also been experimenting with measuring heart rate variability using an app called Welltory. That’s been kind of interesting in terms of raising some new questions though I’m still not sure how I feel about it/how accurate it is for measuring energy levels.</p>
</blockquote>
<p>I’ve bought some Performance Lab products (following a recommendation from Alex in a private conversation). They have <a href="https://www.vaga.org/health/performance-lab-whole-food-multi-review/">better reviews on Vaga</a> and are a bit cheaper than the Athletic Greens.</p>
<h3 id="michael-aird_8"><a class="toclink" href="#michael-aird_8">Michael Aird</a></h3>
<blockquote>
<p>I find that being tired makes my mind wander a lot when reading longform things (e.g., papers, posts, not things like Slack messages or emails), so when I’m tired I usually try to do things other than reading.</p>
<p>If I’m just a bit or moderately tired, I usually find I’m still about as able to write as normal. If I’m very tired, I’ll still often be able to write quickly, but then when I later read what I wrote I’ll feel that it was unclear, poorly structured, and more typo-strewn than usual. So when very tired, I try to avoid writing longform things (e.g., actual research outputs).</p>
<p>Things I find I’m still pretty able to do when tired include commenting on documents people want input on (I think I’m more able to focus on this than on regular reading because it’s more “interactive” or something), writing things like <span class="caps">EA</span> Forum comments, replying to emails and Slack messages and the like, doing miscellaneous admin-y tasks, and reflecting on the last week/month and planning the next. So I often do a disproportionate amount of such tasks during evenings or during days when I’m more tired than normal, and at other times do a disproportionate amount of reading and “substantive” writing.</p>
<p>Also, I’m fortunate enough to have flexible hours. So sometimes I just work less on days when I’m tired (perhaps spending more time with my wife), and then make up for it on other days.</p>
</blockquote>How Might Better Collective Decision-Making Backfire?2020-12-16T00:00:00+00:002020-12-16T00:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2020-12-16:/epistemic-hazards.html<p><a href="https://forum.effectivealtruism.org/posts/nHyuxZHBFBSDERwoz/how-might-better-collective-decision-making-backfire">I’ve asked the titular question in the <span class="caps">EA</span> Forum</a>. I also posted some answers of my own from the get go, but other people added more, and I, too, posted another answer later. You can find all answers in the <span class="caps">EA</span> Forum. (I’ve reproduced my question and my initial answers here.)</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#structure">Structure</a></li>
<li><a href="#summaries">Summaries</a><ul>
<li><a href="#collaborative-systems">Collaborative Systems</a><ul>
<li><a href="#direct-effects">Direct Effects</a></li>
<li><a href="#incidental-concomitant-effects">Incidental, Concomitant Effects</a></li>
</ul>
</li>
<li><a href="#better-decisions-or-epistemics">Better Decisions or Epistemics</a></li>
</ul>
</li>
<li><a href="#conclusion">Conclusion</a></li>
<li><a href="#the-question">The Question</a></li>
<li><a href="#the-answers">The Answers</a><ul>
<li><a href="#legibility">Legibility</a></li>
<li><a href="#modified-ultimatum-game">Modified Ultimatum Game</a></li>
<li><a href="#benefitting-unscrupulous-people">Benefitting Unscrupulous People</a></li>
<li><a href="#psychological-effects">Psychological Effects</a></li>
<li><a href="#social-effects">Social Effects</a></li>
</ul>
</li>
</ul>
</div>
<h2 id="structure"><a class="toclink" href="#structure">Structure</a></h2>
<p>There is a certain structure to the answers or to the effects that the answers anticipate.</p>
<p>Better collective decision-making has the two parts of being <strong>collective</strong> and being <strong>better</strong>. These have their individual failure modes:</p>
<ol>
<li>Failure modes of <strong>collective, cooperative, or collaborative systems</strong> can be <ol>
<li>failure modes that <strong>follow directly from the collaborative nature of the system</strong>, and, a bit more speculatively,</li>
<li>they can also be failure modes that <strong>follow from a potential effect such systems might have to incentivize or facilitate more collaboration generally</strong>.</li>
</ol>
</li>
<li>Failure modes of <strong>better decisions</strong> are a more monolithic group. They could be divided into adverse effects vs. intentional abuse or into effects of the process of determining good decisions vs. the decisions themselves, but in both cases, one category would contain only one answer at the moment, so it doesn’t seem worthwhile to make this distinction.</li>
</ol>
<h2 id="summaries"><a class="toclink" href="#summaries">Summaries</a></h2>
<p>Here I’ll briefly summarize the most worrying answers. These summaries may seem weakly argued or cryptic. If so, please see the full text for further clarifications.</p>
<h3 id="collaborative-systems"><a class="toclink" href="#collaborative-systems">Collaborative Systems</a></h3>
<h4 id="direct-effects"><a class="toclink" href="#direct-effects">Direct Effects</a></h4>
<p>This corresponds to 1.a. of the structure above. Collaborative systems may backfire in at least the following ways:</p>
<ol>
<li><strong>Legibility.</strong> Communication between people will make it necessary to make knowledge more “legible” (in the <em><a href="https://slatestarcodex.com/2017/03/16/book-review-seeing-like-a-state/">Seeing Like A State</a></em> sense<sup id="fnref3:scott-alexander-disclaimer"><a class="footnote-ref" href="#fn:scott-alexander-disclaimer">1</a></sup>). This can have several adverse effects:<ol>
<li><strong>Discrimination against illegible values.</strong> It may be very hard to model the full complexity of moral intuitions of people. In practice, people tend toward systems that greatly reduce the dimensionality of what people typically care about. The result are utils, DALYs, <span class="caps">SWB</span>, consumption, life years, probability of any existential catastrophe, etc. Collaborative systems would incentivize such low-dimensional measures of value, and through training, people may actually come to care about them more. It’s contentious and a bit circular to ask whether this is good or bad. But at least it’s not clearly neutral or good.</li>
<li><strong>Centralized surveillance.</strong> My answer titled “Legibility” argues that collaborative methods will make it necessary to make considerations and values more legible than they are now so they can be communicated and quantified. This may also make them more transparent and thus susceptible to surveillance. That, in turn, may enable more powerful authoritarian governments, which may steer the world into a dystopian lock-in state.</li>
</ol>
</li>
<li><strong>Averaging effect</strong>. Maybe there are people who are particularly inclined toward outré opinions. These people will be either unusually right or unusually wrong for their time. Maybe there’s more benefit in being unusually right than there is harm in being unusually wrong (e.g., thanks to the law). And maybe innovation toward most of what we care about is carried by unusually right people. (I’m thinking of Newton here, whose bad ideas didn’t seem to have much of an effect compared to his good ideas.) Collaborative systems likely harness – explicitly or implicitly – some sort of wisdom of the crowds type of effect. But such an effect is likely to average away the unusually wrong and the unusually right opinions. So such systems might slow progress.</li>
<li><strong>Silencing of lone dissenters.</strong> The averaging effect above may be used intentionally to silence individual or sufficiently small groups that disagree with the majority. E.g., Anna might only be productive at 20°C with <span class="caps">SD</span> of 1°C. She joins a company that has determined in a long, collaborative process that their employees are most productive at 22°C with <span class="caps">SD</span> of 5°C, and so has put the temperature in the office at 22°C. If Anna could restart that long collaborative process, the optimal temperature may now be 21°C. But her boss tells her to just go ahead and rerun the process in her free time, which both know is infeasible given how much work it is and how many people it involves.</li>
<li><strong>More power to the group.</strong> It might be that the behavior of groups (e.g., companies) is generally worse (e.g., more often antisocial) than that of individuals. Collaborative systems would shift more power from individuals to groups. So that may be undesirable.</li>
</ol>
<h4 id="incidental-concomitant-effects"><a class="toclink" href="#incidental-concomitant-effects">Incidental, Concomitant Effects</a></h4>
<p>This corresponds to 1.b. of the structure above. Collaborative systems may incentivize cooperation and collaboration (this is an additional inferential step, though), which are probably vastly positive on balance but may backfire in at least the following ways. I didn’t originally include answers like these because of the additional inferential step, which I think is an unlikely one too, so these are probably particularly incomplete.</p>
<ol>
<li><strong>Collusion.</strong> Cooperation can enable collusion, cartels, and bribes. Various antisocial behaviors that wouldn’t be as easy if there were no way to trust the other party. (See <a href="https://forum.effectivealtruism.org/posts/nHyuxZHBFBSDERwoz/how-might-better-collective-decision-making-backfire?commentId=yuP6kXri7TW9P7Aso">this comment</a>.)</li>
<li><strong>Exploitation, deception, and coercion.</strong> Promoting cooperation will go through promoting capacities that enable cooperation. Such capacities include understanding, recognizing honesty and deception, and commitment. Those same capacities can also be used to understand others’ vulnerabilities, deceive them, and commit to threats. Those, however, can also be prosocial again, e.g., in the case of laws that are enforced through a threat of a fine. (See <a href="https://forum.effectivealtruism.org/posts/nHyuxZHBFBSDERwoz/how-might-better-collective-decision-making-backfire?commentId=yuP6kXri7TW9P7Aso">this comment</a>.)</li>
<li><strong>Reduced competition.</strong> Cooperation may reduce competition, which would’ve facilitated learning or training. (See <a href="https://forum.effectivealtruism.org/posts/nHyuxZHBFBSDERwoz/how-might-better-collective-decision-making-backfire?commentId=yuP6kXri7TW9P7Aso">this comment</a>.)</li>
<li><strong>Value lock-in.</strong> The majority may use the enhanced coordination to lock-in its current values and thereby prevent whatever we regard as moral progress from continuing. (See <a href="https://forum.effectivealtruism.org/posts/nHyuxZHBFBSDERwoz/how-might-better-collective-decision-making-backfire?commentId=ptFfH4Y7d7Wpd2seJ">this answer</a>.)</li>
</ol>
<h3 id="better-decisions-or-epistemics"><a class="toclink" href="#better-decisions-or-epistemics">Better Decisions or Epistemics</a></h3>
<p>This corresponds to 2. of the structure above. Better decisions or better epistemics may backfire in at least the following ways:</p>
<ol>
<li><strong>Overconfidence may be necessary for motivation.</strong> E.g., entrepreneurs are said to be overconfident that their startup ideas will succeed. Maybe increased rationality (individual or collective) will stifle innovation. (See section <a href="#psychological-effects">Psychological Effects</a>.)</li>
<li><strong>The sunk cost fallacy may be necessary for sustained motivation.</strong> E.g., EAs are known to start a lot of projects and abandon them again quickly when they learn of something even better they could do. That might continue indefinitely, so that no project gets off the ground. (See section <a href="#psychological-effects">Psychological Effects</a>.)</li>
<li><strong>Decision theoretic catch-22s.</strong> The Soares and Fallenstein paper “<a href="https://arxiv.org/pdf/1507.01986.pdf">Toward Idealized Decision Theory</a>” presents an example where a dumber agent who knows that they are playing against a smarter agent can exploit the fact that the smarter agent knows more about them than they do about the smarter agent. (See section <a href="#modified-ultimatum-game">Modified Ultimatum Game</a>.)</li>
<li><strong>Less experimentation through convergence to best practices.</strong> Good epistemics may make it patently obvious what the best known (no hyphen) practices are. Most actors benefit individually from adopting the best practices. Hence, fewer of them experiment with new practices that they find promising, and so we may get stuck in some local optimum. (See <a href="https://forum.effectivealtruism.org/posts/nHyuxZHBFBSDERwoz/how-might-better-collective-decision-making-backfire?commentId=qbukeM8nKc26wkys6">this answer</a>.)</li>
<li><strong>Benefiting malevolent actors.</strong> Systems to produce better epistemics will likely be value neutral. Hence they can be abused by malevolent actors. (See section <a href="#benefitting-unscrupulous-people">Benefiting Unscrupulous People</a>.)</li>
<li><strong>Conflict through coordination failures.</strong> Better epistemics may function like weapons that can be used against other groups. This may lead to arms races, full-blown conflicts, or may cause negative externalities for groups that can’t defend themselves – nonhumans, future people, et al. (See <a href="https://forum.effectivealtruism.org/posts/nHyuxZHBFBSDERwoz/how-might-better-collective-decision-making-backfire?commentId=EXZ2kec2YQTkmAdig">this answer</a>.)</li>
<li><strong>Social ramifications.</strong> Using systems to generate better decisions may necessitate actions that have negative social effects. (See section <a href="#social-effects">Social Effects</a>.)</li>
<li><strong>Loss of valuable ambiguity.</strong> It might be that ambiguity plays an important role in social interactions. There is the stylized example where it is necessary to keep the number of rounds of an iterated game a secret or else that knowledge will distort the game. I’ve also read somewhere that there’s the theory that conflicts between two countries can be exacerbated if the countries have too low-quality intelligence about each other but also if they have too high-quality intelligence about each other. But I can’t find the source, so I’m likely to misremember something. Charity evaluators also benefit from ambiguity in that fewer charities would be willing to undergo their evaluation process if the only reason why a charity would either decline it or block the results from being published were reasons that reflect badly on the charity. But there are also good and neutral reasons, so charities will always have plausible deniability.</li>
</ol>
<h2 id="conclusion"><a class="toclink" href="#conclusion">Conclusion</a></h2>
<p>My current take is that the advantages probably outweigh the risks but that it would be negligent not to try to make the risks as low as possible.</p>
<p>My current take is also that the risks from a particular system are probably proportional to the popularity of the system so long as it is marketed to everyone equally. Most likely, this means that we can stay vigilant, observe whether any of these risks or new ones manifest, and then correct course or abandon the project. But it might also mean that in very rare cases we’ll see such run-away growth that we can’t control it anymore.</p>
<p>Anyone might develop such a system at any time, and the independent, almost two-year-old startup <a href="https://www.causal.app">Causal</a> proves that this can happen. So if it were assured that there won’t be any run-away growth, I’d be quite confident that the best approach is to develop such a system, and design and market it carefully with all these risk factors in mind.</p>
<p>But since there is a remote risk of run-away growth, I’m unsure about this conclusion. I’m leaning toward the assessment that the probability of run-away growth is so tiny (and will likely happen within groups that are somewhat cooperative if it happens) that it can be ignored. But maybe one of the best uses of such a system is to test whether it’s a good idea to continue developing the system.</p>
<h2 id="the-question"><a class="toclink" href="#the-question">The Question</a></h2>
<p>I’ve started work on a project that aims to do something like “improving our collective decision-making.”<sup id="fnref:1"><a class="footnote-ref" href="#fn:1">2</a></sup> Broadly, it’s meant to enable communities of people who want to make good decisions to collaboratively work out what these decisions are. Individual rationality is helpful for that but not the focus.</p>
<p>Concretely, it’s meant to make it easier to collaborate on probabilistic models in areas where we don’t have data. You can read more about the vision in <a href="https://www.lesswrong.com/posts/i5BWqSzuLbpTSoTc4/squiggle-an-overview">Ozzie Gooen’s Less Wrong post on Squiggle</a>. But please ignore this for the sake of this question. I hope to make the question and answers to it useful to a broader audience by not narrowly focusing it on the concrete thing I’m doing. Other avenues to improve collective decision-making may be improving prediction markets and developing better metrics for things that we care about.</p>
<p>Before I put a lot of work into this, I would like to check whether this is a good goal in the first place – ignoring tractability and opportunity costs.<sup id="fnref:2"><a class="footnote-ref" href="#fn:2">3</a></sup> By “good” I mean something like “robustly beneficial,” and by “<em>robustly</em> beneficial” I mean something like “beneficial across many plausible worldviews and futures, and morally cooperative.”</p>
<p>My intuition is that it’s about as robustly positive as it gets, but I feel like I could easily be wrong because I haven’t engaged with the question for a long time. Less Wrong and <span class="caps">CFAR</span> seem to have similar goals, though I perceive a stronger focus on individual rationality. So I like to think that people have thought and maybe written publicly about what the major risks are in what they do.</p>
<p>I would like to treat this question like a Stack Exchange question where you can also submit your own answers.[^3] But I imagine there are many complementary answers to this question. So I’m hoping that people can add more, upvote the ones they find particularly concerning, important, or otherwise noteworthy, and refine them with comments.</p>
<p>For examples of pretty much precisely the type of answers I’m looking for, see my answers titled “Psychological Effects” and “The Modified Ultimatum Game.”</p>
<p>A less interesting answer is my answer “Legibility.” I’m less interested in it here because it describes a way in which a system can fail to attain the goal of improving decision-making rather than a way in which the successful realization of that goal backfires. I wanted to include it as a mild counterexample.</p>
<p>If you have ideas for further answers, it would be interesting if you could also think of ways to work around them. It’s usually not beneficial to abandon a project if there is any way in which it can backfire but to work out how the failure mode can be avoided without sacrificing all of the positive effects of the project.</p>
<p>You can also message me privately if you don’t want to post your answer publicly.</p>
<p>Acknowledgements: Thanks for feedback and ideas to Sophie Kwass, Ozzie Gooen, Justin Shovelain, and everyone who answered in the <span class="caps">EA</span> Forum!</p>
<h2 id="the-answers"><a class="toclink" href="#the-answers">The Answers</a></h2>
<h3 id="legibility"><a class="toclink" href="#legibility">Legibility</a></h3>
<p>This is a less interesting failure mode as it is one where the systems that we create to improve our decision-making actually fail to achieve that goal. It’s not one where successfully achieving that goal backfires.</p>
<p>I also think that while this may be a limitation of some collaborative modeling efforts, it’s probably no problem for prediction markets.</p>
<p>The idea is that collaborative systems will always, at some stage, require communication, and specifically communication between brains rather than within brains. To make ideas communicable, they have to be made legible. (Or maybe literature, music, and art are counterexamples.) By <em>legible</em>, I’m referring to the concept from <em><a href="https://slatestarcodex.com/2017/03/16/book-review-seeing-like-a-state/">Seeing Like A State</a></em>.<sup id="fnref:scott-alexander-disclaimer"><a class="footnote-ref" href="#fn:scott-alexander-disclaimer">1</a></sup></p>
<p>In my experience, this can be very limiting. Take for example what I’ll call the <em>Cialdini puzzle</em>:</p>
<blockquote>
<p>Robert Cialdini’s Wikipedia page says “He is best known for his book <em>Influence</em>“. Since its publication, he seems to have spent his time directing an institute to spread awareness of techniques for success and persuasion. At the risk of being a little too cynical – a guy knows the secrets of success, so he uses them to… write a book about the secrets of success? If I knew the secrets of success, you could bet I’d be doing much more interesting things with them. All the best people recommend Cialdini, and his research credentials are impeccable, but I can’t help wondering: If he’s so smart, why isn’t he Emperor?</p>
</blockquote>
<p>It seems to me like a common pattern that for certain activities the ability to do them well is uncorrelated or even anticorrelated with the ability to explain them. Some of that may be just because people want to keep their secrets, but I don’t think that explains much of it.</p>
<p>Hence Robert Cialdini may be > 99th percentile at understanding and explaining social influence, but in terms of doing social influence, that might’ve boosted him from the 40th to the 50th percentile or so. (He says his interest in the topic stems from his being particularly gullible.) Meanwhile, all the people he interviews because they have a knack for social influence are probably 40th to 50th percentile at explaining what they do. I don’t mean that they are average at explaining in general but that what they do is too complex, nuanced, unconscious, intertwined with self-deception, etc. for them to grasp it in a fashion that would allow for anything other than execution.</p>
<p>Likewise, a lot of amazing, famous writers have written books on how to write. And almost invariably these books are… unhelpful. If these writers followed the advice they set down in their own books, they’d be lousy writers. (This is based on a number of Language Log posts on such books.) Meanwhile, some of the most helpful books on writing that I’ve read were written by relatively unknown writers. (E.g., <em>Style: Toward Clarity and Grace</em>.)</p>
<p>My learning of Othello followed a similar trajectory. I got from a Kurnik rating of 1200 up to 1600 quite quickly by reading every explanatory book and text on game strategy that I could find and memorizing hundreds of openings. Beyond that, the skill necessary to progress further becomes too complex, nuanced, and unconscious that, it seems to me, it can only be attained through long practice, not taught. (Except, of course, if the teaching is all about practice.) And I didn’t like practice because it often meant playing against other people. (That is just my experience. If someone is an Othello savant, they may rather feel like some basic visualization practice unlocked the game for them, so that they’d still have increasing marginal utility from training around the area where it started dropping for me.)</p>
<p>Orthography is maybe the most legible illegible skill that I can think of. It can be taught in books, but few people read dictionaries in full. For me it sort of just happened rather suddenly that from one year to the next, I made vastly fewer orthographic mistakes (in German). It seems that my practice through reading must’ve reached some critical (soft) threshold where all the bigrams, trigrams, and exceptions of the language became sufficiently natural and intuitive that my error rate dropped noticeably.</p>
<p>For this to become a problem there’d have to be highly skilled practitioners, like the sort of people Cialdini likes to interview, who are brought together by a team or researchers to help them construct a model of some long-term future trajectory.</p>
<p>These skilled practitioners will do exactly the strategically optimal thing when put in a concrete situation, but in the abstract environment of such a probabilistic model, their predictions may be no better than anyone’s. It’ll take well-honed elicitation methods to get high-quality judgments out of these people, and then a lot of nuance may still be lost because what is elicited and how it fits into the model is probably again something that the researchers will determine, and that may be too low-fidelity.</p>
<p>Prediction markets, on the other hand, tend to be about concrete events in the near future, so skilled practitioners can probably visualize the circumstances that would lead to any outcome in sufficient detail to contribute a high-quality judgment.</p>
<h3 id="modified-ultimatum-game"><a class="toclink" href="#modified-ultimatum-game">Modified Ultimatum Game</a></h3>
<p>A very good example of the sort of risks that I’m referring to is based on a modified version of the ultimatum game and comes from the Soares and Fallenstein paper “<a href="https://arxiv.org/pdf/1507.01986.pdf">Toward Idealized Decision Theory</a>”:</p>
<blockquote>
<p>Consider a simple two-player game, described by Slepnev (2011), played by a human and an agent which is capable of fully simulating the human and which acts according to the prescriptions of [Updateless Decision Theory (<span class="caps">UDT</span>)]. The game works as follows: each player must write down an integer between 0 and 10. If both numbers sum to 10 or less, then each player is paid according to the number that they wrote down. Otherwise, they are paid nothing. For example, if one player writes down 4 and the other 3, then the former gets paid $4 while the latter gets paid $3. But if both players write down 6, then neither player gets paid. Say the human player reasons as follows:</p>
<blockquote>
<p>I don’t quite know how <span class="caps">UDT</span> works, but I remember hearing that it’s a very powerful predictor. So if I decide to write down 9, then it will predict this, and it will decide to write 1. Therefore, I can write down 9 without fear.</p>
</blockquote>
<p>The human writes down 9, and <span class="caps">UDT</span>, predicting this, prescribes writing down 1.</p>
<p>This result is uncomfortable, in that the agent with superior predictive power “loses” to the “dumber” agent. In this scenario, it is almost as if the human’s lack of ability to predict <span class="caps">UDT</span> (while using correct abstract reasoning about the <span class="caps">UDT</span> algorithm) gives the human an “epistemic high ground” or “first mover advantage.” It seems unsatisfactory that increased predictive power can harm an agent.</p>
</blockquote>
<p>A solution to this problem would have to come from the area of decision theory. It probably can’t be part of the sort of collaborative decision-making system that we envision here. Maybe there is a way to make such a problem statement inconsistent because the smarter agent would’ve committed to writing down 5 and signaled that sufficiently long in advance of the game. Ozzie also suggests that introducing randomness along the lines of the <a href="https://en.wikipedia.org/wiki/Madman_theory">madman theory</a> may be a solution concept.</p>
<h3 id="benefitting-unscrupulous-people"><a class="toclink" href="#benefitting-unscrupulous-people">Benefitting Unscrupulous People</a></h3>
<p>A system that improves collective decision making is likely value-neutral, so it can also be used by unscrupulous agents for their nefarious ends.</p>
<p>Moreover unscrupulous people may benefit from it more because they have fewer moral side-constraints. If set A is the set of all ethical, legal, cooperative methods of attaining a goal, and set B is the set of all methods of attaining the same goal, then A ⊆ B. So it should always be as easy or easier to attain a goal by any means necessary than only by ethical, legal, and cooperative means.</p>
<p>Three silver linings:</p>
<ol>
<li>Unscrupulous people probably also have different goals from ours. Law enforcement will block them from attaining those goals, and better decision-making will hopefully not get them very far.</li>
<li>These systems are collaborative, so you can benefit from them more the more people collaborate on them (I’m not saying monotonically, just as a rough tendency). When you invite more people into some nefarious conspiracy, then the risk that one of them blows the whistle increases rapidly. (Though it may depend on the structure of the group. There are maybe some terrorist cells who don’t worry much about whistleblowing.)</li>
<li>If a group is headed by a narcissistic leader, the person may see a threat to their authority in a collaborative decision-making system, so that they won’t adopt it to begin with. (Though it might be that they like that collaborative systems can make it infeasible for individuals to use them to put their individual opinions to the test, so that they can silence individual dissenters. This will depend a lot on implementation details of the system.)</li>
</ol>
<p>More speculatively, we can also promote and teach the system such that everyone who learns to use it also learns about <a href="https://longtermrisk.org/msr">multiverse-wide superrationality alias evidential cooperation in large worlds (<span class="caps">ECL</span>)</a>. Altruistic people with uncooperative agent-neutral goals will reason that they can now realize great gains from trade by being more cooperative or else lose out on them by continuing to defect.</p>
<p>We can alleviate the risk further by marketing the system mostly to people who run charities, social enterprises, prosocial research institutes, and democratic governments. Other people will still learn about the tools, and there are also a number of malevolent actors in those generally prosocial groups, but it may shift the power a bit toward more benevolent people. (<a href="https://forum.effectivealtruism.org/posts/4oGYbvcy2SRHTWgWk/improving-the-future-by-influencing-actors-benevolence">The Benevolence, Intelligence, and Power framework</a> may be helpful in this context.)</p>
<p>Finally, there is the option to make it hard to make models nonpublic. But that would have other downsides, and it’s also unlikely to be a stable equilibrium as others will just run a copy of the software on their private servers.</p>
<h3 id="psychological-effects"><a class="toclink" href="#psychological-effects">Psychological Effects</a></h3>
<p><a href="https://www.lesswrong.com/posts/MkX8D44PFoiNdLMkG/some-potential-dangers-of-rationality-training">Luke Muehlhauser warns</a> that overconfidence and sunk cost fallacy may be necessary for many people to generate and sustain motivation for a project. (But note that the post is almost nine years old.) Entrepreneurs are said to be overconfident that their startup ideas will succeed. Maybe increased rationality (individual or collective) will stifle innovation.</p>
<p>I feel that. When I do calibration exercises, I’m only sometimes mildly overconfident in some credence intervals, and indeed, my motivation usually feels like, “Well, this is a long shot, and why am I even trying it? Oh yeah, because everything else is even less promising.” That could be better.</p>
<p>On a community level it may mean that any community that develops sufficiently good calibration becomes demotivated and falls apart.</p>
<p>Maybe there is a way of managing expectations. If you grow up in an environment where you’re exposed to greatly selection-biased news about successes, your expectations may be so high that any well-calibrated 90th percentile successes that you project may seem disappointing. But if you’re in an environment where you constantly see all the failures around you, the same level of 90th percentile success may seem motivating.</p>
<p>Maybe that’s also a way in which the <span class="caps">EA</span> community backfires. When I didn’t know about <span class="caps">EA</span>, I saw around me countless people who failed completely to achieve my moral goals because they didn’t care about them. The occasional exceptions seemed easy to emulate or exceed. Now I’m surrounded by people who’ve achieved things much greater than my 90th percentile hopes. So my excitement is lower even though my 90th percentile hopes are higher than they used to be.</p>
<h3 id="social-effects"><a class="toclink" href="#social-effects">Social Effects</a></h3>
<p>Beliefs are often entangled with social signals. This can pose difficulties for what I’ll call in the following a “truth-seeking community.”</p>
<p>When people want to disassociate from a disreputable group – say, because they’ve really never had anything to do with the group and don’t want that to change – they can do this in two ways: They can steer clear of anything that is associated with the disreputable group or they can actively signal their difference from the disreputable group.</p>
<p>Things that are associated with the disreputable group are, pretty much necessarily, things that are either sufficiently specific that they rarely come up randomly or things that are common but on which the group has an unusual, distinctive stance. Otherwise these things could not serve as distinguishing markers of the group.</p>
<p>If the disreputable group is small, is distinguished by an unusual focus on a specific topic, and a person wants to disassociate from them, it’s usually enough to steer clear of the specific topic, and no one will assume any association. Others will start out with a prior that the person < 1% likely to be part of the group, and absent signals to the contrary, will maintain that credence.</p>
<p>But if the disreputable group is larger, at least in one’s social vicinity, or the group’s focal topic is a common one, then one needs to countersignal more actively. Others may start out with a prior that the person is ~ 30% likely to be part of the group and may avoid contact with them unless they see strong signals to the contrary. This is where people will find it necessary to countersignal strongly. Moreover, once there is a norm to countersignal strongly, the absence of such a signal or a cheaper signal will be doubly noticeable.</p>
<p>I see two, sometimes coinciding, ways along which that can become a problem. First, the disreputable group may be so because of their values, which may be extreme or uncooperative, and it is just historical contingency that they endorse some distinctive belief. Or second, the group may be disreputable because they have a distinctive belief that is so unusual as to reflect badly on their intelligence or sanity.</p>
<p>The first of these is particularly problematic because the belief can be any random one with any random level of likelihood, quite divorced from the extreme, uncooperative values. It might also not be so divorced, e.g., if it is one that the group can exploit to their advantage if they convince the right people of it. But the second is problematic too.</p>
<p>If a community of people who want to optimize their collective decision-making (let’s call it a “truth-seeking community”) builds sufficiently complex models, e.g., to determine the likelihood of intelligent life re-evolving, then maybe at some point they’ll find that one node in their model (a Squiggle program, a Bayesian network, vel sim.) would be informed by more in-depth research of a question that is usually associated with a disreputable group. They can use sensitivity analysis to estimate the cost it would have to leave the node as it is, but maybe it turns out that their estimate is quite sensitive to that node.</p>
<p>In the first case, in the case of a group that is disreputable by dint of their values, that is clearly a bad catch-22.</p>
<p>But it can also go wrong in the second case, the case of the group that is disreputable because of their unusual beliefs, because people in the truth-seeking community will usually find it <a href="https://www.lesswrong.com/posts/QGkYCwyC7wTDyt3yT/0-and-1-are-not-probabilities">impossible to assign a probability of 0 to any statement</a>. It might be that their model is very sensitive to whether they assign 0.1% or 1% likelihood to a disreputable belief. Then there’s a social cost also in the second case: Even though their credence is low either way, the truth-seeking community will risk being associated with a disreputable group (which may assign > 90% credence to the belief), because they engage with the belief.</p>
<p>I see five ways in which this is problematic:</p>
<ol>
<li>Exploitation of the community by bad actors: The truth-seeking community may be socially adroit, and people will actually grant them some sort of fool’s licence because they trust their intentions. But that may turn out to be exploitable: People with bad intentions may use the guise of being truth-seeking to garner attention and support while subtly manipulating their congregation toward their uncooperative values. (Others may only be interested in the attention.) Hence such a selective fool’s licence may erode societal defenses against extreme, uncooperative values and the polarization and fragmentation of society that they entail. Meanwhile the previously truth-seeking community may be overtaken by such people, who’ll be particularly drawn to its influential positions while being unintimidated with the responsibility that comes with these positions.</li>
<li>Exploitation of the results of the research by bad actors: The same can be exploitable in that the truth-seeking community may find that some value-neutral belief is likely to be true. Regardless of how value-neutral the belief is, the disreputable group may well be able to cunningly reframe it to exploit and weaponize it for their purposes.</li>
<li>Isolation of and attacks on the community: Conversely, the truth-seeking community may also not be sufficiently socially adroit and still conduct their research. Other powerful actors – potential cooperation partners – will consider the above two risks or will not trust the intentions of the truth-seeking community in the first place, and so will withhold their support from the community or even attack it. This may also make it hard to attract new contributors to the community.</li>
<li>Internal fragmentation through different opinions: The question whether the sensitivity of the model to the controversial belief is high enough to warrant any attention may be a narrow one, one that is not stated and analyzed very explicitly, or one that is analyzed explicitly but through models that make contradictory predictions. In such a case it seems very likely that people will arrive at very different predictions as to whether it’s worse to ignore the belief or to risk the previous failure modes. This can lead to fragmentation, which often leads to the demise of a community.</li>
<li>Internal fragmentation through lack of trust: The same internal fragmentation can also be the result of decreasing trust within the community because the community is being exploited or may be exploited by bad actors along the lines of failure mode 1.</li>
<li>Collapse of the community due to stalled recruiting: This applies when the controversial belief is treated as a serious infohazard. It’s very hard to recruit people for research without being able to tell them what research you would like them to do. This can make recruiting very or even prohibitively expensive. Meanwhile there is usually some outflow of people from any community, so if the recruitment is too slow or fully stalled, the community may eventually vanish. This would be a huge waste especially if the bulk of the research is perfectly uncontroversial.</li>
</ol>
<p>I have only very tentative ideas of how these risks can be alleviated:</p>
<ol>
<li>The community will need to conduct an appraisal, as comprehensive and unbiased as possible, of all the expected costs/harms that come with engaging with controversial beliefs.</li>
<li>It will need to conduct an appraisal of the sensitivity of its models to the controversial beliefs and what costs/harms can be averted, say, through more precise prioritization, if the truth value of the beliefs is better known.</li>
<li>Usually, I think, any specific controversial belief will likely be close to irrelevant for a model so that it can be safely ignored. But when this is not the case, further safeguards can be installed:</li>
<li>Engagement with the belief can be treated as an infohazard, so those who research it don’t do so publicly, and new people are onboarded to the research only after they’ve won the trust of the existing researchers.</li>
<li>External communication may take the structure of a hierarchy of tests, at least in particularly hazardous cases. The researchers need to gauge the trustworthiness of a new recruit with questions that, if they backfire, afford plausible deniability and can’t do much harm. Then they only gradually increase the concreteness of the questions if they learn that the recruit is well-intentioned and sufficiently open-minded. But this can be uncooperative if some codes become known, and then people who don’t know them use them inadvertently.</li>
<li>If the risks are mild, there may be some external communication. In it, frequent explicit acknowledgements of the risks and reassurances of the intentions of the researchers can be used to cushion the message. But these signals are cheap, so they don’t help if the risks are grave or others are already exploiting these cheap signals.</li>
<li>Internal communication needs to frequently reinforce the intentions of the participants, especially if there are some among them who haven’t known the others for a long time, to dispel worries that some of them may practice other than prosocial, truth-seeking intentions.</li>
<li>Agreed-upon procedures such as voting may avert some risk of internal fragmentation.</li>
</ol>
<p>An example that comes to mind is a situation when a friend of mine complained about the lacking internal organization of certain unpolitical (or maybe left-wing) groups and contrasted it with a political party that was very well organized internally. It was an, in our circles, highly disreputable right-wing party. His statement was purely about the quality of the internal organization of the party, but I only knew that because I knew him. Strangers at that meetup might’ve increased their credence that he agrees with the policies of that party. Cushioning such a mildly hazardous statement would’ve gone a long way to reduce that risk and keep the discussion focused on value-neutral organizational practices.</p>
<p>Another disreputable opinion is that of Dean Radin who seems to be fairly confident that there is extrasensory perception, in particular (I think) presentiment on the timescale of 3–5 s. He is part of a community that, from my cursory engagement with it, seems to not only assign a nonzero probability to these effects and study them for expected value reasons but seems to actually be substantially certain. This entails an air of disreputability either because of the belief by itself or the particular confidence in it. If someone were to create a model to predict how likely it is that we’re in a simulation, specifically in a stored world history, they may wonder whether cross-temporal fuzziness like this presentiment may be signs of <a href="https://en.wikipedia.org/wiki/Motion_compensation">motion compensation</a>, a technique used in video compression, which may also serve to lossily compress world histories. This sounds wild because we’re dealing with unlikely possibilities, but the simulation hypothesis, if true, may have <a href="https://longtermrisk.org/how-the-simulation-argument-dampens-future-fanaticism">vast effects on the distribution of impacts from interventions in the longterm</a>. These effects may plausibly even magnify small probabilities to a point where they become relevant. Most likely, though, they stem from whatever diverse causes are behind the <a href="https://slatestarcodex.com/2014/04/28/the-control-group-is-out-of-control/">experimenter effect</a>.<sup id="fnref2:scott-alexander-disclaimer"><a class="footnote-ref" href="#fn:scott-alexander-disclaimer">1</a></sup></p>
<p>I imagine that history can also be a guide here as these problems are not new. I don’t know much about religion or history, so I may be mangling the facts, but Wikipedia tells me that the <a href="https://en.wikipedia.org/wiki/First_Council_of_Nicaea">First Council of Nicaea</a> in 325 <span class="caps">CE</span> addressed the question of whether God created Jesus from nothing (Arianism) or whether Jesus was “begotten of God,” so that there was no time when there was no Jesus because he was part of God. It culminated as follows:</p>
<blockquote>
<p>The Emperor carried out his earlier statement: everybody who refused to endorse the Creed would be exiled. Arius, Theonas, and Secundus refused to adhere to the creed, and were thus exiled to Illyria, in addition to being excommunicated. The works of Arius were ordered to be confiscated and consigned to the flames, while his supporters were considered as “enemies of Christianity.” Nevertheless, the controversy continued in various parts of the empire.</p>
</blockquote>
<p>This also seems like a time when, at least in most parts of the empire, a truth-seeking bible scholar would’ve been well advised to consider whether the question has sufficiently vast implication as to be worth the reputational damage and threat of exile that came with engaging with it open-mindedly. But maybe there were monasteries where everyone shared a sufficiently strong bond of trust into one another’s intentions that some people had the leeway to engage with such questions.</p>
<div class="footnote">
<hr>
<ol>
<li id="fn:scott-alexander-disclaimer">
<p>Disclaimer: I don’t generally endorse the works of the author. Alexander originated a wealth of helpful ideas, so that I can’t help but cite him lest it seem that I plagiarize them. Unfortunately, (1) the community around his blog contains some insalubrious factions, and (2) until roughly 2016, he himself still published articles that presented issues in a skewed fashion reminiscent of the very dynamics he warns of in <a href="https://slatestarcodex.com/2014/12/17/the-toxoplasma-of-rage/">Toxoplasma of Rage</a>. I’m adding these disclaimers to avoid the impression that I accept such intellectual wantonness or that it is accepted in my circles. I don’t know whether he still endorses his old approaches. <a class="footnote-backref" href="#fnref:scott-alexander-disclaimer" title="Jump back to footnote 1 in the text">↩</a><a class="footnote-backref" href="#fnref2:scott-alexander-disclaimer" title="Jump back to footnote 1 in the text">↩</a><a class="footnote-backref" href="#fnref3:scott-alexander-disclaimer" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
<li id="fn:1">
<p>How this should be operationalized is still fairly unclear. Ozzie plans to work out more precisely what it is we seek to accomplish. You might as well call it “collaborative truth-seeking,” “improving collective epistemics,” “collaborating on better predictions,” etc. <a class="footnote-backref" href="#fnref:1" title="Jump back to footnote 2 in the text">↩</a></p>
</li>
<li id="fn:2">
<p>I’ve considered a wide range of contributions I could make. Given my particular background, this currently seems to me like my top option. <a class="footnote-backref" href="#fnref:2" title="Jump back to footnote 3 in the text">↩</a></p>
</li>
</ol>
</div>Donations 20202020-12-07T00:00:00+00:002020-12-07T00:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2020-12-07:/donations-2020.html<p>I’ve put approximately $7,000 into the $500,000 <a href="https://app.effectivealtruism.org/lotteries/">donor lottery</a> and donated some odds and ends to the Center on Long-Term Risk.</p><p>Two things happened in 2020: I switched out of earning to give and I looked into <a href="https://longtermrisk.org/msr">evidential cooperation in large worlds</a>. Oh yeah, and a pandemic, various conflagrations, a market crash, Brexit, SpaceX starting the exodus, election fraud in Belarus, protests in Belarus, protests in the <span class="caps">US</span>, protests in Bulgaria, protests in Kyrgyzstan, riots in India, explosions in Beirut, water on the moon, earthquake, hurricane, volcano, terrorist attacks, deployment of murder hornets, black hole, free public transport, Félicien Kabuga arrested, Ladislas Ntaganzwa sentenced, the pope getting into <span class="caps">AI</span> safety, Africa free of polio, Biden-Harris, negligible election fraud in the <span class="caps">US</span>, Ethereum 2.0, first-ascent of the second 9c route, birthday of one of my partners. But those things didn’t influence my donations.</p>
<h2 id="donor-lottery"><a class="toclink" href="#donor-lottery">Donor Lottery</a></h2>
<p>I want to donate less because of my now lower income, and I want to donate to the <a href="https://app.effectivealtruism.org/lotteries/">donor lottery</a> so that 1.4% of my copies on various Everett branches can dedicate a lot of time to allocating a really big chunk of money instead of all of me 80/20ing it again.</p>
<p>The donor lottery <span class="caps">FAQ</span> says, “The winning number for each lottery block will be determined by taking the first ten hexadecimal digits of the <a href="https://www.nist.gov/news-events/news/2018/04/nists-new-quantum-method-generates-really-random-numbers"><span class="caps">NIST</span> Beacon</a> at the lottery draw date.” (Link mine.) So if something like the many-worlds interpretation is true, my plan should go through. If not, I fall back on good-old expected value, and the lottery seems equally attractive.</p>
<h2 id="center-on-long-term-risk"><a class="toclink" href="#center-on-long-term-risk">Center on Long-Term Risk</a></h2>
<p>The Center on Long-Term Risk (<span class="caps">CLR</span>, previously known as Foundational Research Institute), has been my top favorite organization for several years. Because of the diminutive size of this donation, I haven’t re-evaluated that opinion this year.</p>
<p>The reasoning is that, first, all the more mundane things check out, such as that I know that there are smart, cooperative, and committed people working there. But second, <span class="caps">CLR</span> is also the only place that focuses on two of the most important questions I’m aware of: how to avert suffering risks and what the optimal moral compromise looks like.</p>
<p>Extinction risks already receive a fair bit of attention in effective altruism circles. Fates much worse than extinction have received less attention even though averting those is also in the interest of a wide range of value systems, and addressing them may be no less urgent. </p>
<p>The optimal moral compromise, which follows from <a href="https://longtermrisk.org/msr">evidential cooperation in large worlds</a>, is something that <span class="caps">CLR</span> prioritizes only nominally at the moment. As far as I know, no one on staff currently focuses on this topic.</p>
<p>Justin Shovelain (<a href="https://www.convergenceanalysis.org">Convergence Analysis</a>) told me that figuring it out exactly can perhaps wait a bit longer until we have extinction risks a bit more under control.<sup id="fnref:1"><a class="footnote-ref" href="#fn:1">1</a></sup> I would agree that averting extinction risks is very likely to be cooperative according to the compromise. The question whether we’re sufficiently sure that it is in the interest of the compromise is itself an optimal stopping problem that we may want to prioritize, but for now I’m leaning toward yes.</p>
<p>There is the view that the time that we should delegate such investigations to is that of the Long Reflection, after we’ve attained existential security. The Long Reflection may take centuries or millennia.</p>
<p>If I look at the world today, coordination looks very difficult. The Long Reflection would require reducing the risk of irreversible unilateral action from anyone anywhere in earth’s vicinity to a negligible level, and not just a negligible level per year but over the course of a whole millennium. That seems hard, to say the least. Conversely, if the world is very different in that, for example, one singleton is in charge of it, then coordination is easy, but getting such a singleton to care about our values, or the moral compromise in particular, is a notoriously hard problem. I may be overlooking something, but such a Long Reflection seems to me about impossible (<< 0.1%) in the first and very unlikely (< 1%) in the second case.</p>
<p>But I’m still a big fan of all other aspects of the Long Reflection, so I think it’ll more likely take the shape of a race between priorities research and technological progress. In between there’ll be some advocacy to try to influence the (hopefully conscientious) agents driving the technological progress. That race may have to start soon because technological progress won’t wait for us at the starting line.<sup id="fnref:2"><a class="footnote-ref" href="#fn:2">2</a></sup></p>
<p>So in short, I think, we need something like the long reflection soon and with it work to narrow down the optimal moral compromise, but I can also understand that it’s not <span class="caps">CLR</span>’s top priority for now.</p>
<p>In the branches where I win the lottery I may decide to incentivize work on this by people who wouldn’t otherwise work on something even more pressing.</p>
<p>Rethink Priorities also seems well-positioned to launch investigations into such questions. Rethink Priorities and the Wild Animal Initiative have long been close contenders for my personal “favorite organization” spot.</p>
<h2 id="notes"><a class="toclink" href="#notes">Notes</a></h2>
<div class="footnote">
<hr>
<ol>
<li id="fn:1">
<p>It was a verbal conversation, so I can’t check the exact wording. <a class="footnote-backref" href="#fnref:1" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
<li id="fn:2">
<p>Michael Aird remarked that such a race-like “Long Reflection” lacks distinctive features of the Long Reflection and should receive a new name. I have yet to come up with one. <a class="footnote-backref" href="#fnref:2" title="Jump back to footnote 2 in the text">↩</a></p>
</li>
</ol>
</div>Harnessing Cognitive Dissonance2020-11-15T00:00:00+00:002020-11-15T00:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2020-11-15:/harnessing-cognitive-dissonance.html<p>My feeling of motivation seems to stem, in some cases, from cognitive dissonance. I consider how I may intentionally generate cognitive dissonance for the things I want to be motivated for.</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#cognitive-dissonance">Cognitive Dissonance</a></li>
<li><a href="#flow">Flow</a></li>
<li><a href="#tips">Tips</a></li>
</ul>
</div>
<h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<p>I’ve been trying to understand what motivates me. Sometimes I literally fall in love with some area of learning. I think about little else for months or a year and try to spend every waking minute practicing or learning the thing. In particular, I try to eke out even just minutes at a time wherever possible, and I don’t notice feeling tired (even when objectively my performance declines from the lack of sleep). At other times I at least have no trouble concentrating on a task for ten hours almost nonstop even if it doesn’t feel particularly thrilling. Finally, there are many tasks that I have trouble focusing on for even an hour.</p>
<p>This happens involuntarily and is imperfectly aligned with what I consider to be most important – I might call these things System 1 and System 2 motivation. So I want to understand better how to create the sorts of conditions that allow me to be more System 1–motivated.</p>
<h2 id="cognitive-dissonance"><a class="toclink" href="#cognitive-dissonance">Cognitive Dissonance</a></h2>
<p>One factor that I think I’ve identified and that seems somewhat universal (so is not only relevant for me) is cognitive dissonance. When I feel strong cognitive dissonance, it motivates me to investigate the topic. This takes different forms:</p>
<ol>
<li><em>Moral dilemmas.</em> These are morally controversial topics. Sometimes the moral dilemmas aren’t, but so long as that hasn’t been established, I’d like to group such pseudo-dilemmas under this rubric too. The frequent use of moral dilemmas in fiction and the popularity of trolley problem memes indicates that I’m not alone with this. And it makes societal sense to work hard to resolve these things.</li>
<li><em>Taboos.</em> These are “socially controversial” for reasons of, say, signaling group membership or for fear of overstepping one’s social status. But it’s hard to find something morally bad with engaging with them, or whatever morally bad things they tend to come with are avoidable (other than opportunity costs).</li>
<li><em>Empirically controversial topics.</em> These are topics that people have widely different empirical beliefs about. (But only to a degree where they – or I at least – can’t outright dismiss the other view.)</li>
<li><em>Imperfectly understood topics.</em> These are topics where the experimental data contradicts the predictions of all known theories.</li>
<li><em>Unaesthetic resolutions.</em> These may be problems whose solutions strike me as inelegant, partial, ad-hoc, or disproportionate.</li>
</ol>
<h2 id="flow"><a class="toclink" href="#flow">Flow</a></h2>
<p>In addition to the cognitive dissonance, flow is important too. I think flow becomes possible when the cognitive dissonance is only on one (or very few) abstraction layers at a time.</p>
<p>It can be flowy to learn a new programming language with different paradigms and conventions from the ones I know while writing simple examples in the language. But learning such a very different language while working on a complex software written in it whose workings I don’t understand either may already feel less flowy. It can get downright frustrating when the complex software was written by many people with very different levels of seniority, so that I need to constantly evaluate what might make sense within the paradigm I don’t understand yet and what I should refactor thoroughly.</p>
<p>The second one may be a quirk of mine. I’ve found that some smart friends of mine seemed to have learned a lot from unreliable teachers without adopting the teacher’s mistakes. I also have this quirk where I enjoy things more the more they have the shape where you can <a href="/modelers-and-indexers.html">immediately forget all examples</a> because you can rederive them any time. These may or may not apply to you.</p>
<h2 id="tips"><a class="toclink" href="#tips">Tips</a></h2>
<p>I haven’t tested these yet. They are mostly habits or tricks that I’ve observed but never tried because I didn’t understand them. Now I think I understand them better. I’ll update the list once I have first-hand experience with them.</p>
<ol>
<li><em>Thinking first, reading second.</em> Christian Tarsney told me once at a conference that Hilary Greaves has this tip for researchers that they shouldn’t start learning about a topic by reading all the literature on it but that they should start by trying to figure it all out by themselves. (This is some second-level telephone game, so it’s well possible she said something else entirely. Sorry if that’s the case!) That seemed interestingly counterintuitive to me. One benefit that came to mind is that maybe all established researchers are caught in some local optimum of a theory that they fleshed out in such detail that it now works better for them than anything else. So a different theory, fleshed out even to a tenth of that level of detail, would perform better. But from the vantage point of the local optimum that theory looks as inferior as all the other actually inferior theories. But there’s a second benefit: if you form an arbitrary detailed mental model from the start and only then let it clash with reality, the cognitive dissonance may be motivating.</li>
<li><em>Tests</em>. There is no way to generate cognitive dissonance if you can’t test your theory against reality. Extroverts will enjoy conversation as a universal test framework. Introverts can converse with friends, mentally simulate dialogs, monologue toward real or imagined readers, learn formalisms that expose internal inconsistencies, analyze freely available data, or write software (if it doesn’t work, you did it wrong). Please comment if you know more testing methods.</li>
<li><em>Opinionatedness.</em> I’ve known a few smart people who are very opinionated, not in the sense that they fail to update away from wrong opinions because their priors are too strong but in the sense that whatever new thing you throw at them, it takes split seconds before they have a clear opinion on it (which they are happy to dismiss a minute later). My experience, on the other hand, is usually that for me tons of arguments in all directions come to mind, and I need to spend a lot of time assigning the right weights to them and grappling with interdependencies before any fledgling feeling of an opinion emerges. Having a random opinion from the start may generate strong cognitive dissonance (because it’s probably not fully correct), and that may keep you motivated to hone or revise your opinion while I may lose interest halfway through assigning all those weights. For bonus points, you can make it as empirically controversial as possible. (Morally or socially controversial random theories probably come with too costly externalities to be worth it.)</li>
<li><em>Practice.</em> Practice can automate a lot of low-level motions so that one’s focus can rest on just that one abstraction layer where the cognitive dissonance is strongest. That enables flow.</li>
<li><em>Experience.</em> Experience leads to streams of ideas rather than trickles. If I have to wait for new ideas for days or weeks, there’s no flow, but if I have a fresh idea every thirty seconds even before my latest one has even failed, or I have a whole backlog of ideas to try, there’s a lot more flow.</li>
<li><em>Confirmation bias.</em> Maybe confirmation bias is motivating at first because it leads to a random clear model. Then, of course, the confirmation bias needs to be turned off again. Not sure who can do that.</li>
</ol>Levels of Moral Cooperation2020-11-02T00:00:00+00:002020-11-02T00:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2020-11-02:/levels-of-moral-cooperation.html<p>What level of cooperation with other activists or general agenty people is optimal? Everyone needs to “pick their battles” to maximize their impact and stay sane, but what does it look like in practice when you interact with those who’ve picked different “battles”? In this article I sketch some initial ideas I’ve had and hope to encourage others to test and refine them. There’s a lot of potential for someone with more knowledge of game theory to make them greatly more rigorous, nuanced, and reliable.</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#levels">Levels</a></li>
<li><a href="#problems">Problems</a><ul>
<li><a href="#failing-to-notice-trade-opportunities">Failing to Notice Trade Opportunities</a></li>
<li><a href="#many-currencies">Many Currencies</a></li>
<li><a href="#neutrality-and-nuance">Neutrality and Nuance</a></li>
<li><a href="#high-dimensional-beating">High-Dimensional Beating</a></li>
<li><a href="#problems-with-aggregation">Problems with Aggregation</a></li>
</ul>
</li>
<li><a href="#importance">Importance</a></li>
</ul>
</div>
<h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<p>We have a variety of views when it comes to axiology – how good various world histories are relative to one another, whether such assessments are <a href="https://forum.effectivealtruism.org/posts/TwJb75GtbD4LvGiku/moral-anti-realism-sequence-1-what-is-moral-realism">subjective or god- or nature-given</a>, and <a href="https://longtermrisk.org/files/Multiverse-wide-Cooperation-via-Correlated-Decision-Making.pdf">how decision-relevant that is in practice</a> (section 3.1). But there’s also the related question of what individual behaviors or social norms entail the best world histories given some axiology. This article focuses on the second and ignores most of the complexity of the first.</p>
<p>Paul Christiano has argued that <a href="https://forum.effectivealtruism.org/posts/CfcvPBY9hdsenMHCr/integrity-for-consequentialists">integrity is crucial</a>. Scott Behmer warned that <a href="https://forum.effectivealtruism.org/posts/54Cdt4BR84vDcki6i/effective-altruism-and-free-riding">the <span class="caps">EA</span> community tends to free-ride</a>, and that he doesn’t know when that is really appropriate. (I’ve <a href="https://impartial-priorities.org/thoughts-on-free-riding.html">commented</a> on the second post.) </p>
<p>Once you’ve settled on some axiology, calculating and comparing the expected choiceworthiness of all actions may not be feasible. So what considerations should guide your actions? I’ve come across a number of such considerations in isolation. In the following, I want to sketch a framework that I’ve found helpful to consolidate many of them.</p>
<p>It doesn’t answer whether a particular cooperation or moral trade is worth it, but it has helped me structure my thinking about this question. If two people both accept a framework like this one, they can also communicate much better about their needs and what they want to offer in return.</p>
<p>Here a few questions that it has helped me understand better:</p>
<ol>
<li><strong>Intimacy of cooperation</strong>: Do you collaborate with a highly value-aligned cofounder on a charity startup? Or do you cooperate with a somewhat aligned group on some common instrumental goal? Or do you trade with a mostly unaligned group on an issue where some of your opposing preferences are more intense than others, allowing for gains from trade? Or do you clash with an opponent in a verbal debate in which you both refrain from insults, blackmail, and violence? I address this in the Levels section.</li>
<li><strong>Level of escalation</strong>: Do you and an opponent clash in a debate in which you maintain high epistemic standards – you refrain from lies and misrepresenting your opponent’s points, and you concede points that you think they’re right about? Do you and your opponent raise your voices but refrain from blackmailing each other with private information you have? Do you and your opponent scream incomprehensibly at each other but refrain from physical violence? Do you and your opponent beat each other up but leave your guns in the holsters because of a shared understanding that it’s a fist fight? I address this in the Levels section.</li>
<li><strong>Reciprocation of your cooperation</strong>: You may be happy to pay it forward a bit, but you don’t want to be freerode on for long. You also don’t want to end up in mutual defection just because of noise in your communication. I address this in the Levels section.</li>
<li><strong>Reasons for imperfect alignment</strong>: Are you factually wrong about something, and/or are they, and can either/both of you be convinced of the other’s or another position? Or are their terminal moral preferences different from yours? Or is it potentially the first, but finding out would be more costly than compromising? I address this in the Problems section.</li>
<li><strong>W_anting_ to help beings who can’t or are unlikely to reciprocate:</strong> Nonhuman animals or people who’ll be alive only when you’re dead are examples. You need to weigh your resource expenditure on this goal against your resource expenditure on gains from trade. I address this in the Problems section.</li>
<li><strong>What is the neutral level of cooperation</strong>: Is it neutral to be completely silent on a problem or is the problem so widely reviled (in your country, city, or among your peers) that silence is reactionary and the neutral stance is to express support in various customary, low-cost ways? I address this in the Problems section.</li>
</ol>
<p>Some of this will become clear when I introduce the levels of cooperation that I’ve come up with. The other questions are specific to particular situations; these I’ll address in the Problems section. Finally I’ll try to guess how important all of this might be for us.</p>
<h2 id="levels"><a class="toclink" href="#levels">Levels</a></h2>
<p>I’ll for now assume that, all else equal, closer cooperation is preferable to more reserved cooperation or conflict, be it because of reduced overhead, gains from trade, or less zero-sum mutual undoing of efforts. That’s perhaps more plausible if I also consider “cooperation partners” two people who don’t know of each other but work toward the same goal without duplicating efforts. (They might work on something that is not verifiable, so that it’s valuable to know if two people completely independently arrived at similar conclusions.)</p>
<p>Below some illustrative levels of cooperation. These are lines drawn somewhere in the middle of huge gray areas, so don’t take them too seriously. For example, some collaborators may be partially motivated by improving their CVs and so may be more like trade partners, and some people may be collaborators on one issue but mere nonaggressors on another.</p>
<p>This list starts with a level of maximal escalation and goes up to perfect cooperation:</p>
<ol>
<li><strong>Uncivil aggressors</strong>: People outside civilization in the state of nature, be it because they don’t care about anyone including their future person moments or because they are untouchably powerful. They flout all social norms of interaction. They might kill you if it seems auspicious to them.</li>
<li><strong>Ambivalent aggressors</strong>: People who compromise on some social norms. Say, they might lie or blackmail you, but they wouldn’t kill you even if they could get away with it. That’s just an example: Some may find lying more objectionable than killing. You may be able to predict which social norms they are more ready to flout than others.</li>
<li><strong>Civil aggressors</strong>: People who may attack you but in basically civil ways. Say, they may attack you through honest but pointed critiques in public forums. A rough guide may be that they’ll not resort to methods that are illegal in their jurisdictions.</li>
<li><strong>Nonaggressors</strong>: Everyone you trust not to attack you. You can’t trust them to live up to norms that are not legally required of them except that they will stay out of your way.</li>
<li><strong>Neutrals</strong>: Everyone who takes care to do just enough not to hinder your efforts. They may be quite committed to maintaining at least this neutral level. They actively seek out feedback on their behavior. They may not be interested in praise but they’ll change their behavior if they hear complaints.</li>
<li><strong>Trade partners</strong>: Everyone you trade with. Say, a musician you pay to create the theme song for your startup. If you pay them well, you’ll get a great theme song from them. This also includes opponents with whom you can agree on compromise solutions. (Failing that, you’ll probably fall straight to nonaggressor level with them rather than neutral.)</li>
<li><strong>Collaborators</strong>: Everyone you collaborate with because you share some goals, terminal or instrumental. Those might be people of another organization with an only subtly different vision. They don’t require you to pay them if they can afford it.</li>
<li><strong>Unified actors</strong>: You and everyone you’re highly value-aligned with. Maybe your cofounders of a charity startup, your partners, or whole subsets of the <span class="caps">EA</span> community.</li>
</ol>
<p>This progression doesn’t settle how we should resolve the tension between limiting freeriding and paying it forward, or between limiting freeriding and maintaining the moral high ground, or between limiting freeriding and improving social norms. But it can lend some structure and language to discussions of the topic. The following is an example.</p>
<p>At first approximation, it seems that you should engage with someone on the level that they chose for the interaction. If you start the interaction, then you can engage them on the level that is your best guess of the level they would choose.</p>
<p>But I think that’s risky. Communication is noisy. Communication that hasn’t happened yet is particularly noisy. Lifting an interaction with someone to a higher level takes a lot of work. Losing them or getting caught in defect-defect cycles is much easier. The government and our friends and peers give us powers and incentives that make it cheaper and less risky to maintain a moral high ground.</p>
<p>All in all, the costs, risks, and opportunities are sufficiently asymmetrical that it’s usually warranted to err on the side of starting an interaction on too high of a level, say, by one level. (Unless you really can’t afford it.)</p>
<h2 id="problems"><a class="toclink" href="#problems">Problems</a></h2>
<p>Here is how I would frame various problems that I’ve observed.</p>
<h3 id="failing-to-notice-trade-opportunities"><a class="toclink" href="#failing-to-notice-trade-opportunities">Failing to Notice Trade Opportunities</a></h3>
<p>One failure mode is to ignore someone – so behave neutral or even just as nonaggressor toward them – who might be a trade partner or collaborator if only they (or you) realized some minor fact. The information exchange can happen efficiently and is well worth the time investment. It may be particularly worth it for you if you’re more wrong than the other party. Conversely, one might overinvest into trying to resolve differences at the expense of putting more work into finding people or groups who would immediately readily enter into close collaborations or trades.</p>
<p>Examples:</p>
<ol>
<li>I’ve often talked with people who seemed to be opposed to effective altruism for various reasons that were easy to resolve. One person I remember thought that the likes of GiveWell were oblivious to the research of J-<span class="caps">PAL</span> and were trying to reinvent the development-economic wheel. I could easily convince her that that’s not the case.</li>
<li><a href="https://sentientmedia.org/racism-in-animal-advocacy-and-effective-altruism-hinders-our-mission/">Michelle Graham argues</a> that the wild animal welfare community missed out on important contributions from conservationists and that the <span class="caps">EA</span> and animal advocacy communities still miss out on important contributions from people of the global majority (people of color). Perhaps that can also be remedied.</li>
<li>But there’s also often an effect that has been termed “narcissism of small differences.” Very different groups may never talk to each other, but groups that are only very slightly different may take those remaining differences as occasion for hostility or (more relevantly in this context) year-long, tedious debates. There is some point where it’s worth discussing something, but if that fails for too long of a time, it may be cheaper to compromise even if the issue should hypothetically lend itself to an empirical resolution. (The discussion can continue when it’s not blocking the collaboration anymore.)</li>
</ol>
<h3 id="many-currencies"><a class="toclink" href="#many-currencies">Many Currencies</a></h3>
<p>Second, when it comes to the bargaining, there are many currencies involved, and the parties may disagree on how to convert them into each other. Nonhuman animal trade partners or future generations may remunerate us with increased aggregate well-being, which we value. Present human trade partners may contribute back with their skills or contacts or money. Those are already four currencies whose exchange rates potential trade partners can have greatly different opinions on. That can make causal bargaining difficult and gameable and can make acausal bargaining quite unclear. In any case, it’s perhaps helpful to think of all these benefits as having conversion rates that you just need to come to agree on.</p>
<p>Example:</p>
<ol>
<li>One activist may argue that funding should go to the wild animal initiative because the number of wild animals is so enormous and our potential to help them so largely untapped that research into ways to do that improves enormous numbers of animals’ person moments. The other activist may argue that beings who don’t contribute to human society should not be considered moral patients, so that it’s improper to use such beings to inflate the importance of an alleged problem. To resolve this conflict, one would have to bet on moral realism and solve some key issues of ethics in ways that convince both. That may be too hard, so they’ll have to settle on some (perhaps unprincipled) compromise to convert the currency of improved wild animals’ person moments to a currency the other activist values.</li>
</ol>
<h3 id="neutrality-and-nuance"><a class="toclink" href="#neutrality-and-nuance">Neutrality and Nuance</a></h3>
<p>Third, there is neutrality and nuance: Nonaggression is a very fundamental tenet in many parts of the world. Various countries and in particular various small communities and friend groups will have much more refined and varied norms, and many of them will serve good purposes. Even just being neutral will then mean more than just abiding by the law. If you don’t want to hinder the efforts of those in your community, you will need to adhere to a higher standard.</p>
<p>Examples:</p>
<ol>
<li>In much of the world the neutral stance toward animal rights may be to not insult vegans or activists. In some parts of the <span class="caps">EA</span> community, however, the neutral stance may be to be at least a lacto-vegetarian nonactivist.</li>
<li>In some workplaces it’s neutral to refrain from shouting others down when they try to speak. But in others it’s neutral to be respectful of what they have to say and to interrupt someone only in exceptional circumstances.</li>
</ol>
<h3 id="high-dimensional-beating"><a class="toclink" href="#high-dimensional-beating">High-Dimensional Beating</a></h3>
<p>Fourth, it’s hard to recognize multidimensional beating without becoming overly cautious about collaboration. <a href="https://en.wikipedia.org/wiki/Tacking_(sailing)#Beating">Beating</a> is “the procedure by which a ship moves on a zig-zag course to make progress directly into the wind (upwind).” Someone may have instrumental goals opposite to your instrumental goals, but you have resources that the other person desires. So they may pretend to compromise with you to get some of your resources in return. But really they keep shifting (or rotating – imagine a spiral) what they compromise on. At any particular moment they may seem like a trade partner, but over a sufficiently long time it becomes clearer that they are using your resources to work against you.</p>
<p>This can be less obvious than the illustration in the abovelinked Wikipedia article because the tacking points can be smoothed out. It’s also less obvious because social interaction is very high-dimensional. So a slight, gradual reversal on one agreed-upon compromise can be hidden among or cushioned by a strengthening of a number of other compromises. This can arise seemingly naturally if you press the trade partner to adhere to some particular parts of your agreement. They may strengthen their momentary commitment to those parts while reversing their commitment to parts that are not currently in the focus.</p>
<p>Examples:</p>
<ol>
<li>Abusive relationships can work this way.</li>
<li>Some companies with great environmental and welfare externalities (e.g., sellers of animal products) may be trying this when they sign various pledges when there’s a lot of pressure but years later fail to follow through on them when the pressure is on some other issue.</li>
</ol>
<h3 id="problems-with-aggregation"><a class="toclink" href="#problems-with-aggregation">Problems with Aggregation</a></h3>
<p>Fifth and finally, there are sometimes problems with aggregation. For example, when whole organizations want to trade or cooperate with other organizations – or even bigger and looser groups want to do the same – then there may be tensions between what the whole group as a group wants or does and what individual people in the group do. Conversely, when two people trade or cooperate, but one of them is part of a bigger group. The group may violate agreements even when the individual doesn’t. In both cases there are additional costs to expelling a group member or to leaving a group, which may or may not be more important to the trade partner than the trade. This can also be exploited on purpose.</p>
<p>Examples:</p>
<ol>
<li>Whenever I hear someone complain that people with label X believe A and B where A and B contradict each other, I wonder whether really the fraction of Xist who believe A and B is tiny, and really the problem is that Xist are divided over whether A or B. (The division may not be one that they are wont to fight over, so that they hardly notice.)</li>
<li>Conversely, this effect can be exploited by companies that agree with other companies or a government to follow certain guidelines but then implement incentives internally that subtly push against those policies. They can claim that they just haven’t quite figured out the best incentives yet, and there’s never any proof that they set the present incentives on purpose to violate the policy even if recordings of all the relevant meetings should leak.</li>
</ol>
<h2 id="importance"><a class="toclink" href="#importance">Importance</a></h2>
<p>I’m quite unsure how important this is generally. It probably varies a lot and is more important on smaller scales than on greater ones – e.g., it may be important for <span class="caps">EA</span> startups, but a country may only benefit mildly from improving cooperation nationally. Three heuristics that come to mind:</p>
<ol>
<li>A powerful actor (I use “power” and “resources” interchangeably here) whose marginal utility doesn’t diminish quickly may rather want to just do their thing with no help (but also no interference) from others. (Like some fictional superheroes – Batman, Superman, Iron Man.)</li>
<li>Someone who doesn’t have enough power will benefit greatly from trading with a similarly or more powerful partner. (In Germany I saw a lot of animal rights organizations cooperate like that.)</li>
<li>Someone powerful with diminishing marginal utility may get to the point where they are forced to trade with others to make progress. (<a href="https://longtermrisk.org/msr">Evidential cooperation in large worlds</a> may be particularly interesting for them. Thanks to Daniel Kokotajlo for this idea! Maybe tech companies count that can’t find enough developers to hire so that they have to “acqhire” other companies.)</li>
</ol>
<p>Someone may also be powerful in one way (e.g., have a lot of money) but weak in another (e.g., can’t recognize the first excellent developer that can build up a team of excellent developers), which again makes trade more valuable for them.</p>
<p>On a large scale, improved cooperation may not make a big difference. The advent of the internet has probably made a lot of cooperation on a national and international scale a lot easier, and yet <a href="https://ourworldindata.org/economic-growth">global <span class="caps">GDP</span></a> has increased by barely one third since then, even imagining that the 2008 crash didn’t happen. And that’s the product of so much more than just the internet.</p>
<p>But maybe a 5- or 10-fold increase is realistic in more limited contexts that either suffer from low levels of cooperation or have high upsides to cooperation. In fact, I’ve increased the fundraising revenue of an organization 20-fold at about constant cost by adjusting completely to a particular audience, and I might’ve increased it 200-fold or more if I had known then what I know now. That seems similar to a shift from a neutral role to a trade partner or cooperator role for that audience.</p>
<p>Beyond that I suspect that there are risks from failing to keep up (keep on an at least neutral level) with positive developments in one’s social context that can cause hostility rather than just lost gains from trade. The negative effects of such hostility on small or less powerful groups may rival (in absolute terms) the positive effects above. Just as positive effects can have windows that open and close, negative effects can permanently destroy opportunities. So the <span class="caps">EA</span> community may fail to seize an opportunity to forge an important allegiance before such a window closes but it may also be destroyed if it fails to compromise before a powerful opposition forms. The original emergence of the community may have involved a lot of luck. Our new experience and any remaining networks may help to make up for some of the luck that will be lacking the second time around, but such a collapse seems very costly even if it is eventually reversible.</p>Covid-19 Costs2020-10-25T00:00:00+00:002020-10-30T00:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2020-10-25:/covid-19-costs.html<p>A Guesstimate to convert <a href="https://www.microcovid.org/">microcovids</a> to dollars. Microcovid.org is the best calculator of Covid-19 risk that I’m aware of, but the microcovids that it returns are unintuitive for me. <a href="https://docs.google.com/spreadsheets/d/1LBZWHEk2Jo-IFvZK_smrwYoOTOykB7H-oHb0qYjg2ys/edit">Peter Hurford’s spreadsheet</a> converts risks to dollars, which are very intuitive for me. My Guesstimate is a probabilistic version of the lower part of that spreadsheet but also includes risk from long-term damage. The Guesstimate allows you to plug in the microcovids range and get a dollar result. The article is password protected because I’m afraid I made mistakes that might kill people (not because it’s old, yet).</p><div id="pec-encrypted-content" style="display:none">Up3izUFdjtoculAsBYq/nQ==;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;^</div>
<div id="pec-decrypted-content">
<h4><i></i></h4>
</div>
<form id="pec-decrypt-form">
<p>
Old articles can be embarrassing. If you would like to access them anyway, you may
<a href="https://bit.ly/3jPN8tT" target="_blank">request access here</a>. Please indicate
who you are in case I don’t know you or don’t recognize you from your email address.
</p>
<input type="password" id="pec-content-password" placeholder="Password" />
<button type="submit" id="pec-decrypt-content">Decrypt</button>
</form>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/core.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/enc-base64.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/cipher-core.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/pad-nopadding.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/md5.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/aes.js"></script>
<script type="text/javascript">
(function () {
var strip_padding = function (padded_content, padding_char) {
/*
* Strips the padding character from decrypted content.
*/
for (var i = padded_content.length; i > 0; i--) {
if (padded_content[i - 1] !== padding_char) {
return padded_content.slice(0, i);
}
}
};
var decrypt_content = function (password, iv_b64, ciphertext_b64, padding_char) {
/*
* Decrypts the content from the ciphertext bundle.
*/
var key = CryptoJS.MD5(password),
iv = CryptoJS.enc.Base64.parse(iv_b64),
ciphertext = CryptoJS.enc.Base64.parse(ciphertext_b64),
bundle = {
key: key,
iv: iv,
ciphertext: ciphertext
};
var plaintext = CryptoJS.AES.decrypt(bundle, key, { iv: iv, padding: CryptoJS.pad.NoPadding });
try {
return strip_padding(plaintext.toString(CryptoJS.enc.Utf8), padding_char);
}
catch (err) {
// encoding failed; wrong password
console.log(err);
return false;
}
};
var init_decryptor = function () {
var decrypt_btn = document.getElementById('pec-decrypt-content'),
password_input = document.getElementById('pec-content-password'),
encrypted_content = document.getElementById('pec-encrypted-content'),
decrypted_content = document.getElementById('pec-decrypted-content'),
decrypt_form = document.getElementById('pec-decrypt-form');
decrypt_btn.addEventListener('click', function () {
// grab the ciphertext bundle
var parts = encrypted_content.innerHTML.split(';');
// decrypt it
var content = decrypt_content(
password_input.value,
parts[0],
parts[1],
parts[2]
);
if (content) {
// success; display the decrypted content
decrypted_content.innerHTML = content;
decrypt_form.parentNode.removeChild(decrypt_form);
encrypted_content.parentNode.removeChild(encrypted_content);
// any post processing on the decrypted content should be done here
}
else {
// ¯\_(ツ)_/¯
password_input.value = '';
}
});
};
document.addEventListener('DOMContentLoaded', init_decryptor);
})();
</script>Useful Things2020-10-23T00:00:00+00:002020-10-23T00:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2020-10-23:/useful-things.html<p>Some programs, platforms, apps, extensions, services, lifehacks, etc. that I’ve found useful.</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#general">General</a></li>
<li><a href="#linux-mac-os-and-maybe-other-operating-systems-too-but-who-knows">Linux, Mac OS, and Maybe Other Operating Systems Too But Who Knows</a></li>
<li><a href="#chrome-vivaldi-and-maybe-firefox-too">Chrome, Vivaldi, and Maybe Firefox Too</a></li>
<li><a href="#android-and-maybe-ios">Android and Maybe iOS</a></li>
<li><a href="#lifehacks">Lifehacks</a></li>
</ul>
</div>
<h2 id="general"><a class="toclink" href="#general">General</a></h2>
<ol>
<li><a href="https://www.backblaze.com">Backblaze</a>: The cheapest cloud backup solution I’ve found.</li>
<li><a href="https://onetimesecret.com/">One-Time Secret</a>: Sharing passwords is tricky. Even if you share them over an encrypted channel, they may end up unencrypted on a device of the receiver. One-Time Secret alleviates that risk a bit. (H/t Chris.)</li>
<li><a href="https://facebook.com/">Facebook</a>: It has nice people in it. I understand them better than on <span class="caps">VK</span>. I’m also one of those people who don’t dislike whatever the latest design is. WhatsApp is probably pretty secure now. And I really enjoy React and PyTorch!</li>
<li><a href="https://www.google.com">Google</a>: Great search engine, but I can also recommend some of their other tools, like Drive with all its document formats, Photos, Mail, Maps, Colab, etc. YouTube has a lot of cool climbing content. The calendar can be a bit wonky at times. Google Meet usually doesn’t let me in.</li>
<li><a href="https://forum.effectivealtruism.org/"><span class="caps">EA</span> Forum</a>: Lots of good thoughts on effective altruism. (H/t Alex P.)</li>
<li><a href="https://feedly.com/">Feedly</a>: Sort of like Google Reader, except alive.</li>
<li><a href="https://meet.jit.si">Jitsi</a>: Virtual meeting rooms that let you in!</li>
<li><a href="https://www.kraken.com">Kraken</a>: Good for trading cryptocurrencies. (Note that I don’t know what this “Buy Crypto” button does, but it gives me worse conversion rates than normal trades on Kraken. H/t Tanja.)</li>
<li><a href="https://jamstack.org/generators/">Static blog generators</a>: Those are good for websites that are are read much more often than they change. Having unchanging content, but regenerating the same pages every time they’re requested, and then putting a few layers of caching in front seems like the sort of things evolution would come up with. We can do better! ^.^</li>
<li><a href="https://support.google.com/a/users/answer/9308871?hl=en">Domains for creating Google Docs</a>: There are a number of domains that redirect such that they create new Google docs. E.g., you can open doc.new to create a new text document. (H/t Alfredo)</li>
</ol>
<h2 id="linux-mac-os-and-maybe-other-operating-systems-too-but-who-knows"><a class="toclink" href="#linux-mac-os-and-maybe-other-operating-systems-too-but-who-knows">Linux, Mac <span class="caps">OS</span>, and Maybe Other Operating Systems Too But Who Knows</a></h2>
<ol>
<li><a href="https://vivaldi.com">Vivaldi</a> (many OSs): A Chrome-based (or rather Chromium-based) browser with vertical tab bar! Browsers go from unusable to awesome once you add a vertical tab bar. Alternatively, you can turn your screen 90° and have all websites translated to a vertical script. Vivaldi also has a bunch of other features, but the vertical tab bar is what matters to me. Since it’s based on Chromium, you can use it with all the Chrome extensions. (Microsoft Edge, version 89.0.760.0-dev, now also has a vertical tab bar, <a href="edge://flags/#edge-vertical-tabs">which can be activated on the flags page</a>. Once activated you can toggle the vertical bar with a click on a button to the top left.)</li>
<li><a href="https://itunes.apple.com/us/app/magnet/id441258766?mt=12">Magnet</a> (Mac <span class="caps">OS</span>): A window manager for Mac <span class="caps">OS</span>. It’s like an 80/20 compromise between all the features of a tiling window manager (like XMonad) and compatibility with Mac <span class="caps">OS</span>. It saves me a lot of time, though I mostly just use it to maximize windows.</li>
<li><a href="https://github.com/CristianHenzel/ClipIt">ClipIt</a> (Linux), <a href="https://itunes.apple.com/us/app/flycut-clipboard-manager/id442160987?mt=12">Flycut</a> (Mac <span class="caps">OS</span>), <a href="https://apps.apple.com/us/app/copyclip-clipboard-history/id595191960?mt=12">CopyClip</a> (Mac <span class="caps">OS</span>): Some of the most stable clipboard managers I’ve found. I can’t do without a clipboard manager, but a lot of them crash all the time. (H/t Jahlela for Flycut. I haven’t used it.)</li>
<li><a href="https://itunes.apple.com/app/easyres/id688211836?ls=1&mt=12">EasyRes</a> (Mac <span class="caps">OS</span>): Retina screens support insane resolutions but I didn’t notice because I could never set it as high. With EasyRes, you can set resolutions higher than the ones the monitor preferences allow you to select. Suddenly there’s so much space! Especially recommended if stuff on screens usually looks ridiculously big to you while at the same time you can see the space between the pixels.</li>
<li><a href="https://code.visualstudio.com">Visual Studio Code</a> (many OSs): Great for writing. Case in point: I’m writing this in <span class="caps">VS</span> Code.</li>
</ol>
<h2 id="chrome-vivaldi-and-maybe-firefox-too"><a class="toclink" href="#chrome-vivaldi-and-maybe-firefox-too">Chrome, Vivaldi, and Maybe Firefox Too</a></h2>
<ol>
<li><a href="https://1password.com/">1Password</a>: A password manager. Please use one! Then you can use different random passwords for every website, and never reuse passwords anymore. Make encrypted backups from time to time just in case.</li>
<li><a href="https://bitwarden.com/">Bitwarden</a>: Another cheaper password manager. I use this one. Make encrypted backups from time to time just in case.</li>
<li><a href="https://chrome.google.com/webstore/detail/typio-form-recovery/djkbihbnjhkjahbhjaadbepppbpoedaa?hl=en">Typio Form Recovery</a>: A Chrome/Vivaldi extension that automatically backs up form input. Useful to recover long comments after a browser crash, a website malfunction, or after accidentally navigating away from the page.</li>
<li><a href="https://chrome.google.com/webstore/detail/view-image/jpcmhcelnjdmblfmjabdeclccemkghjk?hl=en">View Image</a>: A Chrome/Vivaldi extension that reimplements the Google Images “View Image” and “Search by Image” buttons.</li>
<li><a href="https://chrome.google.com/webstore/detail/i-dont-care-about-cookies/fihnjjcciajhdojfnbdddfaoknhalnja?hl=en">I don’t care about cookies</a>: Firefox and Chrome extension that hides annoying but legally required cookie warnings.</li>
<li><a href="https://chrome.google.com/webstore/detail/text-blaze/idgadaccgipmpannjkmfddolnnhmeklj/">Text Blaze</a>: Similar to Auto Text Expander, but better maintained and adds the templates to the context menu.</li>
<li><a href="https://chrome.google.com/webstore/detail/xpath-helper/hgimnogjllphhhkhlmebbmlgjoejdpjl?hl=en">XPath Helper</a>: Shows you an XPath for arbitrary elements on the page and (often more importantly) shows you the text content of the selected elements. Great for copying things from pages that would be too tedious to copy one by one.</li>
<li><a href="https://chrome.google.com/webstore/detail/event-merge-for-google-ca/idehaflielbgpaokehlhidbjlehlfcep/related?hl=en">Event Merge</a>: Visually merge duplicate events in Google Calendar.</li>
</ol>
<h2 id="android-and-maybe-ios"><a class="toclink" href="#android-and-maybe-ios">Android and Maybe iOS</a></h2>
<ol>
<li><a href="https://play.google.com/store/apps/details?id=com.hyperionics.avar&hl=en_US&gl=US">Voice Aloud Reader</a> (Android): It reads text to you. I went from visual reading at ~ 150 <span class="caps">WPM</span> up to listening at ~ 400 <span class="caps">WPM</span>. Effectively the difference is even greater because reading at a slow rate is so boring that I easily get distracted by my own thoughts and forget to pay attention to what I’m reading. The <span class="caps">UI</span> is terrible, but it extracts PDFs better than other tools like Speechify (which wins in terms of <span class="caps">UI</span>). It also crashes more rarely.</li>
<li><a href="http://savemytime.co/">Save my Time</a> (Android): A time tracking app that asks you periodically what you’ve been up to when you unlock your phone. I’ve tried a dozen or so time-tracking apps (after reviewing almost a hundred), and this one has worked best for me.</li>
<li><a href="https://www.camscanner.com">CamScanner</a> (Android, iOS): Makes scanners obsolete. I think at one point they got banned from Play Store because they installed a backdoor on the phone to display some ads. The app is back in the store, so I hope that’s no longer the case.</li>
<li><a href="https://play.google.com/store/apps/details?id=com.cyberlink.powerdirector.DRA140225_01">PowerDirector</a> (Android): The best Android app I’ve found so far for video editing on the phone.</li>
</ol>
<h2 id="lifehacks"><a class="toclink" href="#lifehacks">Lifehacks</a></h2>
<ol>
<li>Use two-factor authentication but make sure you can restore your access to your accounts if you lose your phone, it’s stolen, or you sell it.</li>
<li>Automate your backups.</li>
<li>Also back up your Google Takeouts, social media data, password manager (but encrypted), etc.</li>
<li>You can turn broken slats of your bed into hangboards for finger training.</li>
<li>Keep as much money as possible in a well-diversified portfolio that matches your investment horizon.</li>
<li>Take a bunch of experimental longevity drugs and adjust your investment horizon accordingly.</li>
<li>Have close friends with opposite opinions on things who (the friends) can argue for them well.</li>
<li>Use parentheticals to resolve referential ambiguity <em>Infinite Jest</em> style.</li>
<li>Do all the effective altruism things, because what’s the point otherwise.</li>
</ol>Building a Climbing Wall2020-09-02T19:00:00+00:002020-10-23T00:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2020-09-02:/building-a-climbing-wall.html<p>I’ve constructed my own home climbing wall. Here I document my plans, processes, and problems.</p><style>
.half-width img,
.half-width p {
display: block;
max-width: 50%;
}
iframe.instagram-media-rendered {
margin-left: auto !important;
margin-right: auto !important;
}
video {
display: block;
margin-left: auto;
margin-right: auto;
margin-bottom: 10px;
}
</style>
<div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#gallery">Gallery</a></li>
<li><a href="#design">Design</a><ul>
<li><a href="#preferred-design-3-modules-of-1-meter">Preferred Design – 3 Modules of 1 Meter</a><ul>
<li><a href="#use-cases">Use Cases</a></li>
<li><a href="#measures">Measures</a></li>
<li><a href="#3-d-model">3-D Model</a></li>
<li><a href="#anticipated-problems">Anticipated Problems</a></li>
</ul>
</li>
</ul>
</li>
<li><a href="#discarded-designs">Discarded Designs</a><ul>
<li><a href="#6-modules-of-50-cm">6 Modules of 50 cm</a></li>
<li><a href="#multi-angle-wall">Multi-Angle Wall</a></li>
<li><a href="#hanging-wall">Hanging Wall</a></li>
</ul>
</li>
<li><a href="#process">Process</a></li>
<li><a href="#problems">Problems</a></li>
<li><a href="#enhancements">Enhancements</a></li>
</ul>
</div>
<h2 id="gallery"><a class="toclink" href="#gallery">Gallery</a></h2>
<blockquote class="instagram-media" data-instgrm-permalink="https://www.instagram.com/tv/CEo1s4NDEJp/?utm_source=ig_embed&utm_campaign=loading" data-instgrm-version="12" style=" background:#FFF; border:0; border-radius:3px; box-shadow:0 0 1px 0 rgba(0,0,0,0.5),0 1px 10px 0 rgba(0,0,0,0.15); margin: 1px; max-width:540px; min-width:326px; padding:0; width:99.375%; width:-webkit-calc(100% - 2px); width:calc(100% - 2px);"><div style="padding:16px;"> <a href="https://www.instagram.com/tv/CEo1s4NDEJp/?utm_source=ig_embed&utm_campaign=loading" style=" background:#FFFFFF; line-height:0; padding:0 0; text-align:center; text-decoration:none; width:100%;" target="_blank"> <div style=" display: flex; flex-direction: row; align-items: center;"> <div style="background-color: #F4F4F4; border-radius: 50%; flex-grow: 0; height: 40px; margin-right: 14px; width: 40px;"></div> <div style="display: flex; flex-direction: column; flex-grow: 1; justify-content: center;"> <div style=" background-color: #F4F4F4; border-radius: 4px; flex-grow: 0; height: 14px; margin-bottom: 6px; width: 100px;"></div> <div style=" background-color: #F4F4F4; border-radius: 4px; flex-grow: 0; height: 14px; width: 60px;"></div></div></div><div style="padding: 19% 0;"></div> <div style="display:block; height:50px; margin:0 auto 12px; width:50px;"><svg width="50px" height="50px" viewBox="0 0 60 60" version="1.1" xmlns="https://www.w3.org/2000/svg" xmlns:xlink="https://www.w3.org/1999/xlink"><g stroke="none" stroke-width="1" fill="none" fill-rule="evenodd"><g transform="translate(-511.000000, -20.000000)" fill="#000000"><g><path d="M556.869,30.41 C554.814,30.41 553.148,32.076 553.148,34.131 C553.148,36.186 554.814,37.852 556.869,37.852 C558.924,37.852 560.59,36.186 560.59,34.131 C560.59,32.076 558.924,30.41 556.869,30.41 M541,60.657 C535.114,60.657 530.342,55.887 530.342,50 C530.342,44.114 535.114,39.342 541,39.342 C546.887,39.342 551.658,44.114 551.658,50 C551.658,55.887 546.887,60.657 541,60.657 M541,33.886 C532.1,33.886 524.886,41.1 524.886,50 C524.886,58.899 532.1,66.113 541,66.113 C549.9,66.113 557.115,58.899 557.115,50 C557.115,41.1 549.9,33.886 541,33.886 M565.378,62.101 C565.244,65.022 564.756,66.606 564.346,67.663 C563.803,69.06 563.154,70.057 562.106,71.106 C561.058,72.155 560.06,72.803 558.662,73.347 C557.607,73.757 556.021,74.244 553.102,74.378 C549.944,74.521 548.997,74.552 541,74.552 C533.003,74.552 532.056,74.521 528.898,74.378 C525.979,74.244 524.393,73.757 523.338,73.347 C521.94,72.803 520.942,72.155 519.894,71.106 C518.846,70.057 518.197,69.06 517.654,67.663 C517.244,66.606 516.755,65.022 516.623,62.101 C516.479,58.943 516.448,57.996 516.448,50 C516.448,42.003 516.479,41.056 516.623,37.899 C516.755,34.978 517.244,33.391 517.654,32.338 C518.197,30.938 518.846,29.942 519.894,28.894 C520.942,27.846 521.94,27.196 523.338,26.654 C524.393,26.244 525.979,25.756 528.898,25.623 C532.057,25.479 533.004,25.448 541,25.448 C548.997,25.448 549.943,25.479 553.102,25.623 C556.021,25.756 557.607,26.244 558.662,26.654 C560.06,27.196 561.058,27.846 562.106,28.894 C563.154,29.942 563.803,30.938 564.346,32.338 C564.756,33.391 565.244,34.978 565.378,37.899 C565.522,41.056 565.552,42.003 565.552,50 C565.552,57.996 565.522,58.943 565.378,62.101 M570.82,37.631 C570.674,34.438 570.167,32.258 569.425,30.349 C568.659,28.377 567.633,26.702 565.965,25.035 C564.297,23.368 562.623,22.342 560.652,21.575 C558.743,20.834 556.562,20.326 553.369,20.18 C550.169,20.033 549.148,20 541,20 C532.853,20 531.831,20.033 528.631,20.18 C525.438,20.326 523.257,20.834 521.349,21.575 C519.376,22.342 517.703,23.368 516.035,25.035 C514.368,26.702 513.342,28.377 512.574,30.349 C511.834,32.258 511.326,34.438 511.181,37.631 C511.035,40.831 511,41.851 511,50 C511,58.147 511.035,59.17 511.181,62.369 C511.326,65.562 511.834,67.743 512.574,69.651 C513.342,71.625 514.368,73.296 516.035,74.965 C517.703,76.634 519.376,77.658 521.349,78.425 C523.257,79.167 525.438,79.673 528.631,79.82 C531.831,79.965 532.853,80.001 541,80.001 C549.148,80.001 550.169,79.965 553.369,79.82 C556.562,79.673 558.743,79.167 560.652,78.425 C562.623,77.658 564.297,76.634 565.965,74.965 C567.633,73.296 568.659,71.625 569.425,69.651 C570.167,67.743 570.674,65.562 570.82,62.369 C570.966,59.17 571,58.147 571,50 C571,41.851 570.966,40.831 570.82,37.631"></path></g></g></g></svg></div><div style="padding-top: 8px;"> <div style=" color:#3897f0; font-family:Arial,sans-serif; font-size:14px; font-style:normal; font-weight:550; line-height:18px;"> View this post on Instagram</div></div><div style="padding: 12.5% 0;"></div> <div style="display: flex; flex-direction: row; margin-bottom: 14px; align-items: center;"><div> <div style="background-color: #F4F4F4; border-radius: 50%; height: 12.5px; width: 12.5px; transform: translateX(0px) translateY(7px);"></div> <div style="background-color: #F4F4F4; height: 12.5px; transform: rotate(-45deg) translateX(3px) translateY(1px); width: 12.5px; flex-grow: 0; margin-right: 14px; margin-left: 2px;"></div> <div style="background-color: #F4F4F4; border-radius: 50%; height: 12.5px; width: 12.5px; transform: translateX(9px) translateY(-18px);"></div></div><div style="margin-left: 8px;"> <div style=" background-color: #F4F4F4; border-radius: 50%; flex-grow: 0; height: 20px; width: 20px;"></div> <div style=" width: 0; height: 0; border-top: 2px solid transparent; border-left: 6px solid #f4f4f4; border-bottom: 2px solid transparent; transform: translateX(16px) translateY(-4px) rotate(30deg)"></div></div><div style="margin-left: auto;"> <div style=" width: 0px; border-top: 8px solid #F4F4F4; border-right: 8px solid transparent; transform: translateY(16px);"></div> <div style=" background-color: #F4F4F4; flex-grow: 0; height: 12px; width: 16px; transform: translateY(-4px);"></div> <div style=" width: 0; height: 0; border-top: 8px solid #F4F4F4; border-left: 8px solid transparent; transform: translateY(-4px) translateX(8px);"></div></div></div> <div style="display: flex; flex-direction: column; flex-grow: 1; justify-content: center; margin-bottom: 24px;"> <div style=" background-color: #F4F4F4; border-radius: 4px; flex-grow: 0; height: 14px; margin-bottom: 6px; width: 224px;"></div> <div style=" background-color: #F4F4F4; border-radius: 4px; flex-grow: 0; height: 14px; width: 144px;"></div></div></a><p style=" color:#c9c8cd; font-family:Arial,sans-serif; font-size:14px; line-height:17px; margin-bottom:0; margin-top:8px; overflow:hidden; padding:8px 0 7px; text-align:center; text-overflow:ellipsis; white-space:nowrap;"><a href="https://www.instagram.com/tv/CEo1s4NDEJp/?utm_source=ig_embed&utm_campaign=loading" style=" color:#c9c8cd; font-family:Arial,sans-serif; font-size:14px; font-style:normal; font-weight:normal; line-height:17px; text-decoration:none;" target="_blank">A post shared by Denis Drescher (@telofy)</a> on <time style=" font-family:Arial,sans-serif; font-size:14px; line-height:17px;" datetime="2020-09-02T15:03:07+00:00">Sep 2, 2020 at 8:03am <span class="caps">PDT</span></time></p></div></blockquote>
<script async src="//www.instagram.com/embed.js"></script>
<div class="lightbox-block align-center half-width">
<img
src="images/building-a-climbing-wall/IMG_20200902_142256.jpg"
alt="Climbing wall from the front"
/>
</div>
<div class="lightbox-block align-center half-width">
<img
src="images/building-a-climbing-wall/IMG_20200902_141046.jpg"
alt="Climbing wall from the side"
/>
</div>
<div class="lightbox-block align-center half-width">
<img
src="images/building-a-climbing-wall/IMG_20200902_141401.jpg"
alt="Climbing wall from the side"
/>
</div>
<div class="lightbox-block align-center half-width">
<img
src="images/building-a-climbing-wall/IMG_20200902_141417.jpg"
alt="“Shoes” of the climbing wall"
/>
</div>
<div class="lightbox-block align-center half-width">
<img
src="images/building-a-climbing-wall/IMG_20200901_204859.jpg"
alt="Eevee through climbing wall"
/>
</div>
<h2 id="design"><a class="toclink" href="#design">Design</a></h2>
<p>First, I ordered the T-nuts. They take about 1–2 months to arrive, are almost certainly needed, and are not a great sunk cost even if I decide not to go ahead with the board. I felt like I was 60% likely to go ahead with the climbing board plans, and in this majority of worlds, I didn’t want to be blocked on an order later on.</p>
<p>Then I started designing different wall concepts, watched YouTube videos on home climbing walls, and read tutorials on the topic.</p>
<p>I came up with about four designs, detailed below.</p>
<h3 id="preferred-design-3-modules-of-1-meter"><a class="toclink" href="#preferred-design-3-modules-of-1-meter">Preferred Design – 3 Modules of 1 Meter</a></h3>
<h4 id="use-cases"><a class="toclink" href="#use-cases">Use Cases</a></h4>
<ol>
<li>It should be suitable for using inside my room, at least most of the time.<ol>
<li>I’ve achieved this.</li>
</ol>
</li>
<li>It needs to be easy to take apart to take outside for a day or for moving flats. Each part should easily fit through doors and be less than 40 kg, ideally less than 30 kg.<ol>
<li>They’re about 32 kg.</li>
</ol>
</li>
<li>It should be extensible for adjustment for flats with higher ceiling.<ol>
<li>It is.</li>
</ol>
</li>
<li>It should rest at an angle of ~ 0–45°, i.e. vertical, diagonal, or anything in between. The climbable side is the one facing downwards.<ol>
<li>The angle is 48–49° off the vertical.</li>
</ol>
</li>
<li>Ideally, this angle should be flexible.<ol>
<li>It would take a lot of work, but in a larger room, I could angle it differently.</li>
</ol>
</li>
<li>The budget is < <span class="caps">CHF</span> 500 excluding climbing holds.<ol>
<li>It was about <span class="caps">CHF</span> 420 without holds.</li>
</ol>
</li>
<li>Assembly should not require any power tools beyond a drill because we don’t have those.<ol>
<li>Check.</li>
</ol>
</li>
<li>It should not require drilling of any holes in the apartment walls, floor, or ceiling.<ol>
<li>No walls or ceilings have been drilled into in the making of this wall.</li>
</ol>
</li>
<li>It should be positioned such that I can use my 2 m x 1.6 m bed as a mat.<ol>
<li>Check.</li>
</ol>
</li>
</ol>
<h4 id="measures"><a class="toclink" href="#measures">Measures</a></h4>
<ol>
<li>The beams are 100 x 8 x 8 cm. (The model below currently shows 100 x 8 x 6 cm – I’d need to update that but realistically never will.) Each module has two beams on each side, four in total.</li>
<li>The boards are 200 x 50 x 2 cm. Each module has two boards that are firmly screwed to the beams with ~ 12–14 wood screws per 50 cm.</li>
<li>My ceiling is at 240 cm. The board reaches almost all the way up to it.<ol>
<li>Various slight error added up to a few wasted cm.</li>
</ol>
</li>
<li>The board rests at a ~ 47° angle, a bit more overhanging than I had hoped but it’s fine. (To decrease the angle without it going through my ceiling, I would have to make it shorter. But I can’t make it shorter or it would have to go through my bed instead. I also don’t want to make it shorter than ~ 3 m.)<ol>
<li>It ended up at an angle 48–49° off the vertical.</li>
</ol>
</li>
<li>The kick board at the bottom is ~ 25 cm high.<ol>
<li>I decided to skip the kick board. It would’ve been hard to reach anyway because my bed is so high.</li>
</ol>
</li>
</ol>
<h4 id="3-d-model"><a class="toclink" href="#3-d-model">3-D Model</a></h4>
<p>The wood is all one shade. I’m using different shades in the model to distinguish the modules that can be taken apart easily from all the little pieces that are firmly attached to each other.</p>
<p>The 1 m beams are connected to each other with two metal bolts and by having their ends rest against each other. If this is insufficient, I can wrap a metal band around each end to keep it from splitting.</p>
<p>The boards have each three rows of holes for the T nuts. Every other row is offset by half a hole distance, which, I hope, improves stability and maybe also flexibility for the boulder problems.</p>
<p>I want to place a lamp roughly where the camera is to improve the lighting.</p>
<p>The whole climbing board is in my room. The “feet” of the kick board press against one wall (I’ll put something soft there to protect the wall from damage) and the top rests in the corner between the opposite wall and the ceiling. I’ll put a blanket there to protect the wall from the board and from my shoes if they hit the wall during a climb.</p>
<p>The whole board is flush against one side wall of the room. I’ll put some foam or a blanket there as well to protect the wall.</p>
<video controls width="640" height="360">
<source src="images/building-a-climbing-wall/0001-0210.mkv"
type="video/mp4">
Your browser doesn’t seem to support this format. <a href="https://photos.app.goo.gl/5aggbj3DdRVP6h17A">You can watch it here.</a>
</video>
<div class="lightbox-block align-center half-width">
<img
src="images/building-a-climbing-wall/Screenshot from 2020-07-14 12-23-41.png"
alt="Climbing wall from behind"
/>
</div>
<div class="lightbox-block align-center half-width">
<img
src="images/building-a-climbing-wall/Screenshot from 2020-07-14 12-27-39.png"
alt="Climbing wall from behind"
/>
</div>
<div class="lightbox-block align-center half-width">
<img
src="images/building-a-climbing-wall/Screenshot from 2020-07-14 12-30-11.png"
alt="Climbing wall from the front"
/>
</div>
<h4 id="anticipated-problems"><a class="toclink" href="#anticipated-problems">Anticipated Problems</a></h4>
<ol>
<li>A single 350 x 8 x 8 cm beam would deform only ~ 1 cm in the middle even when I do dynamic exercises on it. That would be unproblematic for climbing but it may loosen the outer ones of the screws that hold the boards. The modular design will probably lead to more bending at the joints, and I can avoid putting screws between beam joints and the edges of the boards. Still, I want to anticipate this problem better.<ol>
<li>One idea is to put double-sided adhesive tape between the boards and the beams. I’ll probably do that anyway to facilitate the assembly. The tape is about 1 mm thick. Over 50 cm, the beam may not deform very much. ~ 1 mm or a bit less of that can be absorbed by the tape. There’ll be a bit of room there anyway because of the backs of the T nuts. I’ll try to avoid having T nuts press against the beams in the middle of the boards.</li>
<li>I might put rubber under the heads of the screws that hold the boards. That might add another ~ 0.5 mm of tolerance before the screws start to suffer.</li>
<li>The boards themselves are probably also somewhat flexible.</li>
</ol>
</li>
<li>I haven’t decided how I want to connect the board proper to the beams that emanate from the kick board. If I use only one bolt there, I’m worried that it’ll break off the kick board and fold upward.<ol>
<li>I could prevent that by placing something heavy on these beams.</li>
<li>I could put a 90° angled metal piece where the beams meet to keep them in place. That would make the bottom piece harder to take apart. </li>
<li>I could also saw pieces out of the beams to fit them into each other. But I’d like to avoid that since I’d have to do it with a hand saw and a chisel, and it seems like a lot of work.</li>
<li>I could rethink the whole bottom section.</li>
</ol>
</li>
<li>The board might suffer some parallel deformation, i.e. deformation along a horizontal axis. The boards should prevent that, but maybe the screws aren’t strong enough for that.<ol>
<li>I don’t think this is likely to be a problem, but if it does become a problem, I can put an X of ropes across the back of the board. If that’s not enough, I can put an X of beams across the back of the board.</li>
</ol>
</li>
</ol>
<p>I ended up addressing none of these either because the risk seemed low or because I can address them as they arise.</p>
<h2 id="discarded-designs"><a class="toclink" href="#discarded-designs">Discarded Designs</a></h2>
<h3 id="6-modules-of-50-cm"><a class="toclink" href="#6-modules-of-50-cm">6 Modules of 50 cm</a></h3>
<div class="lightbox-block align-center half-width">
<img
src="images/building-a-climbing-wall/Screenshot from 2020-07-18 02-45-06.png"
alt="Climbing wall from behind"
/>
</div>
<p>The idea here was to have particularly small, light, and interchangeable modules. They are all identical but fit together because the beams are angled just right over the 50 cm for the next beam to fit next to them. I’ve discarded this design because of doubts about the stability of the connections of the beams.</p>
<h3 id="multi-angle-wall"><a class="toclink" href="#multi-angle-wall">Multi-Angle Wall</a></h3>
<p>Another way of getting a longer wall into my small room was to have, say, four modules at angles 0°, 20°, 45°, and 90° or so in that order one after another. They’d form a bent wall to the ceiling and then along the ceiling. It would’ve enabled me to train on various angles of walls. But I couldn’t think of a way to make it flexible enough for moving and adjusting to other room heights.</p>
<h3 id="hanging-wall"><a class="toclink" href="#hanging-wall">Hanging Wall</a></h3>
<p>In this idea the modules were connected by ropes rather than beams. At the top, the ropes go through a hole into the board from the front to the back where there’s a knot. Then they continue behind the board to the bottom. Below the board there’s a gap where they thread back to the front, and then go back through a hole in the top of the next board. And so on.</p>
<p>Downward pressure on protruding holds on the lower half of the board results in a downward and inward (into the board) force. But also in a tautening of the rope. The rope then keeps the board from tilting forward from the inward force.</p>
<p>The problem is of course that this only works for forces on the lower half of the board, so I’d have to use rather many boards to be able to place the holds at a sensible distance. And despite all the tautening, it still seemed like a rather shaky contraption.</p>
<p>Furthermore, I can’t think of a way to hang it in my room without drilling holes.</p>
<h2 id="process"><a class="toclink" href="#process">Process</a></h2>
<ol>
<li>Order T-nuts.</li>
<li>Come up with different designs. About four in my case.</li>
<li>Get inspiration online.</li>
<li>3-D-model the designs. I used first Vectary and then Blender for this.</li>
<li>Describe designs and ask around for feedback on them.</li>
<li>Rearrange furniture.</li>
<li>Order materials and tools.</li>
<li>Wait. Wait a lot. Wait long and hard.</li>
<li>Once the materials arrive, put it all together.</li>
</ol>
<h2 id="problems"><a class="toclink" href="#problems">Problems</a></h2>
<ol>
<li>The beams are connected with 1 cm x 20 cm metal bolts. The plan was to use wingnuts on some of these to enable easy attaching and detaching. This failed because (1) the bolts are very hard to get into the holes even though they are also 1 cm and (2) my room is too small to put the wall horizontally on the floor, so that there is always weight on the bolts.</li>
<li>Using continuous beams would’ve sacrificed the modularity but made the building process somewhat easier (or greatly easier had the room been 20 cm longer). It would’ve also made the wall 10–20% lighter.</li>
<li>It would’ve been useful to determine the point where the beams touch their shoes experimentally. I call “shoes” the beams that I put on the ground and that the ends of the 49° beams are grounded in. My 3-D model was off by 2–3 cm due to all sorts of small modifications that I made later. That created some extra work.</li>
<li>It would’ve made sense to generally optimize more for lightness. The wall now weighs about 100 kg, so we needed five people to lift it and put it in its shoes.</li>
</ol>
<p>You’ll see more of it on my Instagram. At least once I come up with a problem I can solve.</p>
<h2 id="enhancements"><a class="toclink" href="#enhancements">Enhancements</a></h2>
<ol>
<li>Since the first setup, I’ve moved the wall by 66 cm toward the window to have more space underneath. That way I can swing out when I lose feet without hitting the wall (or without hitting it hard). The move also allows me to climb up a bit farther because I get get head up higher, and it allows me to start lower, because I can climb the first horizontal meter or so over the floor rather than the bed. That makes the area accessible that was previously blocked by the 25 cm high bed.</li>
<li>One of the bed slats broke, and I turned it into eight nice crimps and a hangboard. Very recommeded!</li>
</ol>Summary of Evidence, Decision, and Causality2020-08-22T13:00:00+00:002020-08-22T13:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2020-08-22:/summary-of-evidence-decision-and-causality.html<p>Arif Ahmed’s <a href="https://www.cambridge.org/core/books/evidence-decision-and-causality/7077949D2CD42E99C08D4FBFE5321148"><em>Evidence, Decision, and Causality</em> (2014)</a> is a dense, mathematical book-length argument against causalism and for the merits of evidential over causal decision theory. It’s not a light read, so I decided that others, including future me, may benefit from a short, informal summary. I think this summary will be most interesting for people who are new to decision theory. The subsection <a href="/summary-of-evidence-decision-and-causality.html#edt-money-pump">“<span class="caps">EDT</span> Money Pump”</a> may be more generally interesting, unless I’m wrong.</p><style>
.half-width img,
.half-width p {
display: block;
max-width: 50%;
}
</style>
<div class="lightbox-block align-center half-width">
<img
src="images/summary-evidence-decision-causality/i-know-dt.jpg"
alt="“I know decision theory.” “Show me.”"
/>
</div>
<div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#the-book">The Book</a></li>
<li><a href="#responses">Responses</a></li>
<li><a href="#game-theory">Game Theory</a></li>
<li><a href="#newcombs-problem">Newcomb’s Problem</a></li>
<li><a href="#vaguely-newcomb-like-problems">Vaguely Newcomb-like Problems</a><ul>
<li><a href="#tickle-defense">Tickle Defense</a></li>
<li><a href="#comment">Comment</a></li>
</ul>
</li>
<li><a href="#proper-newcomb-like-problems">Proper Newcomb-like Problems</a><ul>
<li><a href="#prisoners-dilemma">Prisoner’s Dilemma</a><ul>
<li><a href="#comment_1">Comment</a></li>
</ul>
</li>
<li><a href="#betting-on-past-and-laws">Betting on Past and Laws</a><ul>
<li><a href="#comment_2">Comment</a></li>
</ul>
</li>
</ul>
</li>
<li><a href="#quantum-mechanical-cases">Quantum-Mechanical Cases</a><ul>
<li><a href="#comment_3">Comment</a></li>
</ul>
</li>
<li><a href="#newcomb-again">Newcomb Again</a><ul>
<li><a href="#the-case-for-one-boxing">The Case for One-Boxing</a><ul>
<li><a href="#comment_4">Comment</a><ul>
<li><a href="#edt-money-pump">EDT Money Pump</a></li>
<li><a href="#better-irrational-than-crazy">Better Irrational Than Crazy</a></li>
</ul>
</li>
</ul>
</li>
<li><a href="#the-case-for-two-boxing">The Case for Two-Boxing</a></li>
</ul>
</li>
<li><a href="#the-ultimate-contingency">The Ultimate Contingency</a></li>
<li><a href="#conclusion">Conclusion</a></li>
<li><a href="#footnotes">Footnotes</a></li>
</ul>
</div>
<h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<p>Decision theory felt like an unusually obvious personal knowledge gap to me as it plays an important role in <a href="https://www.alignmentforum.org/posts/DbuCdEbkh4wL5cjJ5/preface-to-clr-s-research-agenda-on-cooperation-conflict-and">s-risk research</a>, <a href="https://www.lesswrong.com/posts/i3BTagvt3HbPMx6PN/embedded-agency-full-text-version">some approaches to general <span class="caps">AI</span> alignment</a>, and <a href="https://longtermrisk.org/msr">evidential cooperation in large worlds</a> (<span class="caps">ECL</span>). <span class="caps">ECL</span> in particular only works for cooperators that implement noncausal decision theories (or have <a href="https://casparoesterheld.com/2017/05/12/anthropic-uncertainty-in-the-evidential-blackmail/">particular indexical uncertainty</a>). For them it may have vast moral implications.</p>
<p>Arif Ahmed’s particular thesis allows him to not only argue why evidential decision theory (<span class="caps">EDT</span>) lives up to various demands that strike him (and me) as reasonable but also why causal decision theory (<span class="caps">CDT</span>) fails to live up to various demands regardless of how reasonable we deem them. These demands can’t serve as reasons to prefer <span class="caps">CDT</span> over <span class="caps">EDT</span>, so if both decision theories fail them rather than just <span class="caps">EDT</span>, that strengthens <span class="caps">EDT</span>’s relative position.<sup id="fnref:plates-vs-bowls"><a class="footnote-ref" href="#fn:plates-vs-bowls">1</a></sup> My interest was primarily in the first type of argument, though, and I may do a bad job at covering the second type here.</p>
<p>Note that the book doesn’t touch on interesting concepts like updatelessness because they’re not distinguishing features of <span class="caps">CDT</span> and <span class="caps">EDT</span>. If you want to read more about updatelessness, then this <a href="https://www.alignmentforum.org/posts/9sYzoRnmqmxZm4Whf/conceptual-problems-with-udt-and-policy-selection">2019 discussion of updateless decision theory</a> is one that I enjoyed, and the 2015 paper <a href="https://arxiv.org/pdf/1507.01986.pdf">“Toward Idealized Decision Theory”</a> sounds exciting!</p>
<p>In the following, page and chapter references refer to <em>Evidence, Decision, and Causality</em> (2014) unless noted otherwise.<sup id="fnref:alterations"><a class="footnote-ref" href="#fn:alterations">2</a></sup></p>
<p>I’ve learned a lot from this book and thank <a href="https://casparoesterheld.com/">Caspar Oesterheld and Johannes Treutlein</a> for recommending it in various articles.</p>
<h2 id="the-book"><a class="toclink" href="#the-book">The Book</a></h2>
<p>Arif Ahmed’s own abstract describes <em>Evidence, Decision, and Causality</em> (2014) as follows:</p>
<blockquote>
<p>Most philosophers agree that causal knowledge is essential to decision-making: agents should choose from the available options those that probably cause the outcomes that they want. This book argues against this theory and in favour of Evidential or Bayesian Decision Theory, which emphasizes the symptomatic value of options over their causal role. It examines a variety of settings, including economic theory, quantum mechanics and philosophical thought-experiments, where causal knowledge seems to make a practical difference. The arguments make novel use of machinery from other areas of philosophical inquiry, including first-person epistemology and the free-will debate. The book also illustrates the applicability of decision theory itself to questions about the direction of time and the special epistemic status of agents.</p>
</blockquote>
<p>So the general structure of the book is this: It argues against a certain doctrine (<em>causalism</em>, defined below) by showing that a corresponding decision theory (<span class="caps">CDT</span>) correctly implements this doctrine but that this decision theory ends up making the same or worse recommendations than a rival decision theory (<span class="caps">EDT</span>) that ignores the doctrine. So it sets up <span class="caps">EDT</span> as the more parsimonious decision theory (e.g., p. 81) that is preferable unless it can be shown that <span class="caps">CDT</span> makes better recommendations.</p>
<blockquote>
<p>Causalism is the doctrine that rational choice must take account of causal information. Specifically it must attend to whether and how an agent’s available acts are causally relevant to the outcomes that he desires or dreads. …</p>
<p>Evidentialism, which [this book] prefers, is the contrary view that only the diagnostic bearing of acts is of practical concern. It only matters to what extent this or that act is evidence of this or that outcome, regardless of whether the act causes the outcome or is merely symptomatic of it. (p. 1)</p>
</blockquote>
<p>The book also explains what recommendations <span class="caps">EDT</span> makes in <em>common cause</em> type of decision problems like Smoking Lesion (see <a href="https://johannestreutlein.files.wordpress.com/2018/10/a-typology-of-newcomblike-problems.pdf">A Typology of Newcomb-like Problems</a>) because of the tickle defense; and it clarifies a number of open questions I had about decision theory in general.<sup id="fnref:textbook"><a class="footnote-ref" href="#fn:textbook">3</a></sup></p>
<p>There are places in the book where there’s a block of math followed by the words “more formally” and then a longer block of math. But I could generally follow the arguments, so no math degree is needed to make sense of them. Another feature or drawback is of course that it’s written as a defense of <span class="caps">EDT</span> for proponents of <span class="caps">CDT</span>, so it disproves at length dozens of arguments that never seemed plausible to me to begin with.</p>
<p>But all of that is by design. Most of my knowledge of decision theory now stems from the book, so I’m in no position to critique it in light of its design. Instead I will summarize it and merely comment occasionally when I continue to be confused about some matters.</p>
<p>I also noticed that Ahmed shares my predilection for recondite words like <em>recondite</em> and seems to be generally a chill fellow:</p>
<blockquote>
<p>None of the arguments for incompatibilism start to look convincing when I start doing things, and then stop being convincing on reversion to my customary torpor. (p. 218)</p>
</blockquote>
<p>Ahmed is a grantee of the <a href="https://longtermrisk.org/grantmaking/#accordion-arif">Center on Long-Term Risk Fund</a> and attends online seminars of the Global Priorities Institute. (I saw him ask a question in one seminar.) I hope that has resulted and will result in fruitful collaborations on some of the most important questions of our time – in a very long sense of “our time.”</p>
<p><a href="https://www.youtube.com/watch?v=bZAg-IPADBQ">This is how he looks, moves, and speaks.</a> I find it helpful to be able to visualize the author who is speaking to me in some detail.</p>
<h2 id="responses"><a class="toclink" href="#responses">Responses</a></h2>
<p><em><a href="https://doi.org/10.1017/9781316847893">Newcomb’s Problem</a></em> is a 2018 book edited by Ahmed to which he also contributed the introduction and one chapter. I read most of it, and it does contain many new thoughts beyond the purview of the 2014 book. I did not, however, find corrections of any possible errors in <em>Evidence, Decision, and Causality</em> in it. One exception is maybe the chapter contributed by James Joyce (chapter 7, “<a href="https://pdfs.semanticscholar.org/e630/e6a355988d22e4229fd5f63036156c62c2f7.pdf">Deliberation and Stability in Newcomb Problems and Pseudo-Newcomb Problems</a>”), which defends <span class="caps">CDT</span> against some critiques in the 2014 book. I haven’t taken the time to understand this reply to the point where I could summarize it.</p>
<p>Chapter 6, “<a href="https://static1.squarespace.com/static/56d45b30f699bb6f0beeaf5a/t/5abc9d0e88251b734ff195d4/1522310416943/Greene%2C+Success-First+Decision+Theories.pdf">Success-First Decision Theories</a>,” by Preston Greene, is probably the chapter that I would recommend most highly. It fleshes out an “experimental” approach to decision theory that is motivated by a “success-first” criterion of rightness, which sounds synonymous to <span class="caps">MIRI</span>’s “winning.” That, I suppose, is all I care about when it comes to my decision theories, so I welcome more academic attention to this criterion. Greene also mentions <a href="https://arxiv.org/pdf/1710.05060.pdf">Functional Decision Theory</a> as an improved version of <a href="https://philpapers.org/archive/MEABAI.pdf">Cohesive Decision Theory</a>, the latter of which I hadn’t heard of. (Chapter 2, by Chrisoula Andreou, seemed interesting too for its discussion of something called “restraint,” which sounds similar in effect to precommitments.)</p>
<p>Another response (or review rather) that I’ve found is the descriptively titled “<a href="https://doi.org/10.1086/684183">Review of Evidence, Decision and Causality</a>” by H. Orri Stefánsson. I have yet to read it.</p>
<h2 id="game-theory"><a class="toclink" href="#game-theory">Game Theory</a></h2>
<p>The first chapter introduces Leonard J. Savage’s axioms from his 1954 book <em>The Foundations of Statistics</em> which entail the principle of strategic dominance: In some games there is one strategy that is better than all others for the player regardless of what happens afterwards, e.g., regardless of how another player responds or how a chance process turns out. The argument goes that it’s the instrumentally rational thing to do for a player to follow the dominant strategy.</p>
<p>Say, you want to build a climbing wall, but you don’t know whether, in the long run, you’ll prefer a vertical climbing wall or one at an angle. You can choose between two climbing wall designs, (1) a rigidly vertical one and (2) a wall that you can tilt to whatever angle you like. All else equal, choosing wall 2 is the dominant strategy because you’ll be indifferent between the walls if it turns out that you’ll prefer a vertical wall, but you’ll be happier with it if it turns out that you prefer a tilted wall.</p>
<p>But there are cases where dominance reasoning leads to funny implications. Say, you’re planning a climbing trip. You’re worried about dying because then it’d be your last climbing trip. You have no rope or harness, so the main risk you’re worried about is from falling and dying upon impact on the ground. But rope and harness cost money, which you could donate to an <span class="caps">EA</span> think tank directly or in your will. And a sharp edge might shear the rope, so you’ll die anyway.</p>
<p>You employ dominance reasoning: If you die, you’ll prefer to not have wasted money on rope and harness. If you don’t die, you’ll also prefer not to have wasted money on rope and harness. So either way, you’ll prefer not buying rope and harness, making it the dominant strategy.</p>
<p>Even <a href="https://en.wikipedia.org/wiki/Alex_Honnold">Alex Honnold</a> would agree that something feels off about this reasoning. The problem is that whether or not you buy rope and harness tells you something about the probability of your death from falling. And this is where decision theory comes in.</p>
<ol>
<li><span class="caps">CDT</span> asserts that it’s the causal link from buying rope and harness to the reduced risk of death that makes dominance reasoning inapplicable in this case.</li>
<li><span class="caps">EDT</span>, on the other hand, asserts that it’s the fact that buying rope and harness gives you evidence (or news) that you’ll have a reduced risk of death that makes it inapplicable.</li>
</ol>
<p>In this example, <span class="caps">CDT</span> and <span class="caps">EDT</span> agree (for different reasons) that dominance reasoning is inapplicable.</p>
<p>A few clarifications:</p>
<ol>
<li>Decision theory is all about trying to determine what acts are rational. So we start from a position of confusion about what “rational” means and seek to alleviate this confusion. <span class="caps">CDT</span> and <span class="caps">EDT</span> are two possible ways to define what is rational.</li>
<li>Decision theory, in this article, is always about subjective probabilities. Ahmed briefly discusses something called “objective decision theory,” but I’ll ignore that here. That means that it’s irrelevant for these thought experiments how terribly wrong the agents may be about the facts of the world but only that their decisions are instrumentally rational given whatever they believe to be true.</li>
<li>News is a great term for the sort of evidence that <span class="caps">EDT</span> is concerned with in that it is only additional evidence that is interesting. A repetition of known evidence is not. This has maybe led to misunderstandings of <span class="caps">EDT</span> in the past (p. 88).</li>
<li>It’s a bit unclear to me what causality actually is or how it would be best defined (if there’s no fact of the matter). I’ve briefly read up on John Mackie’s account of it, but I continue to feel confused about it. Ahmed addresses different attempts to define causation in chapter 6.5 (pp. 157–165). In the following, I’ll assume that an intuitive understanding of it is enough.</li>
</ol>
<h2 id="newcombs-problem"><a class="toclink" href="#newcombs-problem">Newcomb’s Problem</a></h2>
<p>People have constructed thought experiments to find out whether it’s the evidence or the causal link that is important for dominance reasoning. The most famous one is Newcomb’s problem. A succinct formulation from <a href="https://en.wikipedia.org/wiki/Newcomb's_paradox">Wikipedia</a>, citing Wolpert and Benford (2013):</p>
<blockquote>
<p>There is an infallible predictor [or, in some formulations, a merely very reliable predictor], a player, and two boxes designated A and B. The player is given a choice between taking only box B [one-boxing], or taking both boxes A and B [two-boxing]. The player knows the following:</p>
<ol>
<li>Box A is clear, and always contains a visible $1,000.</li>
<li>Box B is opaque, and its content has already been set by the predictor:<ol>
<li>If the predictor has predicted the player will take both boxes A and B, then box B contains nothing.</li>
<li>If the predictor has predicted that the player will take only box B, then box B contains $1,000,000.</li>
</ol>
</li>
</ol>
<p>The player does not know what the predictor predicted or what box B contains while making the choice.</p>
</blockquote>
<p>So the hypothetical outcomes are as follows:</p>
<table>
<thead>
<tr>
<th style="text-align: left;"></th>
<th style="text-align: right;">One-boxing predicted</th>
<th style="text-align: right;">Two-boxing predicted</th>
</tr>
</thead>
<tbody>
<tr>
<td style="text-align: left;">You one-box</td>
<td style="text-align: right;">$1,000,000</td>
<td style="text-align: right;">$0</td>
</tr>
<tr>
<td style="text-align: left;">You two-box</td>
<td style="text-align: right;">$1,001,000</td>
<td style="text-align: right;">$1,000</td>
</tr>
</tbody>
</table>
<p>The reasoning of <span class="caps">EDT</span> and <span class="caps">CDT</span> respectively, from <a href="https://faculty.arts.ubc.ca/rjohns/nozick_newcomb.pdf">Nozick (1969)</a>:</p>
<blockquote>
<p><strong>First Argument [<span class="caps">EDT</span>]</strong>: If I take what is in both boxes, the being, almost certainly, will have predicted this and will not have put the $1,000,000 in the second box, and so I will, almost certainly, get only $1,000. If I take only what is in the second box, the being, almost certainly, will have predicted this and will have put the $1,000,000 in the second box, and so I will, almost certainly, get $1,000,000. Thus, if I take what is in both boxes, I, almost certainly, will get $1,000. If I take only what is in the second box, I, almost certainly, will get $1,000,000. Therefore I should take only what is in the second box.</p>
<p><strong>Second Argument [<span class="caps">CDT</span>]</strong>: The being has already made his prediction, and has already either put the $1,000,000 in the second box, or has not. The $1,000,000 is either already sitting in the second box, or it is not, and which situation obtains is already fixed and determined. If the being has already put the $1,000,000 in the second box, and I take what is in both boxes I get $1,000,000 + $1,000, whereas if I take only what is in the second box, I get only $1,000,000. If the being has not put the $1,000,000 in the second box, and I take what is in both boxes I get $1,000, whereas if I take only what is in the second box, I get no money. Therefore, whether the money is there or not, and which it is already fixed and determined, I get $1,000 more by taking what is in both boxes rather than taking only what is in the second box. So I should take what is in both boxes.</p>
</blockquote>
<p>So in each case it’s clear what the outcome is for both decision theories.<sup id="fnref:combs"><a class="footnote-ref" href="#fn:combs">4</a></sup> But the dual problem in the field of decision theory is that it’s not only contested how one arrives at a rational decision but also how to recognize a rational decision in the first place: The criterion (or criteria) of rightness is (are) contested too.</p>
<p>A few criteria that I’ve come across:</p>
<ol>
<li>Maximal terminal utility (Ahmed’s favored criterion).</li>
<li>Minimal foreseeable regret (pp. 59, 68, and many more).</li>
<li>Maximal utility across a probability distribution of all world-histories. (This is my attempt to <a href="https://www.alignmentforum.org/posts/9sYzoRnmqmxZm4Whf/conceptual-problems-with-udt-and-policy-selection#Terminology_Notes_References">extract the criterion from <span class="caps">UDT</span> 1.0</a>. Does that make sense?)</li>
<li><a href="https://casparoesterheld.com/2017/01/17/decision-theory-and-the-irrelevance-of-impossible-outcomes/">Irrelevance of impossible outcomes</a> (not meant to be used in isolation, I think).</li>
</ol>
<p>To my knowledge, it’s uncontested that <span class="caps">EDT</span> maximizes terminal utility in Newcomb’s problem. But using the criterion of minimal regret, one-boxing could be said to foreseeably lead to regret that could be avoided by two-boxing: An agent who one-boxes will be in a world in which there is money under both boxes, so by one-boxing, they miss out on $1,000. Two-boxers, on the other hand, will be in worlds in which there is only $1,000 under both boxes in total and so not miss out on any money under any present boxes. (To me this sounds like a funny form of regret, but that, I suppose, is because I intuitively evaluate things by their terminal utility and can’t fully, intuitively switch into a mindset in which I care only about regret.)</p>
<p>Ahmed discusses regret in the context of the “Why Ain’cha Rich?” argument (ch. 7.3.1, pp. 181–183) and other places (ch. 3.2, pp. 68–73; ch. 7.4.2, pp. 199–201). Apart from the, to me, more interesting arguments for why regret (alone) is not a good criterion, he also shows that <span class="caps">CDT</span> leads to foreseeable regret in other cases (e.g., ch. 7.4.3, pp. 201–211), so that it’s in particular no argument to prefer <span class="caps">CDT</span> to <span class="caps">EDT</span>.</p>
<p>I mentioned above an example in which there is a causal but no evidential effect:</p>
<blockquote>
<p><strong>Armour</strong>: You are a medieval soldier and tomorrow is the day of the big battle. You have the option to buy a suit of armour for thirty florins. But the predictor tells you that you will be dead by the end of tomorrow. The predictor’s strike rate for this sort of prediction is 99 per cent, whether or not the predictee wears armour. Should you spend what is probably your last thirty florins on a new suit of armour or on getting drunk? (p. 197)</p>
</blockquote>
<p>Apart from the prediction there is nothing supernatural meant to be going on in this example. Armour still has a causal effect on protecting you. (Maybe without armor you die within 10 seconds of the start of the battle and with armor it takes a minute?) But the prediction is so good that you gain no consummate evidence of survival from wearing armor. <span class="caps">EDT</span> here applies dominance reasoning and recommends getting drunk.</p>
<p>I’m unsure what <span class="caps">CDT</span>’s recommendation would be: Ahmed writes in a footnote “it is arguable that Causal Decision Theory itself says nothing about this case, at least not in the present formulation of it.”</p>
<h2 id="vaguely-newcomb-like-problems"><a class="toclink" href="#vaguely-newcomb-like-problems">Vaguely Newcomb-like Problems</a></h2>
<p>Ahmed’s phrasing is consistent in that it always frames <span class="caps">EDT</span> as the more parsimonious alternative to <span class="caps">EDT</span> that has to shoulder the burden of proof to show that causal relations are relevant for decision making. (But his arguments don’t rely on that.) The math for deriving an <span class="caps">EDT</span> decision was usually simple and the math to derive the <span class="caps">CDT</span> decision required a lot more hoop-jumping to get right (e.g., ch. 26–2.7, pp. 48–53). In some cases, deriving the <span class="caps">CDT</span> decision required resolving complicated metaphysical issues and hinged on definitorial minutia (ch. 6, pp. 146–165).</p>
<p>But so far nothing has quite convinced me that it wouldn’t be possible for a proponent of <span class="caps">CDT</span> to find a different mathematical formalism that makes <span class="caps">CDT</span> very simple to work with. Often the intuitiveness of concepts is not so much a proxy for some principled measure such as Kolmogorov complexity but for the <a href="https://wiki.lesswrong.com/wiki/Inferential_distance">inferential distance</a> of the concepts for the particular person working with them. (<a href="https://docs.google.com/document/d/1JoHVvXBpD3Gb34HYwCRM6hT_IECkR8ljQOv0lHCCsZ4/edit">See also this draft by Linda Linsefors.</a>) So I share the intuition that <span class="caps">CDT</span> is needlessly complex, but I’m not convinced that that’s not just some uninformative personal contingency (that may even be related to why I’m reading Ahmed and not Pearl).</p>
<p>But that said, Ahmed describes the discussion that followed the Newcomb’s problem as follows:</p>
<ol>
<li>Intuitions with regard to the correct decision in the Newcomb’s problem diverge. Or as he puts it: “The strange thing about Newcomb’s problem is that there is widespread agreement that it is obvious what the agent should do, but not so much over what that is.”</li>
<li>Newcomb’s problem – at least in its most obviously credible formulation – is science fictional and currently unrealistic.<sup id="fnref:transparent-boxes"><a class="footnote-ref" href="#fn:transparent-boxes">5</a></sup>
It follows from observation 1 that you can’t easily argue that Newcomb’s problem is a reason to prefer <span class="caps">CDT</span> over <span class="caps">EDT</span> (or vice versa if someone disagrees about the parsimony of <span class="caps">EDT</span>); and it follows from observation 2 that even if our intuitions were more unanimous, we could still not rely on them very well because they are unlikely to be informed by a lot of experience from evolution, culture, or life (my framing).</li>
</ol>
<p>So the discussion instead focused on (1) trying to find realistic alternatives over which <span class="caps">CDT</span> and <span class="caps">EDT</span> disagree and showing that one makes a better recommendation than another, or (2) argue against the veracity or relevance of these observations and show that one decision theory makes a better recommendation than another in the original Newcomb’s problem.</p>
<p>Chapters 4–6 respond to the first line of argument. Chapter 7 responds to the second.</p>
<h3 id="tickle-defense"><a class="toclink" href="#tickle-defense">Tickle Defense</a></h3>
<p>Predictions are not the only way in which a noncausal effect can emerge. Another purported way is for your action and the outcome to have a common cause. Ahmed argues that these types of setups fail in that they don’t elicit different recommendations from <span class="caps">EDT</span> and <span class="caps">CDT</span>. A helpful illustration from <a href="https://johannestreutlein.files.wordpress.com/2018/10/a-typology-of-newcomblike-problems.pdf">A Typology of Newcomb-like Problems</a>:</p>
<div class="lightbox-block align-center half-width">
<img
src="images/summary-evidence-decision-causality/reference-class-and-common-cause.png"
alt="Visualization of references class and common cause problems"
/>
</div>
<p>The first graph shows a reference class problem like Newcomb’s problem; the second one a common cause type of problem like the one in this section. <em>A</em> is the action, <em>O</em> is the outcome, <em>R</em> is the reference class (e.g., you in the Newcomb problem), and <em>C</em> is the common cause. The dashed line indicates a noncausal relationship.</p>
<p>A prominent example:</p>
<blockquote>
<p><strong>Smoking Lesion</strong>: Susan is debating whether not to smoke or to smoke. She believes that smoking is strongly correlated with lung cancer, but only because there is a common cause – a lesion that tends to cause both smoking and cancer. Once we fix the absence or presence of this lesion, there is no additional correlation between smoking and cancer. Susan prefers smoking without cancer to not smoking without cancer; and she prefers smoking with cancer to not smoking with cancer. Should Susan smoke? It seems clear that she should. (p. 90)</p>
</blockquote>
<p>I’ve noticed three problems with this thought experiment that often need clarifying:</p>
<ol>
<li>Some friends of mine abhor smoking so much that they have trouble accepting the hypothetical that Susan actually enjoys it. This is similar to the problem that the <a href="https://www.lesswrong.com/posts/HFyWNBnDNEDsDNLrZ/the-true-prisoner-s-dilemma">True Prisoner’s Dilemma</a> solves. If you have this problem, just mentally rephrase the Smoking Lesion to be about chewing gum or, if you’re like a colleague of mine who hates chewing gum, maybe ice cream or potato chips.</li>
<li>It’s sometimes hard to accept the hypothetical that smoking does not directly cause lung cancer because it does in the real world. Maybe that also gets easier with chewing gum or ice cream. (Subjective decision theories are about subjective beliefs that can be perfectly false. So a thought experiment can posit that an actor has any false but consistent belief whatsoever. I don’t think my friends had trouble accepting that but had trouble accessing their intuitions for a highly counterintuitive situation.)</li>
<li>Personally, I had the problem that, in fact, it did <em>not</em> seem clear to me what to do in this instance! I seemed to have been the exception to “everyone agrees that Susan should smoke in Smoking Lesion.” I understood the tickle defense better when I discovered why that intuition was not more widely shared. More on that in the comment.</li>
</ol>
<p>It was important for me to realize that this is a problem that is designed to be realistic. So what Susan considers realistic is probably along the lines of this:</p>
<blockquote>
<p>The more plausible hypothesis is that the lesion produces smoking via its effect on the agent’s desires and beliefs. We are familiar with cases of this sort. For instance, real life offers plenty of cases in which prior states of the person affect his present desires. Smoking itself is thought to increase the desire to smoke by activating the mesolimbic or ‘reward’ pathway. Something similar may be true of many addictive drugs. (p. 93)</p>
</blockquote>
<p>But there’s a tension here. To see where it lies, it is helpful to divide the decision process into three phases: (1) the situation before the decision, characterized by pre-existing desires and beliefs; (2) the decision theoretic derivation of the decision, so here the application of <span class="caps">CDT</span> or <span class="caps">EDT</span>; and (3) the execution of the decision.</p>
<p>The above description of what a realistic effect of the lesion might be is such that it affects phase 1 but not any of the later phases. In particular it does not bear on how you think about decision theory. Therefore, the very exposition of Smoking Lesion already stipulates everything (desires and beliefs) that is at all evidentially relevant to the decision. The output of the decision theory (being fully and, together with the algorithm, exclusively dependent on the pre-existing desires and beliefs) adds nothing in terms of news. Hence, Susan’s stipulated desire to smoke tells us that she has the lesion with such and such probability. That’s bad news. But the decision to smoke or not to smoke adds nothing to that. So <span class="caps">CDT</span> and <span class="caps">EDT</span> both advise Susan to smoke.</p>
<p>Ahmed puts it as follows. For his detailed arguments for why it’s sensible to suppose that these apply to the Smoking Lesion and a number of similar problems, see ch. 4.3.</p>
<blockquote>
<p>(4.7) Mental causation: the undesirable event produces the dominant option via and only via its operation upon the agent’s current predecision desires and beliefs.</p>
<p>(4.8) Transparency: the agent knows enough about her own current desires and beliefs to make the dominated option evidentially irrelevant to them.</p>
<p>(4.9) Therefore, the agent takes neither option to be evidentially relevant to the undesirable event. (Ch. 4.3, p. 91.)</p>
</blockquote>
<p><a href="https://johannestreutlein.files.wordpress.com/2018/10/a-typology-of-newcomblike-problems.pdf">Treutlein and Oesterheld (2018)</a> put it this way – and also add a paragraph of explanation that I’ll omit here:</p>
<blockquote>
<p>(5.1) A rational agent’s action can be caused by modifying (i) the agent’s beliefs, (ii) the agent’s desires, and (iii) the agent’s decision algorithm.</p>
<p>(5.2) An agent employing <span class="caps">EDT</span> has sufficient knowledge of (i) and (ii) to make their actions evidentially irrelevant to (i) and (ii).</p>
<p>(5.3) Therefore, Newcomblike problems in which a common cause influences only (i) or (ii) (or both) are incoherent.</p>
</blockquote>
<p>The only difference to my above explanation and, I think, Ahmed’s is that they argue that “the problem is incoherent, because it stipulates beliefs and desires that an agent will never hold in a realistic version of the problem; in reality, the most efficacious option has highest news value.” I don’t understand why the problem is incoherent as opposed to uninteresting. But I expect there to be a simple explanation that I just don’t know yet.</p>
<p>I have, for simplicity, phrased these arguments in the language of certainty – e.g., that Susan is certain that the lesion works through her desires and not through her decision theoretic deliberation. Ahmed, however, almost always includes calculations of the thresholds of credences in such propositions, credences above or below which his arguments go through. See, for example, pages 105 and 110.</p>
<h3 id="comment"><a class="toclink" href="#comment">Comment</a></h3>
<p>I mentioned above that my first intuition was that Susan should abstain from smoking. I double-checked that intuition with chewing gum, which I sometimes enjoy, and it remained stable. My problem was that I imagined the lesion as some sort of highly advance toxoplasma that manipulates me over decades all throughout my life: It increases endorphin levels when I think hard; releases disproportionate amounts of serotonin when I discover a highly parsimonious, elegant explanation; and increases cortisol unless I have a lot of external validation for the things I believe. It thereby incentivizes me to look for intellectual communities that put a strong emphasis on simple explanations. And it does all of that with the goal of gradually, over years, shaping my thinking such that I abandon belief in needless crutches like time, space, and causality unless there is very, very strong evidence for their necessity and only reason directly from wave function–related first principles. Then <span class="caps">EDT</span> (along with utilitarianism and Haskell) will seem much more appealing to me than alternatives, and the toxoplasma lesion will have a good chance to control my thinking from the first tickle to smoke all the way to deliberating and carrying out the decision.</p>
<p>From what I understand, the tickle defense would, in fact, not apply in such a case and <span class="caps">EDT</span> would recommend that I avoid smoking – or try to – in line with my intuition. If I then actually manage to not smoke for a considerable time, I can increase my confidence that I don’t have the lesion.</p>
<p>But the crux for me was that Smoking Lesion was meant to be realistic. I could argue that hallucinogens have systematic longterm effects on openness, which encourages some frameworks of thought (that tend to sound metaphorical and inscrutable to me), discourages others, and changes tribal identification. But that’d be beside the point: Someone could just add to the thought experiment a detailed account of what Susan considers realistic. We’d get a few different variations of the Smoking Lesion, and the Tickle Defense would apply to some and not to others.</p>
<p>Another thought experiment I hear often is the Chewing Gum Throat Abscesses problem from <a href="https://intelligence.org/files/TDT.pdf">Yudkowsky (2010)</a>. Here the original setup with minor alterations: </p>
<blockquote>
<p>Suppose that a recently published medical study shows that chewing gum seems to cause throat abscesses – an outcome-tracking study showed that of people who chew gum, 90% died of throat abscesses before the age of 50. Meanwhile, of people who do not chew gum, only 10% die of throat abscesses before the age of 50. … But now a second study comes out, which shows that most gum-chewers have a certain gene, <span class="caps">CGTA</span>, and the researchers produce a table showing the following mortality rates:</p>
</blockquote>
<table>
<thead>
<tr>
<th></th>
<th style="text-align: right;"><span class="caps">CGTA</span> present</th>
<th style="text-align: right;"><span class="caps">CGTA</span> absent</th>
</tr>
</thead>
<tbody>
<tr>
<td>Chew gum</td>
<td style="text-align: right;">89% die</td>
<td style="text-align: right;">8% die</td>
</tr>
<tr>
<td>Don’t chew gum</td>
<td style="text-align: right;">99% die</td>
<td style="text-align: right;">11% die</td>
</tr>
</tbody>
</table>
<blockquote>
<p>The researchers hypothesize that because people with the gene <span class="caps">CGTA</span> are highly susceptible to throat abscesses, natural selection has produced in them a tendency to chew gum, which protects against throat abscesses. … Having learned of this new study, would you choose to chew gum? Chewing gum helps protect against throat abscesses whether or not you have the gene <span class="caps">CGTA</span>. Yet a friend who heard that you had decided to chew gum (as people with the gene <span class="caps">CGTA</span> often do) would be quite alarmed to hear the news – just as she would be saddened by the news that you had chosen to take both boxes in Newcomb’s problem. This is a case where evidential decision theory seems to return the wrong answer, calling into question the validity of the evidential rule “Take actions such that you would be glad to receive the news that you had taken them.”</p>
</blockquote>
<p>I like this version a lot more than the Smoking Lesion because it probably evokes the strong intuitive aversive reaction to the substance in fewer people. It’s also more open-ended because it doesn’t give away the agent’s desire in the setup.</p>
<p>Yudkowsky resolves it in chapter 9.1 of his paper. In my words: If the problem is meant to be realistic, it seems unlikely that a gene could’ve evolved whose effect goes through causing people to seek out a study that hasn’t even been around for long enough to have an effect on natural selection. It’s again more likely that the gene influencers desires. So this <span class="caps">EDT</span> friend would be reassured if you told her that you never had an unusual desire to chew gum (have never chewed gum at an unusual rate, if you don’t trust your introspection) and have just decided to do so because of the study results. Or else, if you’ve long been a chewing gum addict, then she does have reason to worry, but the study and subsequent decision to continue with the gum chewing doesn’t add to that worry.</p>
<p>Yudkowsky also mentions a complicated “metatickle defense” due to <a href="https://psycnet.apa.org/doi/10.1007/BF00140057">Eells (1984)</a>, in which <span class="caps">EDT</span> recursively conditions on the current preliminary decision until it reaches a stable point – if I understand the summary correctly. <a href="https://impartial-priorities.org/thoughts-on-the-transparent-newcombs-problem.html#updatelessness">I’ve run into situations</a> where I thought I ought to reason along such lines, but Ahmed’s analysis of the Ramsey Thesis on pages 220–226 makes me think that he would not be sympathetic to this approach. He also does not in fact use it.</p>
<p>Surprisingly, Ahmed has a paper on PhilPapers, “<a href="https://philpapers.org/archive/AHMSCA.docx">Sequential Choice and the Agent’s Perspective</a>,” in which he uses the Smoking Lesion as an example without taking the Tickle Defense into account or even mentioning it. It was informally published there in 2018 after having been edited (according to timestamps in the file) in 2017. So it’s much more recent than the book. (It also uses some exact sentences from the book.) Perhaps he no longer endorses the Tickle Defense for reasons I don’t know of, or he considers such examples to be science-fictional (like my intuitive reading) unless stated otherwise.</p>
<h2 id="proper-newcomb-like-problems"><a class="toclink" href="#proper-newcomb-like-problems">Proper Newcomb-like Problems</a></h2>
<h3 id="prisoners-dilemma"><a class="toclink" href="#prisoners-dilemma">Prisoner’s Dilemma</a></h3>
<p>Ahmed concludes that the attempts to create realistic Newcomb-like problems through the “common cause” setup have failed. If that were all that can be said on the issue, the case would rest – unsatisfyingly for me – on <span class="caps">EDT</span>’s seemingly greater parsimony. But luckily for me, Ahmed did find – and develop too – actual realistic Newcomb-like problems!</p>
<p>One well-known one is the prisoner’s dilemma (or “prisoners’ dilemma” as Ahmed calls it<sup id="fnref:prisoners-dilemma"><a class="footnote-ref" href="#fn:prisoners-dilemma">6</a></sup>), covered in chapter 4.6, pages 108–119:</p>
<blockquote>
<p>Alice and Bob are strapped to chairs facing in opposite directions. In front of each of them is 1,000 dollars. Each must now choose between taking it and leaving it. Each of them will receive 1 million dollars if and only if the other player doesn’t take the 1,000 dollars.</p>
</blockquote>
<p>He gives us the payoff table too (M for million, K for thousand):</p>
<table>
<thead>
<tr>
<th style="text-align: left;"></th>
<th style="text-align: right;">Bob cooperates</th>
<th style="text-align: right;">Bob defects</th>
</tr>
</thead>
<tbody>
<tr>
<td style="text-align: left;">Alice cooperates</td>
<td style="text-align: right;">M, M</td>
<td style="text-align: right;">0, M + K</td>
</tr>
<tr>
<td style="text-align: left;">Alice defects</td>
<td style="text-align: right;">M + K, 0</td>
<td style="text-align: right;">K, K</td>
</tr>
</tbody>
</table>
<p>Here, Ahmed argues, it is realistic for Alice to suppose in some situations that Bob is very similar to her – be it because they’re twins or because the payoffs are so different that even a slight similarity is enough for the following argument to go through. This similarity might be one of their initial intuitions about the game, say, because something about their characters makes them favor cooperation or defection on some belief or desire level. (In which case the Tickle Defense would apply.) But the similarity might also be one of their reasoning processes or of their ways of executing decisions. It’s not important that this is actually the case but that it’s realistic (because the game is supposed to be realistic) for Alice to think it might be so with a sufficient probability.</p>
<p>Given the payoffs of the game, Ahmed calculates that the conditional probability of Bob making a certain choice given that Alice made that choice needs to exceed 50.05% to be worth it in expectation (p. 110).</p>
<p>Ahmed mentions the “false consensus effect” – where people take their own preference to be indicative of how widespread that preference is – as a reason to think that reasoning of this sort is itself widespread. He also argues that this need not be irrational, despite the name “<em>false</em> consensus effect.”</p>
<p>This is bad news in a way: If the intuition is that you should always defect in a prisoner’s dilemma–like situations and <span class="caps">EDT</span> in some cases recommends cooperating, that’d be a disproof of <span class="caps">EDT</span>. But not if you can show that <span class="caps">EDT</span> makes that recommendation only exactly in the prisoner’s dilemma–like situations where it’s correct to do so.</p>
<p>With this realistic example of a decision situation in which <span class="caps">EDT</span> and <span class="caps">CDT</span> diverge in their recommendations, the argument can no longer be about showing that any situations in which this happens are too unrealistic to be worthy of consideration and then hoping that <span class="caps">EDT</span> will prevail on its parsimony alone. Now the argument needs to be about showing that <span class="caps">EDT</span> actually recommends the better decision in these divergent cases.</p>
<p>Ahmed argues for this in the rest of chapter 4.6 (pp. 115–119) based on strange discontinuities in <span class="caps">CDT</span>’s recommendations. Please read these if you’re unconvinced of the correctness of <span class="caps">EDT</span>’s recommendation in proper Newcomb-like problems.</p>
<h4 id="comment_1"><a class="toclink" href="#comment_1">Comment</a></h4>
<p>What also comes to mind is that studies that determine such things as a willingness to pay are generally considered more reliable if the subjects work with actual monetary payoffs that they can keep after the study rather than with hypothetical money. It might be that this is fully due to some effects like social desirability that subjects deem less valuable than certain amounts of real money. But maybe it’s also partially about the deliberation or execution after it’s clear that the study is about real money. (Perhaps this could be tested by picking a context in which this effect is strong and testing a few common explanations like social desirability by changing, say, how many people learn about the subjects’ decisions. If the effect can largely be explained, my hypothesis is wrong. I didn’t try to find out whether such a study already exists.)</p>
<p>But more generally, we routinely trust that study results extrapolate to people who were not part of the study – at least if the study lives up to our quality standards. This trust is usually not considered to be sensitive to whether the subjects had time to think about their choices or had to make them rapidly or while being mentally occupied. It is, however, sensitive to how representative we think the subjects were of some population or other and how many were employed. Extrapolations from one’s own person, a sample of one, require very high representativeness to make up for the extremely small sample size.</p>
<p>Even more generally, it seems to be common for me and among friends of mine (and with all the topical caveats about external validity also maybe the rest of humanity) to reason like so: “I want to do X. Would Bob mind if I do X? He’s on the phone, so I can’t ask. Would I mind if Bob did X? Not really. So it’s probably fine if I do X.” I usually adjust this line of reasoning for differences between me and Bob that I know about. But even so it’s probably my own character that does the heavy lifting in this simulation of Bob, not the minor adjustments that I make consciously. I’ve found this to be viable when I know the other well or we’re very similar, and when I know that we share relevant knowledge.</p>
<p>The last reservation feels particularly interesting: If I wonder whether Bob would mind if I ate some of his blueberries, the outcome might be (1) Bob would mind, but (2) he wouldn’t mind if he realized just how much I love blueberries. If I suppose that I told him about it weeks ago, then I think it’d be realistic for me to suppose that he’d probably remember it but only with a delay. So he would start out in state 1 but switch to state 2 after a few seconds of reflection. This would make me more hesitant to eat his blueberries. But up the ante sufficiently – say, I’d pass out without blueberries – and I’d still eat them. (Not all of them of course. Just enough not to pass out. I’m not some mean meany-pants blueberry stealer person!) This seems to me to resemble the step from simulating desires and beliefs to simulating a reasoning process too.</p>
<h3 id="betting-on-past-and-laws"><a class="toclink" href="#betting-on-past-and-laws">Betting on Past and Laws</a></h3>
<p>But Ahmed didn’t stop there! The book also includes contributions of his own that he had separately published as the papers “<a href="https://doi.org/10.1093/bjps/axt021">Causal Decision Theory and the Fixity of the Past</a>” and “<a href="https://doi.org/10.1215/00318108-1963725">Causal Decision Theory: A Counterexample</a>.” Again Alice has to make a hard decision:</p>
<blockquote>
<p><strong>Betting on the Past</strong>: In my pocket (says Bob) I have a slip of paper on which is written a proposition P. You must choose between two bets. Bet 1 is a bet on P at 10:1 for a stake of one dollar. Bet 2 is a bet on P at 1:10 for a stake of ten dollars. So your pay-offs are as in Table 5.1. Before you choose whether to take Bet 1 or Bet 2 I should tell you what P is. It is the proposition that the past state of the world was such as to cause you now to take Bet 2. (p. 120)</p>
</blockquote>
<p>Table 5.1 is as follows:</p>
<table>
<thead>
<tr>
<th style="text-align: left;"></th>
<th style="text-align: right;">P</th>
<th style="text-align: right;">¬P</th>
</tr>
</thead>
<tbody>
<tr>
<td style="text-align: left;">Take bet 1</td>
<td style="text-align: right;">10</td>
<td style="text-align: right;">−1</td>
</tr>
<tr>
<td style="text-align: left;">Take bet 2</td>
<td style="text-align: right;">1</td>
<td style="text-align: right;">−10</td>
</tr>
</tbody>
</table>
<p>Ahmed describes this version of the thought experiment as “completely vague about (a) the betting mechanism, vague and slightly inaccurate about (b) soft determinism, and slightly inaccurate about (c) the content of P.” He then constructs a (superficially strange) version of it that is not vague and inaccurate like that, and defends all these choices. Still the above is more intuitive, so I’ll skip over the clarifications.</p>
<p><span class="caps">CDT</span> observes that Alice can’t affect the past state of the world, so that dominance reasoning applies. And because of the particular odds ratios, bet 1 always dominates bet 2 (see p. 126). But that means that Alice would take bet 1 thereby betting that she’s determined to take bet 2. That’s of course patently false in that case and would lose her $1.</p>
<p><span class="caps">EDT</span>, on the other hand, would observe that the payoffs of 10 and −10 are impossible to realize so that bet 2 promises the good news of winning $1 and bet 1 promises the bad news of losing $1. So it recommends that Alice take bet 2.</p>
<p>Ahmed in particular contrasts the following principles:</p>
<blockquote>
<p>(5.13) Causal dominance: The dominance principle applies whenever the relevant events are causally independent of the agent’s options.</p>
<p>(5.14) Nomological dominance: the dominance principle applies only when the relevant events are nomologically independent of the agent’s options.</p>
</blockquote>
<p>(“Only when” in the definition of nomological dominance I take to mean really only “only when” and not “only and exactly when.” After all, in maybe most cases it’s uncontroversial that there’s no way to apply dominance reasoning.)</p>
<p>A second thought experiment is Ahmed’s Betting on the Laws:</p>
<blockquote>
<p>The next day Alice faces a second problem.</p>
<p><strong>Betting on the Laws</strong>: On this occasion and for some irrelevant reason, Alice must commit herself in print to the truth or falsity of her favoured system of laws [L]. She has two options: … to affirm [L]; that is, to assert that our world in fact conforms to the (jointly) deterministic generalizations that it conjoins; and … to deny it; that is, to assert that at some time and place our world violates at least one of these generalizations. (p. 130)</p>
</blockquote>
<p>The paper “<a href="https://doi.org/10.1215/00318108-1963725">Causal Decision Theory: A Counterexample</a>” makes another stipulation more salient than the book. My intuition was that Alice’s “favoured system of laws” L (it’s called L* in the book but that seems redundant here because I’m not using L for anything yet) is one that she has, say, 30% credence in, with the remaining 70% equally distributed over 10 other systems of laws plus the hypothesis that the correct system of laws is still unknown. But that is not how this is meant. Somewhat unrealistically, Alice is assumed to have almost 100% credence in L. More realistically, the argument goes through if she has > 50% credence in it, if I’m not mistaken. From the paper, but the italics are mine:</p>
<blockquote>
<p><strong>Causal Betting Principle (<span class="caps">CBP</span>)</strong>: If you face a choice between a unit bet [the sort of bet we’re dealing with here] on P and a unit bet on ¬P, and if you are certain that P is causally independent of your bet, and <em>if you are much more confident of P than of ¬P</em>, then it is rational to take the unit bet on P and irrational to take the unit bet on ¬P.</p>
</blockquote>
<p>The payoff table is even simpler than the Betting on the Past one:</p>
<table>
<thead>
<tr>
<th style="text-align: left;"></th>
<th style="text-align: right;">L</th>
<th style="text-align: right;">¬L</th>
</tr>
</thead>
<tbody>
<tr>
<td style="text-align: left;">Affirm L</td>
<td style="text-align: right;">1</td>
<td style="text-align: right;">0</td>
</tr>
<tr>
<td style="text-align: left;">Deny L</td>
<td style="text-align: right;">0</td>
<td style="text-align: right;">1</td>
</tr>
</tbody>
</table>
<p>Ahmed’s argument has two parts, the argument that it’s rational to affirm L and the argument that <span class="caps">CDT</span> recommends denying L. I don’t see a reason to doubt that affirming L is rational given the intuition from the causal betting principle above. <span class="caps">EDT</span>, according to Ahmed, recommends exactly the intuitive thing.</p>
<p>But I continue to be puzzled over some aspects of the proof that <span class="caps">CDT</span> endorses denying L. I say “endorses” because Ahmed’s argument only implies that <span class="caps">CDT</span> is neutral over affirming or denying L. In a footnote in the paper he notes: “Strictly speaking, I haven’t shown that <span class="caps">CDT</span> strictly prefers [denying L to affirming L] but only that the converse is false. But this is enough to condemn <span class="caps">CDT</span>: endorsing an irrational option is just as bad as preferring it to a rational one. In any case, the fact that <span class="caps">CDT</span> does not prefer [affirming L] … implies that it does strictly prefer [denying it] given an arbitrarily small incentive to [deny it].”</p>
<p>The argument seems to be along the lines of: <span class="caps">CDT</span> forbids conditioning on whether L is true because the laws of nature are very much out of anyone’s causal control. What remains is Alice’s credence that L implies that she’ll take one or the other of her available options. Ahmed’s calculation is perfectly parallel for both of these cases implying no preference for one over the other. And so <span class="caps">CDT</span> is neutral about the choice or, as he puts it, endorses denying L (just as it endorses affirming L).</p>
<p>The part I don’t understand is where Alice’s credence goes that L implies that she’ll take one option or the other. In the paper, the credence doesn’t factor into the calculation in the first place while in the book it features in footnote 13 on page 132 but then cancels out (or otherwise disappears) in the last one or two steps, which I don’t understand. But intuitively, I suppose, it’s unrealistic that Alice, as an embedded agent, can deduce a credence for an action of her own from L. This strikes me as something that should only be possible in exceptional cases. So I’m not particularly motivated to investigate this further. (Besides, I’m way over my timebox for this chapter’s summary.)</p>
<h4 id="comment_2"><a class="toclink" href="#comment_2">Comment</a></h4>
<p>Ahmed writes that “This argument for nomological over causal dominance is a special case of a more general argument for evidentialism.” It’s certainly an argument against causalism, which is his whole project. But I suppose evidentialism could still be incomplete, or the argument in particular could be used by someone who prefers to two-box in Newcomb’s problem to argue for something like ”nomological decision theory.”</p>
<h2 id="quantum-mechanical-cases"><a class="toclink" href="#quantum-mechanical-cases">Quantum-Mechanical Cases</a></h2>
<p>In chapter 6 (pp. 146–165), Ahmed assures us that <a href="https://www.semanticscholar.org/paper/Quantum-Mysteries-for-Anyone-Mermin/856ccd030d12d89f65e3b85f3430194cc4281ab4">Mermin (1981)</a> assured him that the following setup is realistic: There’s a source and two receivers A and B. In each run, the source emits two signals to the receivers. It has no switches and receives no signals. The receivers, however, have switches that you can set to 1, 2, or 3. The receivers can’t communicate with one another or, as mentioned, with the source.</p>
<p>The receivers have displays where they display “yes” or “no,” so the outcome of a run can be denoted as 12ny. What looks like a street in New York City, denotes that receiver A was set to 1, B to 2 and the readings on them were “no” and “yes.”</p>
<p>What’s surprising about the setup is that the following statistical observations obtain:</p>
<ol>
<li>Whenever the switches on A and B are on the same setting (i.e. both on 1, both on 2 or both on 3) the devices display the same reading: either both say ‘y’ or both say ‘n’. So we sometimes get runs like this: ‘11yy’, ‘22yy’. But we never get runs like this: ‘22yn’, ‘33ny’.</li>
<li>When the switches on A and B are on different settings (e.g. A on 1, B on 3), the devices display the same reading about 25 per cent of the time. So we get runs like ‘12yn’ and ‘23ny’ about three times as often as we get runs like ‘12yy’ and ‘13nn’.</li>
</ol>
<p>The first observation could obtain if the receivers just always displayed yy (or always displayed nn) or displayed something in accordance with a specification in the signal. But the second one requires that they switch outputs, and that’s where the behavior starts to look spooky in ways that are perhaps explained by books like <em><a href="https://www.math.columbia.edu/~woit/wordpress/?p=11128">Something Deeply Hidden</a></em> (though this particular device is not mentioned). Ahmed argues for this on page 158, citing John Stewart Bell.</p>
<p>In Ahmed’s book it is assumed that there is no causal connection here, in the intuitive sense of “causal”:</p>
<blockquote>
<p><strong>Causal independence</strong>: whatever happens at either receiver is causally independent of anything that happens at the other.</p>
</blockquote>
<p>This leads to a money pump against <span class="caps">CDT</span> that is based on <span class="caps">CDT</span>’s refusal to take the distribution of chance events into account because it is causally independent of one’s choices. The payoff table looks as follows where “het” is short for “ny or yn” and “hom” is short for “yy or nn.”</p>
<table>
<thead>
<tr>
<th style="text-align: left;"></th>
<th style="text-align: right;">yy</th>
<th style="text-align: right;">yn</th>
<th style="text-align: right;">ny</th>
<th style="text-align: right;">nn</th>
</tr>
</thead>
<tbody>
<tr>
<td style="text-align: left;">12hom</td>
<td style="text-align: right;">2</td>
<td style="text-align: right;">0</td>
<td style="text-align: right;">0</td>
<td style="text-align: right;">2</td>
</tr>
<tr>
<td style="text-align: left;">13hom</td>
<td style="text-align: right;">2</td>
<td style="text-align: right;">0</td>
<td style="text-align: right;">0</td>
<td style="text-align: right;">2</td>
</tr>
<tr>
<td style="text-align: left;">23hom</td>
<td style="text-align: right;">2</td>
<td style="text-align: right;">0</td>
<td style="text-align: right;">0</td>
<td style="text-align: right;">2</td>
</tr>
<tr>
<td style="text-align: left;">12het</td>
<td style="text-align: right;">0</td>
<td style="text-align: right;">1</td>
<td style="text-align: right;">1</td>
<td style="text-align: right;">0</td>
</tr>
<tr>
<td style="text-align: left;">13het</td>
<td style="text-align: right;">0</td>
<td style="text-align: right;">1</td>
<td style="text-align: right;">1</td>
<td style="text-align: right;">0</td>
</tr>
<tr>
<td style="text-align: left;">23het</td>
<td style="text-align: right;">0</td>
<td style="text-align: right;">1</td>
<td style="text-align: right;">1</td>
<td style="text-align: right;">0</td>
</tr>
</tbody>
</table>
<p>That is, there are no options with identical receiver settings, and a winning bet on any of the “hom” options pays twice as much as a winning bet on any of the “het” options.</p>
<p>For <span class="caps">EDT</span> the case is clear: “het” results are three times as likely to come up than “hom” results, so the doubled payoff does nothing to make bets on “hom” more attractive.</p>
<p>The derivation of the <span class="caps">CDT</span> recommendation spans pages 150 to 155 and is based on <span class="caps">CDT</span> foundations that I haven’t even introduced in this summary, so I’ll skip to the end: “<span class="caps">EDT</span> recommends only 12het, which gets a V-score 0.75 over 0.5 for 12hom. <span class="caps">CDT</span> endorses 12hom, which is getting some unknown U-score that is no less than U(12het).” (Ahmed chose the ”12” setting in a “without loss of generality” type of move.)</p>
<p>So again we have a case where <span class="caps">CDT</span> is neutral between options even though one (or one set) is intuitively better than another.</p>
<p>The objections section of this chapter (almost every chapter has one) is particularly interesting because it goes more in depth on the arcana of how different scholars have attempted to define what causality is. This is something I would’ve found interesting for every derivation of a <span class="caps">CDT</span> recommendation, though Ahmed probably omitted them in many cases because the leading definitions don’t disagree on the recommendation.</p>
<p>Ahmed lists the definitions or accounts of causality of <a href="https://www.google.com/books/edition/_/LTaU5JTj3mUC">Reichenbach</a>, <a href="https://www.google.com/books/edition/A_Probabilistic_Theory_of_Causality/Ff4HAQAAIAAJ">Suppes</a>, <a href="https://www.google.com/books/edition/A_Treatise_of_Human_Nature/a-rWAAAAMAAJ">Hume</a>, <a href="https://www.google.com/books/edition/A_System_of_Logic_Ratiocinative_and_Indu/y4MEAAAAQAAJ">Mill</a>, and <a href="https://books.google.com/books?id=KhJ99xSUkPAC">Mackie</a> as ones that support a view according to which the correlation between the outputs of the receivers could be called causal – but only if we ignore the requirement of the first four definitions that the cause needs to occur prior to the effect in sidereal time or, in Mackie’s case, logical time, and doesn’t require faster-than-light propagation (p. 159). <a href="https://doi.org/10.1086/289652">Lewis</a>’s definition makes it plausible with fewer reservations (ibid.). Under these definitions, <span class="caps">CDT</span> will make the same recommendation as <span class="caps">EDT</span>.</p>
<p><a href="https://www.google.com/books/edition/Scientific_Explanation_and_the_Causal_St/2ug9DwAAQBAJ">Salmon</a>, <a href="http://web.mit.edu/bskow/www/215-S12/price-menzies-cause-as-secondary-quality.pdf">Menzies and Price</a>, <a href="https://www.google.com/books/edition/Making_Things_Happen/vvj7frYow6IC">Woodward</a>, and <a href="https://doi.org/10.1023/A:1005281714660">Hausman</a> have proffered definitions according to which there is no such causal loop between the receivers (ibid.).</p>
<p><a href="https://doi.org/10.1111/j.1475-4975.1984.tb00062.x">Skyrms</a>’s “family resemblance” theory of causation is silent on the issue (p. 161).</p>
<p><span class="caps">CDT</span>’s neutrality between the options seemed to me to imply that any <span class="caps">CDT</span> agent with some nonzero credence in the hypothesis that there may be something causal going on between the receivers will have a tiebreaker in favor of the <span class="caps">EDT</span> recommendation. That there is some normative uncertainty about what causality really is – if there’s a fact of the matter – seemed to imply to me that <span class="caps">CDT</span> agents would typically have such uncertainty.</p>
<p>But to the contrary, Ahmed shows in chapter 6.5.2 that even in that case, the payoffs of the “hom” options can always be adjusted such that a <span class="caps">CDT</span> agent with any degree of uncertainty will (also) endorse the option that <span class="caps">EDT</span> recommends against.</p>
<h3 id="comment_3"><a class="toclink" href="#comment_3">Comment</a></h3>
<p>The device is realistic, so Ahmed even offers a challenge:</p>
<blockquote>
<p>On your turns, I pay you what you win. On my turns, you pay me what I win. So if I follow <span class="caps">EDT</span> and you follow <span class="caps">CDT</span> then I will on average win one dollar from you every four runs. I hereby publicly challenge any defender of <span class="caps">CDT</span> who accepts [the causal independence premise] to play this game against me.</p>
</blockquote>
<p>This seems risky. A defender of <span class="caps">CDT</span> can pretend to just “arbitrarily” choose the more lucrative option at every turn. Since <span class="caps">CDT</span> is still the dominant view among decision theorists, they can thereby effectively filibuster Ahmed and slow or reduce the chance of a development away from the status quo. (Not that anyone would actually want to do that.)</p>
<p>Page 162 has a very interesting footnote that I’ll reproduce here in full since I’ve not yet followed up on it, i.e. read at least Hilary Greaves’s paper. (Links are mine.)</p>
<blockquote>
<p>I have in this discussion altogether ignored the ‘many-worlds’ theory according to which all possible results of any run appear in one of the branches into which the world divides after the readings have been taken (Blaylock 2010: 116–17). One reason for this is that it is hard to see what any decision theory should say about a case in which the uncertainty of each run gets resolved in different ways according to equally real future counterparts of the experimenter: in particular, it is not clear to me why the ‘weights’ that the many-worlds theory attaches to each branch should play the decision-theoretic role that subjective uncertainties play in cases of ‘single branch’ decisions under uncertainty (for reasons arising from Price’s (2010) comments on Wallace 2007 and <a href="http://philsci-archive.pitt.edu/3103/2/pitei.pdf">Greaves 2007</a>). But in any case and as section 6.5.2 argues, anyone who gives the many-worlds interpretation some credence should still agree that <span class="caps">EPR</span> drives <span class="caps">EDT</span> and <span class="caps">CDT</span> apart, as long as he also gives some credence to the ‘single-world’ proposal on which the receivers are causally independent of one another.</p>
</blockquote>
<p>This seems to contradict my treatment of measure in ethical decision making, which I’ve so far treated analogously to probability.</p>
<h2 id="newcomb-again"><a class="toclink" href="#newcomb-again">Newcomb Again</a></h2>
<p>At this stage, Ahmed has shown that there are various realistic cases in which the recommendations of <span class="caps">CDT</span> and <span class="caps">EDT</span> diverge. So the disagreement cannot just be due to intuitions that are badly calibrated in situations that are impossible in the real world. This is bad news for <span class="caps">EDT</span>, which would’ve otherwise, according to Ahmed, been preferable on parsimony grounds. Or it is bad news for <span class="caps">EDT</span> unless it can be shown that <span class="caps">EDT</span> actually makes the better recommendations in the divergent cases. Then it may even be good news, since that would constitute an even stronger case for <span class="caps">EDT</span> than the one from parsimony alone.</p>
<p>In the first section of the chapter, Ahmed reiterates this reasoning and explains it in some more depth, and in the second, he defuses the “discontinuous strategy” according to which you one-box only when the predictor is absolutely infallible and two-box otherwise, even for (merely) arbitrarily reliable fallible predictors (pp. 170–179).</p>
<h3 id="the-case-for-one-boxing"><a class="toclink" href="#the-case-for-one-boxing">The Case for One-Boxing</a></h3>
<p>Section 3 focuses on the case for one-boxing and in particular on the “Why Ain’cha Rich?” (<span class="caps">WAR</span>) argument (pp. 182–183, I’ve changed the numbering and removed internal references):</p>
<ol>
<li>The expected return to one-boxing exceeds that to two-boxing.</li>
<li>Everyone agrees on 1 (premise).</li>
<li>Therefore one-boxing foreseeably does better than two-boxing.</li>
<li>Therefore <span class="caps">CDT</span> is committed to the foreseeably worse option for anyone facing Newcomb’s problem.</li>
<li>Therefore <span class="caps">CDT</span> gives the wrong recommendation in the standard Newcomb problem.</li>
</ol>
<p>The main contentious point about <span class="caps">WAR</span> is whether conclusion really follows from statement 4.</p>
<p>One such challenge is due to <a href="https://www.google.com/books/edition/The_Foundations_of_Causal_Decision_Theor/LYTMhPzCUxYC">Joyce (1999)</a>, in which he reframes the Newcomb problem as one in which the agent believes to be the “type” of person who two-boxes and believes that this determined the predictor’s decision. This, however, is the type of problem where the Tickle Defense is applicable and <span class="caps">CDT</span> and <span class="caps">EDT</span> would both recommend two-boxing. (Also if you believe yourself the type of person who one-boxes, because then you can get a thousand dollars extra by two-boxing.) Joyce’s chapter in <a href="https://doi.org/10.1017/9781316847893">Ahmed (2018)</a> seems to me like an improved version of this argument.</p>
<p>Next, on pages 187–194, Ahmed addresses an objection due to <a href="https://doi.org/10.1007/s10670-007-9084-8">Arntzenius (2008)</a>:</p>
<blockquote>
<p><strong>Yankees v Red Sox</strong>: The Yankees and the Red Sox are going to play a lengthy sequence of games. The Yankees win 90 per cent of such encounters. Before each game Mary has the opportunity to bet on either side. [The payoffs are as in the table below.] Just before each bet, a perfect predictor tells her whether her next bet is going to be a winning bet or a losing bet. Mary knows all this.</p>
</blockquote>
<table>
<thead>
<tr>
<th></th>
<th style="text-align: right;">Red Sox win</th>
<th style="text-align: right;">Yankees win</th>
</tr>
</thead>
<tbody>
<tr>
<td>Bet on Red Sox</td>
<td style="text-align: right;">2</td>
<td style="text-align: right;">−1</td>
</tr>
<tr>
<td>Bet on Yankees</td>
<td style="text-align: right;">−2</td>
<td style="text-align: right;">1</td>
</tr>
</tbody>
</table>
<p>Here, Arntzenius argues, <span class="caps">EDT</span> is doomed to always bet on the Red Sox and lose 90% of the time, because if the predictor says that they’ll win their next bet, they’ll win most by betting on the Red Sox and if the predictor says they’ll lose their next bet, they’ll lose least if they also bet on the Red Sox. This would constitute a parallel case to <span class="caps">WAR</span> where <span class="caps">EDT</span> fails to get rich. (<a href="https://www.youtube.com/watch?v=c2rZTzrn18s">Here a video in case you’re unfamiliar with the sport.</a>)</p>
<p>Ahmed objects that it’s not so much that Arntzenius’s analysis of what <span class="caps">EDT</span> would do in this situation is mistaken but that the argument is not parallel to (or a mirror image of) <span class="caps">WAR</span>. That seems true enough given the particular syllogisms Ahmed uses, but I have the impression that the more general point stands: The agent who’s been using <span class="caps">CDT</span> can still ask the agent who’s been using <span class="caps">EDT</span>: Why ain’tcha rich?</p>
<h4 id="comment_4"><a class="toclink" href="#comment_4">Comment</a></h4>
<h5 id="edt-money-pump"><a class="toclink" href="#edt-money-pump"><span class="caps">EDT</span> Money Pump</a></h5>
<p>Ahmed argues that Yankees vs. Red Sox is parallel to Check-up (p. 189):</p>
<blockquote>
<p><strong>Check-up</strong>: Every Monday morning everyone has the chance to pay one dollar for a medical check-up at which the doctor issues a prescription if necessary. Weeks in which people take this opportunity are much more likely than other weeks to be weeks in which they fall ill. In fact on average, 90 per cent of Mondays on which someone does have a check-up fall in weeks when he or she is ill. And only 10 per cent of Mondays on which someone doesn’t go for a check-up fall in weeks when he or she is ill. There is nothing surprising or sinister about this. It is just that one is more likely to go for a check-up when one already has reason to think that one will fall ill. \</p>
<p>…</p>
<p>All weekend you have suffered from fainting and dizzy spells. You’re pretty sure that something’s wrong. Should you go for the check-up on Monday morning? Clearly if you are ill this week, it will be better to have the prescription than not, so the check-up will have been worth your while. But if you are not ill this week then the check-up will have been a waste of time and money.</p>
</blockquote>
<p>The payoff table:</p>
<table>
<thead>
<tr>
<th style="text-align: left;"></th>
<th style="text-align: right;">Well this week</th>
<th style="text-align: right;">Ill this week</th>
</tr>
</thead>
<tbody>
<tr>
<td style="text-align: left;">Check-up</td>
<td style="text-align: right;">1</td>
<td style="text-align: right;">0</td>
</tr>
<tr>
<td style="text-align: left;">No check-up</td>
<td style="text-align: right;">2</td>
<td style="text-align: right;">−1</td>
</tr>
</tbody>
</table>
<p>In Check-up it is clear that one can not just calculate the expected return of going and not going to the check-up given the population-wide correlation statistics. The fainting and dizzy spells put you in the reference class of people who are very likely sick, so unless almost the whole population is sick, these broad statistics are uninformative. Rather you can be almost certain that you fall into the “ill this week” group, in which case going to the check-up is the better option.</p>
<p>Ahmed suggests that (1) the population-wide statistics are the equivalent of the long-run Yankees vs. Red Sox statistics, that (2) the fainting and dizzy spells are the equivalent of the predictor, and that (3) the decision is the equivalent of the bet. So there are three variables, as it were.</p>
<p>Ahmed does not consider the population-level statistics to be variable, hence the “as it were.” Whether you’re sick or not and how you decide do not influence these statistics in the limit. In footnote 41 he explicitly says: “So even if she learns that she will win her next bet, is she not still entitled to be as confident in [the statistics] as she was before? … But the point is not that Mary’s information makes [the statistics] false but that [the statistics] no longer [entail] anything about what she should now do. Certainly her next bet belongs to a population of bets of which [the statistics are] true.” That makes a lot of sense.</p>
<p>But early signs of sickness are not sensitive to whether we later go to a doctor. A prediction, given outcomes of games that are held constant, is sensitive to how we bet. So we still end up with two worrisome outcomes:</p>
<ol>
<li>An <span class="caps">EDT</span> agent will have a predictor who predicts a series of 90% losses. The <span class="caps">EDT</span> agent will – <span class="caps">EDT</span>-rationally, as Ahmed shows – bet on the Red Sox to minimize the loss.</li>
<li>A <span class="caps">CDT</span> agent will have a predictor who predicts 90% wins. The <span class="caps">CDT</span> agent will – ignoring the predictor – bet on the Yankees to maximize unconditional expected value.</li>
</ol>
<p>So the <span class="caps">EDT</span> agent predictably loses money while the <span class="caps">CDT</span> agent predictably wins money, and the <span class="caps">CDT</span> agent is in a perfect position to ask, “Why ain’tcha rich?” All the <span class="caps">EDT</span> agent would have to do is to ignore the prediction.</p>
<p>But I still don’t think Yankees vs. Red Sox is parallel to Newcomb’s problem: The predictor in Newcomb’s problem only predicts your actions. The predictor in Yankees vs. Red Sox predicts the outcome of the game and your bet, two independent events. So in the latter case these two can clash, while in the first case inconsistent scenarios are simply impossible without paradox. (A feature that, to my understanding, <span class="caps">MIRI</span>-style decision theories try to capitalize on.)</p>
<p>Consider a robot with no concept of decision theory, money, or utility. The robot follows the simple algorithm to only bet on the Yankees (without loss of generality) unless the prediction is that it’ll lose, in which case it bets on the Red Sox. I’ll call such an agent a “Yankee fan” because they want the Yankees to win every game: The Yankee fan makes worlds in which the Yankees lose inconsistent. Or if the world is malleable, then being a Yankee fan is a way of making sure that the Yankees never lose. (You could get their coach to pay you much more than the payoffs from the bets to follow this algorithm.) But in the following I will assume that the outcome of the game is completely independent.</p>
<p>Say the predictor runs simulations to compute its predictions: At one point it’ll predict that the Yankees will lose. The predictor will run a simulation of the Yankee fan in which they announce that the Yankee fan will win the next bet. They’ll bet on the Yankees. This case is inconsistent. Then they’ll run a simulation in which they tell the Yankee fan that they’ll lose the next bet. They’ll bet on the Red Sox. This case, too, is inconsistent.</p>
<p>But this behavior can be fixed. In a fixed version of the problem, the predictor can choose from three pronouncements: “You’ll win,” “You’ll lose,” and “I don’t know.”</p>
<p>These are the bets of the agents conditional on the predictions:</p>
<table>
<thead>
<tr>
<th></th>
<th style="text-align: right;">You’ll win</th>
<th style="text-align: right;">You will lose</th>
<th style="text-align: right;">I don’t know</th>
</tr>
</thead>
<tbody>
<tr>
<td>Yankee bettor</td>
<td style="text-align: right;">Yankees</td>
<td style="text-align: right;">Yankees</td>
<td style="text-align: right;">Yankees</td>
</tr>
<tr>
<td>Red Sox bettor</td>
<td style="text-align: right;">Red Sox</td>
<td style="text-align: right;">Red Sox</td>
<td style="text-align: right;">Red Sox</td>
</tr>
<tr>
<td>Yankee fan</td>
<td style="text-align: right;">Yankees</td>
<td style="text-align: right;">Red Sox</td>
<td style="text-align: right;">Red Sox</td>
</tr>
<tr>
<td>Red Sox fan</td>
<td style="text-align: right;">Red Sox</td>
<td style="text-align: right;">Yankees</td>
<td style="text-align: right;">Yankees</td>
</tr>
</tbody>
</table>
<p>The predictor always (1) simulates the game to determine the winner; (2) runs two simulations in which it announces that the agent will win and lose respectively; and finally (3) announces, in the real world, <em>the</em> overall consistent prediction, <em>one of them</em> if both are consistent, or <em>“I don’t know”</em> if neither is consistent.</p>
<table>
<thead>
<tr>
<th></th>
<th style="text-align: right;">Yankees win</th>
<th style="text-align: right;">Red Sox win</th>
</tr>
</thead>
<tbody>
<tr>
<td>Yankee bettor</td>
<td style="text-align: right;">You’ll win</td>
<td style="text-align: right;">You’ll lose</td>
</tr>
<tr>
<td>Red Sox bettor</td>
<td style="text-align: right;">You’ll lose</td>
<td style="text-align: right;">You’ll win</td>
</tr>
<tr>
<td>Yankee fan</td>
<td style="text-align: right;">Either</td>
<td style="text-align: right;">I don’t know</td>
</tr>
<tr>
<td>Red Sox fan</td>
<td style="text-align: right;">I don’t know</td>
<td style="text-align: right;">Either</td>
</tr>
</tbody>
</table>
<p>Here it becomes clearer why (the the first table) the Yankee fan bets on the Red Sox in the “I don’t know” case (and vice versa for the Red Sox fan). The “I don’t know” case is the one where that team wins. Their chagrin over the lost game will be alleviated by the elation over the won bet. This results in the following expected payoffs:</p>
<table>
<thead>
<tr>
<th></th>
<th style="text-align: right;">Yankees win</th>
<th style="text-align: right;">Red Sox win</th>
</tr>
</thead>
<tbody>
<tr>
<td>Yankee bettor</td>
<td style="text-align: right;">0.9</td>
<td style="text-align: right;">−0.2</td>
</tr>
<tr>
<td>Red Sox bettor</td>
<td style="text-align: right;">−0.9</td>
<td style="text-align: right;">0.2</td>
</tr>
<tr>
<td>Yankee fan</td>
<td style="text-align: right;">0.9 or −0.9</td>
<td style="text-align: right;">0.2</td>
</tr>
<tr>
<td>Red Sox fan</td>
<td style="text-align: right;">0.9</td>
<td style="text-align: right;">0.2 or −0.2</td>
</tr>
</tbody>
</table>
<ol>
<li>The Yankee bettor makes 0.7 in expectation.</li>
<li>The Red Sox bettor makes −0.7 in expectation.</li>
<li>The Yankee fan makes between −0.7 and 1.1 in expectation depending on whether the predictor chooses to announce that they’ll win, lose, or some randomized answer in the case where both are consistent.</li>
<li>The Red Sox fan makes between 0.7 and 1.1 in expectation for the same reason.</li>
</ol>
<p>So being a Red Sox fan seems best overall.</p>
<p>Note that the predictor needn’t always run two simulations. It could also run the second simulation only when the first one led to an inconsistent case. This would only change the outcome of the “Either” cases in that they’ll be “You’ll win” or “You’ll lose” depending on which the predictor tests first.</p>
<p>All in all, Yankees vs. Red Sox is similar to the transparent Newcomb problem (as well as the standard version) because that one also has inconsistencies for similar reasons that need to be fixed. But even that likeness seems imperfect to me. The problems behave quite differently (e.g., the predictor is more constrained in Yankees vs. Red Sox if it wants to avoid prediction errors). Or maybe I just haven’t thought of the right way to map one onto the other.</p>
<p><a href="https://impartial-priorities.org/thoughts-on-the-transparent-newcombs-problem.html">In a related post on my blog</a> I speculate a bit about how such simulation schemes might work and what that says about precommitments and updateless behavior of <span class="caps">CDT</span> and <span class="caps">EDT</span> in the transparent Newcomb problem.</p>
<h5 id="better-irrational-than-crazy"><a class="toclink" href="#better-irrational-than-crazy">Better Irrational Than Crazy</a></h5>
<p>On a lighter note: <a href="https://doi.org/10.1007/s10670-007-9084-8">Arntzenius (2008)</a>, where Yankees vs. Red Sox originates, argues for Piaf’s maxim, which Ahmed refutes in chapter 3. Arntzenius concludes the section with a mention of the transparent Newcomb’s problem:</p>
<blockquote>
<p>Consider again a Newcomb situation. Now suppose that the situation is that one makes a choice after one has seen the contents of the boxes, but that the predictor still rewards people who, insanely, choose only box A even after they have seen the contents of the boxes. What will happen? Evidential decision theorists and causal decision theorists will always see nothing in box A and will always pick both boxes. Insane people will see $10 in box A and $1 in box B and pick box A only. So insane people will end up richer than causal decision theorists and evidential decision theorists, and all hands can foresee that insanity will be rewarded. This hardly seems an argument that insane people are more rational than either of them are. Let me turn to a better argument against causal decision theory.</p>
</blockquote>
<p>History repeats itself. It’s ironic how parallel this “argument” is to the one that the standard Newcomb’s problem merely rewards irrationality. I, for one, am happy to be irrational and insane in <em>that</em> sense if Omega pays me enough.</p>
<p>Ahmed also rejects one-boxing in the transparent Newcomb’s problem and argues that this says nothing about the applicability of <span class="caps">WAR</span> to the standard case:</p>
<blockquote>
<p>A similar point applies to the similar objection to <span class="caps">WAR</span> that Gibbard and Harper (1978: 369–70) based on a variant of the Newcomb case in which both boxes are transparent. ‘Most subjects find nothing in the first [i.e. what used to be the opaque] box and then take the contents of the second box. Of the million subjects tested, 1% have found 1 million dollars in the first box, and strangely enough only 1% of those … have gone on to take the 1,000 dollars they could each see in the second box.’ Everyone should agree (certainly <span class="caps">EDT</span> and <span class="caps">CDT</span> agree) that it would be rational to take both boxes in this case, whatever you see in them. But according to Gibbard and Harper (1978: 371), <span class="caps">WAR</span> recommends only taking one box in this modified version of the Newcomb problem. But if it goes wrong here – the argument runs – then there must have been something wrong with it as applied to the standard Newcomb case. But that is not so: the reason it goes wrong here is that the agent has information that puts him not only in the class of people who face this case, within which indeed those who take just one box do better on average, but also within the narrower class of people who face this case and can see 1 million dollars in the first box, if that is what he sees. Within that narrower class, the expected return to two-boxing exceeds that to one-boxing by 1,000 dollars. Similarly if the agent is in the narrower class of people who can already see that the first box is empty. But the fact that nothing like <span class="caps">WAR</span> applies in this modified Newcomb case doesn’t stop it applying to the standard Newcomb case, in which no more specific information makes the statistically superior overall performance of one-boxing irrelevant to you.</p>
</blockquote>
<p>As of 2018, he doesn’t seem to have quite warmed up to updateless behavior yet. Ahmed in the introduction to his 2018 book <em><a href="https://doi.org/10.1017/9781316847893">Newcomb’s Problem</a></em> (he is the editor and contributed this introduction and one chapter):</p>
<blockquote>
<p>The people who do typically end up rich are the very few who decline to take both boxes even in this case. That seems insane, but these “lunatics” could retort, to the rest of us who have faced this problem: “If you’re so smart, why aren’t you rich?” (For further discussion, see the chapters by Ahmed and Greene.)</p>
</blockquote>
<h3 id="the-case-for-two-boxing"><a class="toclink" href="#the-case-for-two-boxing">The Case for Two-Boxing</a></h3>
<p>This subsection doesn’t so much present completely new arguments for two-boxing than provide more detail on ones that were mentioned throughout the book.</p>
<p>It starts out with the dominance reasoning that I already summarized in the section Newcomb’s Problem. Then it describes something called “information dominance”:</p>
<blockquote>
<p>A related but distinct argument for two-boxing imagines that your accomplice has seen what the predictor put in the opaque box and is willing to tell you what it is, perhaps for a fee not exceeding 1,000 dollars. If on receiving the fee he tells you that the predictor put 1 million dollars in the opaque box, your best response is clearly to take both boxes, and both <span class="caps">EDT</span> and <span class="caps">CDT</span> will recommend that. And if he tells you that the predictor put nothing in the opaque box, your best response is again to take both boxes, as <span class="caps">EDT</span> and <span class="caps">CDT</span> again agree. So why bother paying the fee? If you know that whatever extra information you are about to get, it will, on getting it, be rational for you to two-box, isn’t it already rational to two-box before you get it? In fact, isn’t it rational to do so even if there is no accomplice?</p>
</blockquote>
<p>Ahmed argues that it’s a nonsequitur – i.e. the version with the accomplice and without the accomplice are different problems: “So there is nothing wrong with taking different acts to be rational <em>ex post</em> and <em>ex ante</em>.” </p>
<p>That seems true enough since this problem, too, is equivalent to the transparent Newcomb’s problem. But see <a href="https://impartial-priorities.org/thoughts-on-the-transparent-newcombs-problem.html">this related post of mine</a> on why I think <span class="caps">EDT</span> may still one-box even with accomplice or transparent boxes so long as it has a choice to precommit.</p>
<p>The last reason to two-box is “Could have done better” (p. 202):</p>
<blockquote>
<p>If you know that a certain available option makes you worse off, given your situation, than you would have been on some identifiable alternative, then that first option is irrational.</p>
</blockquote>
<p>I don’t see how this criterion of rightness is different from Piaf’s maxim. Ahmed’s argumentation follows a different structure (if you’re interested in his analysis of sequential choice, see pp. 201–210), but the upshot is the same, namely that this maxim or criterion is a reason to two-box in Newcomb’s problem but that <span class="caps">CDT</span> and <span class="caps">EDT</span> likewise violate it in different cases so that it is no grounds on which to prefer <span class="caps">CDT</span> over <span class="caps">EDT</span>.</p>
<h2 id="the-ultimate-contingency"><a class="toclink" href="#the-ultimate-contingency">The Ultimate Contingency</a></h2>
<p>This is a wonderfully opaque title for a chapter that I’ll leave untouched in this summary as a teaser for you to buy the book.</p>
<p>My other reason for omitting it is that it covers a, to me, highly unintuitive and unparsimonious thesis and concludes that the thesis is, in fact, implausible. It leads to <span class="caps">CDT</span> and <span class="caps">EDT</span> violating such useful properties as dynamic consistency, i.e., that they don’t predictably change their strategy halfway through a multistage decision problem.</p>
<h2 id="conclusion"><a class="toclink" href="#conclusion">Conclusion</a></h2>
<p>I’ve found <em><a href="https://smile.amazon.com/Evidence-Decision-Causality-Arif-Ahmed-ebook/dp/B00KL8CK86/">Evidence, Decision, and Causality</a></em> to be very insightful and engagingly written. 4.7 out of 5 stars. Sadly, I’m not eligible to write reviews on Amazon.</p>
<p>There are many exciting areas of decision theory that are outside the purview of the book. That may be a reason not to read it, but the book makes it clear from the start what its project is and that it is not intended as a textbook on decision theory. For a non-textbook, though, it’s surprisingly didactic.</p>
<p>In particular its explanation of the Tickle Defense was the most lucid one that I could find. Many other summaries of it that I had read before were very condensed and left many reservations I had unanswered. Ahmed’s explanation answered almost all of them.</p>
<p>My only remaining reservations are that I’m unsure whether the book still reflects the latest knowledge of the author since it’s already six years old and that I wasn’t able to find an errata online.</p>
<h2 id="footnotes"><a class="toclink" href="#footnotes">Footnotes</a></h2>
<div class="footnote">
<hr>
<ol>
<li id="fn:plates-vs-bowls">
<p>For example, two people may argue whether plates are better than bowls. One of them may argue that bowls can’t hover in the air while you’re eating. Now the argument may turn on whether this is a reasonable demand, because if it is not reasonable – e.g., if you can always find some surface to put your bowl on to eat or you can always eat with one hand leaving the other to hold the bowl – then bowls may be no worse than plates. But the other party can also sidestep that argument by instead arguing that plates can’t hover in the air either, which works regardless of how reasonable of a demand this hovering ability is. <a class="footnote-backref" href="#fnref:plates-vs-bowls" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
<li id="fn:alterations">
<p>In almost all longer citations I’ve made such alterations as transposing tables, expanding abbreviations, and correcting typos without highlighting them. I’m confident that none of these changed the meaning of the sentences. <a class="footnote-backref" href="#fnref:alterations" title="Jump back to footnote 2 in the text">↩</a></p>
</li>
<li id="fn:textbook">
<p>But you may want to heed the warning of the author that “because of this aim the book couldn’t serve as a stand-alone introduction to decision theory. Anyone who took it that way would find it distorted and lacunary. Distorted because of the focus on the dominance principle at the expense of almost everything else that Savage’s axioms entail; lacunary because of the complete absence from the story of any approach outside the Ramsey–Savage expected-utility paradigm.” <a class="footnote-backref" href="#fnref:textbook" title="Jump back to footnote 3 in the text">↩</a></p>
</li>
<li id="fn:combs">
<p>If you’re confused about the name “Newcomb’s problem,” imagine all the new combs you can buy from all that money! <a class="footnote-backref" href="#fnref:combs" title="Jump back to footnote 4 in the text">↩</a></p>
</li>
<li id="fn:transparent-boxes">
<p>Clearly, we need a lot more progress in manufacturing technology to produce empty <a href="https://smile.amazon.com/OfficeGoods-accesorios-contenedor-almacenamiento-exhibici%C3%B3n/dp/B07RWTQQ5P/">transparent boxes that don’t themselves cost some $20–30</a>. <a class="footnote-backref" href="#fnref:transparent-boxes" title="Jump back to footnote 5 in the text">↩</a></p>
</li>
<li id="fn:prisoners-dilemma">
<p>I don’t suppose the naming makes a difference, right? The first version focuses on one of two unnamed prisoners in a parallel situation, so there’s no loss of generality, just like in Ahmed’s version that is explicitly about both prisoners. Sadly, that doesn’t make it any easier to decide on one spelling, so I’ve gone with the more common one. <a class="footnote-backref" href="#fnref:prisoners-dilemma" title="Jump back to footnote 6 in the text">↩</a></p>
</li>
</ol>
</div>Thoughts: The Transparent Newcomb’s Problem2020-08-18T22:05:00+00:002020-08-18T22:05:00+00:00Dawn Dreschertag:impartial-priorities.org,2020-08-18:/thoughts-on-the-transparent-newcombs-problem.html<p>Writing <a href="https://impartial-priorities.org/summary-of-evidence-decision-and-causality.html">my summary of <em>Evidence, Decision, and Causality</em></a>, I got interested in how <span class="caps">EDT</span> might be able to succeed by precommitting, and how different simulation schemes that a predictor might run of the transparent Newcomb’s problem might affect the way <span class="caps">EDT</span> and <span class="caps">CDT</span> reason about the problem. Precommitments hinge on an aspect of sophisticated choice that I haven’t been able to find information on. Indexical (or anthropic) uncertainty seems to do the trick unambiguously. Be warned that these things were probably only new to me, so if you know some decision theory you may end up bored, and I wouldn’t want that.</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#precommitments">Precommitments</a></li>
<li><a href="#updatelessness">Updatelessness</a><ul>
<li><a href="#some-observations">Some Observations</a></li>
<li><a href="#simulation-regimes">Simulation Regimes</a><ul>
<li><a href="#2-sim">2 Sim.</a></li>
<li><a href="#15-sim-consistent">1.5 Sim. (Consistent)</a></li>
<li><a href="#1-sim">1 Sim.</a></li>
</ul>
</li>
<li><a href="#reasoning-under-indexical-uncertainty">Reasoning Under Indexical Uncertainty</a></li>
</ul>
</li>
<li><a href="#notes">Notes</a></li>
</ul>
</div>
<h2 id="precommitments"><a class="toclink" href="#precommitments">Precommitments</a></h2>
<p>In the transparent Newcomb’s problem, <span class="caps">CDT</span> and <span class="caps">EDT</span> are both said to two-box and fail to become millionaires. Most decision theorists would call me insane (or maybe they’d be more polite about their disapproval) for wanting to one-box in this situation, but maybe I just really like money.</p>
<p>So I wonder whether I need to adopt a <span class="caps">UDT</span>-like decision theory to achieve this or whether I have other options, in particular options compatible with following the advice of <span class="caps">EDT</span>.</p>
<p><a href="https://casparoesterheld.com/2016/11/21/thoughts-on-updatelessness/">Friends of mine said that <span class="caps">EDT</span> would update on the nonoptimality of its approach and choose to precommit.</a> For that to work out, it’ll need (1) a full understanding of the choice situation and (2) a chance to precommit. It’ll need both of these before learning some crucial information, in this case the contents of the boxes. Thereby it would, in effect, implement updateless behavior in individual choice situations.</p>
<p>But I somehow can’t make that happen. I turn it into a sequential choice problem by adding a chance to precommit when the agent knows the choice situation but hasn’t seen the (transparent) boxes. To precommit, the agent puts $1,001 in escrow, which they’ll only get back if they one-box.</p>
<p>I’ll write M and K for the potential money under the boxes ($1,000,000 and $1,000 respectively) and E = $1,001 for the potential amount in escrow. The predictor is known to be perfect, so some cases are impossible. First the <strong>myopic case</strong>:</p>
<table>
<thead>
<tr>
<th style="text-align: left;"></th>
<th style="text-align: right;">Pred. to one-box</th>
<th style="text-align: right;">Pred. to two-box</th>
</tr>
</thead>
<tbody>
<tr>
<td style="text-align: left;">Escrow, one-box</td>
<td style="text-align: right;">M</td>
<td style="text-align: right;">–</td>
</tr>
<tr>
<td style="text-align: left;">Escrow, two-box</td>
<td style="text-align: right;">–</td>
<td style="text-align: right;">K - E = −$1</td>
</tr>
<tr>
<td style="text-align: left;">No escrow, one-box</td>
<td style="text-align: right;">M</td>
<td style="text-align: right;">–</td>
</tr>
<tr>
<td style="text-align: left;">No escrow, two-box</td>
<td style="text-align: right;">–</td>
<td style="text-align: right;">K</td>
</tr>
</tbody>
</table>
<p>Myopic <span class="caps">EDT</span> may endorse escrow but there’s nothing here that makes it recommend it over no escrow. The case where the agent gets the million is, by construction, identical to the case where they get the million and get the escrow money back. In practice, precommitments will come with small costs – transaction costs, lost interest, option value, vel sim. – but may still be very much worth it. Yet myopic <span class="caps">EDT</span> would then no longer even endorse them.</p>
<table>
<thead>
<tr>
<th style="text-align: left;"></th>
<th style="text-align: right;">Pred. to one-box</th>
<th style="text-align: right;">Pred. to two-box</th>
</tr>
</thead>
<tbody>
<tr>
<td style="text-align: left;">Escrow, one-box</td>
<td style="text-align: right;">M</td>
<td style="text-align: right;">–</td>
</tr>
<tr>
<td style="text-align: left;">No escrow, two-box</td>
<td style="text-align: right;">–</td>
<td style="text-align: right;">K</td>
</tr>
</tbody>
</table>
<p>The <strong>sophisticated choice</strong> situation is unclear to me. The materials on decision theory that I’ve read didn’t feature a case where the likely payoffs of the final stage of a multi-stage decision problem varied depending on decisions at earlier stages. Usually the preferences changed or nothing changed. <span class="caps">EDT</span> normally recommends two-boxing if it encounters the second stage in isolation. If that recommendation were the starting point of the backward induction, it would make the escrow seem highly undesirable.</p>
<p>But at least intuitively it seems unproblematic to me (without a textbook to back me up, I’m afraid I may be overlooking some incoherency) to take the bad news into account that two-boxing confers when it comes to the then-empty box. I just don’t know if this is in the spirit of sophisticated choice or already touches on policy selection. But if this is compatible with how decision theorists typically understand sophisticated choice and <span class="caps">EDT</span> to operate, then <span class="caps">EDT</span> recommends escrow. The table above is the result of eliminating “escrow, two-box” because two-boxing at the final stage would violate the agent’s preferences, and eliminating “No escrow, one-box” for the same reason and the reason that the actual preferences would make the case inconsistent. What remains is for the agent to choose “escrow, one-box.” Yay! I just hope this way of reasoning is consistent.</p>
<p>Resolute choice solves the issue, but then resolute choice doesn’t require any separate precommitment stage to begin with.</p>
<h2 id="updatelessness"><a class="toclink" href="#updatelessness">Updatelessness</a></h2>
<p>There is also the consideration that such thorough reasoners as these idealized agents will consider that they have some degree of indexical (or anthropic<sup id="fnref:1"><a class="footnote-ref" href="#fn:1">1</a></sup>) uncertainty. The predictor may determine the agent’s bet conditional on the prediction by running a simulation of them, and the agent may currently be inside that simulation. This consideration would make both <span class="caps">EDT</span> and <span class="caps">CDT</span> updateless and thereby rich.</p>
<p>For concision I’ll write “one-boxer” for “an agent that has been found to de facto follow the one-boxer policy.” Calling the agent a one-boxer makes it sound like this is an indelible property of the agent rather than an observation. This would be problematic in the standard Newcomb’s problem since such an inherent property is likely to be known to the agent themselves, and then the Tickle Defense would apply to the resulting problem. Here I intend “one-boxer” etc. as mere shorthands without all these implications.</p>
<p>In the transparent Newcomb’s problem, the agents can update on the payoffs in both boxes, so there are four instead of two policies (see the table below):</p>
<ol>
<li>One-boxer: Take one box (the formally opaque one) no matter what.</li>
<li>Two-boxer: Take both boxes no matter what.</li>
<li>Conformist: One-box if there is a million in the box (M, K); otherwise two-box.</li>
<li>Rebel: One-box if the box is empty (0, K); otherwise two-box.</li>
</ol>
<table>
<thead>
<tr>
<th style="text-align: left;"></th>
<th style="text-align: right;">M, K</th>
<th style="text-align: right;">0, K</th>
</tr>
</thead>
<tbody>
<tr>
<td style="text-align: left;">One-boxer</td>
<td style="text-align: right;">M</td>
<td style="text-align: right;">0</td>
</tr>
<tr>
<td style="text-align: left;">Two-boxer</td>
<td style="text-align: right;">M+K</td>
<td style="text-align: right;">0+K</td>
</tr>
<tr>
<td style="text-align: left;">Conformist</td>
<td style="text-align: right;">M</td>
<td style="text-align: right;">0+K</td>
</tr>
<tr>
<td style="text-align: left;">Rebel</td>
<td style="text-align: right;">M+K</td>
<td style="text-align: right;">0</td>
</tr>
</tbody>
</table>
<h3 id="some-observations"><a class="toclink" href="#some-observations">Some Observations</a></h3>
<ol>
<li>It’s unclear what box configuration Omega should choose for conformists or rebels, so a <em>default</em> (M, K or 0, K) needs to be chosen for those cases. 0, K somehow strikes me as the more natural default, but see also point 5 below where I suggest a new configuration 0, 0.</li>
<li>There may be different simulation regimes (see the table below):<ol>
<li>Simulating both box configurations multiple times (“n sim.”) to capture whether agents follow different policies with some probability. This would need to be real randomness, not some pseudo-randomness that can be simulated. The predictor would then need a separate policy for how to respond to, say, agents that one-box only most of the time. I’ll omit this case here.</li>
<li>Simulating both box configurations once (“2 sim.”).</li>
<li>Simulating one box configuration, and only simulating the other one if the agent makes a decision that is inconsistent (or consistent, but that seems almost equivalent and intuitively odd) with seeing that box configuration (“1.5 sim.”).</li>
<li>Simulating only one box configuration (“1 sim.”) and accepting that you can’t distinguish the conformists and rebels from one of the other strategies.</li>
</ol>
</li>
<li>In cases 1.5 sim. and 1 sim. it is important which scenario is simulated first. The resulting differences are in the table below. A bold choice in the 1.5 sim. table means that the simulation completed after a single run. Regular type means that it took two runs.</li>
<li>Encoding perfect predictions in the two configurations of boxed money is no longer possible as there are only two such configurations in the original problem. A rebel bot that always follows the rebel strategy could thwart any such prediction. But you could introduce a 0, 0 payoff combination as default (in the above sense) to discourage some strategies that update on the boxed money:<ol>
<li>The 0, 0 payoff could await both conformist and rebel. A conformist bot would then see 0, 0 even though they would’ve one-boxed if the payoffs had been M, K.</li>
<li>The predictor could also choose to only use the default for rebels since that is the one policy that would otherwise lead to paradox. The predictor could then randomize what to give to the conformist.</li>
<li>Finally, the predictor could use the (perhaps cheaper) 1.5 sim. regime to determine the payoffs in option b. It won’t be able to tell the conformist from the one-boxer or the two-boxer, but it will recognize the rebel.</li>
</ol>
</li>
<li>If you know which strategy you’ll follow, say, because you’ve precommited to it, and you encounter an inconsistent situation, you still can’t exploit your knowledge because, less interestingly, you’ll have precommitted and because, more interestingly, you’d need to behave in two different ways depending on whether you’re in the simulation or outside, but you can’t because you don’t which is which: If you’ve secretly and nonbindingly committed to two-boxing, and you see M, K, it may be because you’re in a simulation and should one-box or because you’ve one-boxed in the simulation should now really two-box. I’ll mention another reservation later.</li>
</ol>
<h3 id="simulation-regimes"><a class="toclink" href="#simulation-regimes">Simulation Regimes</a></h3>
<h4 id="2-sim"><a class="toclink" href="#2-sim">2 Sim.</a></h4>
<p>In one column the M, K configuration is tested first and then the 0, K configuration. In the other column the order is reversed. It makes no difference in this case – the resulting configurations are all the same.</p>
<table>
<thead>
<tr>
<th style="text-align: left;"></th>
<th style="text-align: right;">M, K; 0, K</th>
<th style="text-align: right;">0, K; M, K</th>
</tr>
</thead>
<tbody>
<tr>
<td style="text-align: left;">One-boxer</td>
<td style="text-align: right;">M, K</td>
<td style="text-align: right;">M, K</td>
</tr>
<tr>
<td style="text-align: left;">Two-boxer</td>
<td style="text-align: right;">0, K</td>
<td style="text-align: right;">0, K</td>
</tr>
<tr>
<td style="text-align: left;">Conformist</td>
<td style="text-align: right;">default</td>
<td style="text-align: right;">default</td>
</tr>
<tr>
<td style="text-align: left;">Rebel</td>
<td style="text-align: right;">default</td>
<td style="text-align: right;">default</td>
</tr>
</tbody>
</table>
<h4 id="15-sim-consistent"><a class="toclink" href="#15-sim-consistent">1.5 Sim. (Consistent)</a></h4>
<p>In each column, the second configuration is now bracketed because it is only tested if the agent behaved inconsistently in the first configuration. If they behaved consistently right away, the simulation is complete and the resulting configuration is bolded.</p>
<table>
<thead>
<tr>
<th style="text-align: left;"></th>
<th style="text-align: right;">M, K; (0, K)</th>
<th style="text-align: right;">0, K; (M, K)</th>
</tr>
</thead>
<tbody>
<tr>
<td style="text-align: left;">One-boxer</td>
<td style="text-align: right;">M, K</td>
<td style="text-align: right;">M, K</td>
</tr>
<tr>
<td style="text-align: left;">Two-boxer</td>
<td style="text-align: right;">0, K</td>
<td style="text-align: right;">0, K</td>
</tr>
<tr>
<td style="text-align: left;">Conformist</td>
<td style="text-align: right;">M, K</td>
<td style="text-align: right;">0, K</td>
</tr>
<tr>
<td style="text-align: left;">Rebel</td>
<td style="text-align: right;">default</td>
<td style="text-align: right;">default</td>
</tr>
</tbody>
</table>
<h4 id="1-sim"><a class="toclink" href="#1-sim">1 Sim.</a></h4>
<p>With a single simulation, the conformist and the one-boxer, and the rebel and the two-boxer behave the same if only the M, K configuration is tested. Otherwise, if only the 0, K configuration is tested, the one-boxer and the rebel, and the two-boxer and the conformist behave the same.</p>
<table>
<thead>
<tr>
<th style="text-align: left;"></th>
<th style="text-align: right;">M, K</th>
<th style="text-align: right;">0, K</th>
</tr>
</thead>
<tbody>
<tr>
<td style="text-align: left;">One-boxer</td>
<td style="text-align: right;">M, K</td>
<td style="text-align: right;">M, K</td>
</tr>
<tr>
<td style="text-align: left;">Two-boxer</td>
<td style="text-align: right;">0, K</td>
<td style="text-align: right;">0, K</td>
</tr>
<tr>
<td style="text-align: left;">Conformist</td>
<td style="text-align: right;">M, K</td>
<td style="text-align: right;">0, K</td>
</tr>
<tr>
<td style="text-align: left;">Rebel</td>
<td style="text-align: right;">0, K</td>
<td style="text-align: right;">M, K</td>
</tr>
</tbody>
</table>
<h3 id="reasoning-under-indexical-uncertainty"><a class="toclink" href="#reasoning-under-indexical-uncertainty">Reasoning Under Indexical Uncertainty</a></h3>
<p>There are two things that I’m uncertain about:</p>
<ol>
<li>The materials on decision theory that I’ve read didn’t feature a case where the payoffs of the final stage of a multi-stage decision problem varied depending on decisions at earlier stages. So here I hope that a sophisticated agent (as opposed to a myopic or resolute one) will foresee in earlier stages that they’ll, at the final stage, prefer to be in one rather than another decision situation. I don’t know whether this is actually in accordance with the definitions of the terms.</li>
<li>I’m also not clear on whether decision theoretic agents can realize that all their copies will make identical decisions in identical situations (regardless of the stage).</li>
</ol>
<p>But if this is all the case and maybe also if only the first assumption is true, it seems to me that the argument that <span class="caps">EDT</span> and <span class="caps">CDT</span> one-box if they realize their indexical uncertainty goes through for this transparent Newcomb’s problem:</p>
<p>First, even if the probability an agent assigns to being in a simulation were small, the opportunity to causally increase the final payoff by a factor of 1,000 by one-boxing would overwhelm even such rather small probabilities.</p>
<p>Second, the probabilities are actually greater than 50%, so substantial. The self-sampling assumption (<span class="caps">SSA</span>) and the self-indication assumption (<span class="caps">SIA</span>) agree in the case of 2 Sim. and assign probability ⅔ to being in a simulation. (See, e.g., <a href="https://www.fhi.ox.ac.uk/wp-content/uploads/Anthropic_Decision_Theory_Tech_Report.pdf">Anthropic Decision Theory</a>.) In 1.5 Sim. they diverge: I’ll assume that the agents have a flat prior over whether the predictor starts by simulating 0, K or M, K. This uncertainty is indicated by the configurations in bold vs. in square brackets in the graphic below.</p>
<p><svg version="1.1" viewBox="0.0 0.0 578.0 328.43832020997377" fill="none" stroke="none" stroke-linecap="square" stroke-miterlimit="10"
xmlns:xlink="http://www.w3.org/1999/xlink"
xmlns="http://www.w3.org/2000/svg">
<clipPath id="p.0">
<path d="m0 0l578.0 0l0 328.43832l-578.0 0l0 -328.43832z" clip-rule="nonzero"/>
</clipPath>
<g clip-path="url(#p.0)">
<path fill="#000000" fill-opacity="0.0" d="m0 0l578.0 0l0 328.43832l-578.0 0z" fill-rule="evenodd"/>
<path fill="#000000" fill-opacity="0.0" d="m5.7690287 128.0971l49.29134 0l0 68.66142l-49.29134 0z" fill-rule="evenodd"/>
<path fill="#000000" d="m17.48778 175.34087l0.359375 -2.484375l2.796875 0.15625l3.40625 0l0 -0.203125q-3.203125 -1.953125 -4.703125 -4.6875q-1.5 -2.75 -1.5 -6.578125q0 -3.6875 1.34375 -6.390625q1.34375 -2.703125 3.859375 -4.109375q2.515625 -1.421875 5.953125 -1.421875q5.124998 0 8.046873 3.09375q2.921875 3.09375 2.921875 8.546875q0 3.796875 -1.5 6.640625q-1.5 2.828125 -4.65625 4.90625l0 0.203125l3.46875 0l2.84375 -0.203125l0.15625 0.28125l-0.3125 2.453125l-2.765625 -0.046875l-4.484375 0l-1.6406231 0.078125l-0.15625 -0.15625l0.046875 -2.96875q2.999998 -2.0 4.374998 -4.546875q1.375 -2.5625 1.375 -6.171875q0 -4.59375 -2.046875 -7.046875q-2.03125 -2.46875 -5.828123 -2.46875q-3.765625 0 -5.78125 2.4375q-2.015625 2.4375 -2.015625 7.046875q0 3.59375 1.359375 6.125q1.359375 2.515625 4.4375 4.625l0.046875 2.96875l-0.203125 0.15625l-1.609375 -0.078125l-4.546875 0l-2.8125 0.125l-0.234375 -0.28125z" fill-rule="nonzero"/>
<path fill="#000000" fill-opacity="0.0" d="m130.26772 73.283646l0 0c0 -4.943123 4.007187 -8.95031 8.950317 -8.95031l91.23323 0c2.373764 0 4.6503143 0.9429703 6.328827 2.6214828c1.6785126 1.678505 2.6214905 3.9550552 2.6214905 6.328827l0 35.800163c0 4.943123 -4.007202 8.95031 -8.950317 8.95031l-91.23323 0c-4.9431305 0 -8.950317 -4.007187 -8.950317 -8.95031z" fill-rule="evenodd"/>
<path stroke="#000000" stroke-width="1.0" stroke-linejoin="round" stroke-linecap="butt" d="m130.26772 73.283646l0 0c0 -4.943123 4.007187 -8.95031 8.950317 -8.95031l91.23323 0c2.373764 0 4.6503143 0.9429703 6.328827 2.6214828c1.6785126 1.678505 2.6214905 3.9550552 2.6214905 6.328827l0 35.800163c0 4.943123 -4.007202 8.95031 -8.950317 8.95031l-91.23323 0c-4.9431305 0 -8.950317 -4.007187 -8.950317 -8.95031z" fill-rule="evenodd"/>
<path fill="#000000" d="m161.72408 98.10372l-2.234375 0l-0.1875 -3.640625l-0.421875 -4.625l-0.1875 0l-2.6875 7.0l-1.96875 0l-2.4375 -7.0l-0.171875 0l-0.46875 4.390625l-0.25 3.875l-2.125 0l1.203125 -12.015625l2.90625 0l0.484375 1.765625l1.90625 5.703125l0.171875 0l2.0 -5.703125l0.53125 -1.765625l2.859375 0l1.078125 12.015625zm3.64682 -2.46875q0.828125 0.671875 0.828125 1.609375q0 1.625 -2.25 3.234375l-0.796875 -0.71875l-0.03125 -0.1875q1.03125 -0.859375 1.03125 -1.828125q0 -0.625 -0.59375 -1.109375l0.015625 -0.203125q0.859375 -0.53125 1.796875 -0.796875zm10.870621 -3.875q2.0 2.78125 4.609375 5.828125l-0.03125 0.265625q-1.046875 0.265625 -2.46875 0.4375l-0.296875 -0.15625q-2.4375 -3.234375 -4.34375 -6.21875l-0.046875 2.375l0.09375 3.8125l-2.515625 0l0.125 -3.578125l-0.125 -8.4375l2.5625 0l-0.09375 5.578125l3.375 -4.421875l0.6875 -1.15625l2.71875 0l-4.25 5.671875z" fill-rule="nonzero"/>
<path fill="#000000" d="m189.02754 86.1506l0 0.703125l-2.21875 0.203125l0 12.453125l2.21875 0.203125l0 0.703125l-0.09375 0.09375l-3.34375 -0.1875l0.078125 -7.046875l-0.078125 -7.0625l3.34375 -0.1875l0.09375 0.125zm5.138626 2.546875q1.890625 0 2.921875 1.234375q1.046875 1.21875 1.046875 3.46875q0 2.3125 -1.09375 3.578125q-1.078125 1.25 -3.0625 1.25q-1.890625 0 -2.9375 -1.203125q-1.03125 -1.21875 -1.03125 -3.4375q0 -2.34375 1.078125 -3.609375q1.09375 -1.28125 3.078125 -1.28125zm-0.125 1.359375q-1.421875 0 -2.140625 0.859375q-0.71875 0.859375 -0.71875 2.5625q0 1.71875 0.71875 2.59375q0.734375 0.859375 2.15625 0.859375q1.421875 0 2.15625 -0.84375q0.734375 -0.859375 0.734375 -2.515625q0 -1.734375 -0.734375 -2.625q-0.734375 -0.890625 -2.171875 -0.890625zm7.3192596 6.421875q0.78125 0.484375 0.78125 1.25q0 1.21875 -1.75 2.5l-0.515625 -0.46875l0 -0.125q1.03125 -0.796875 1.03125 -1.640625q0 -0.5 -0.578125 -0.875l0 -0.15625q0.515625 -0.359375 1.03125 -0.484375zm8.14743 -2.109375l0.0625 3.734375l-1.484375 0l0.0625 -3.59375l-0.0625 -8.359375l1.5 0l-0.078125 8.21875zm1.84375 -2.609375q2.484375 3.203125 5.1875 6.03125l-0.015625 0.1875q-0.5625 0.15625 -1.40625 0.296875l-0.1875 -0.09375q-2.9375 -3.34375 -5.140625 -6.34375l0 -0.109375l3.984375 -4.53125l0.75 -1.046875l1.6875 0l-4.859375 5.609375zm5.2987213 8.640625l0 -0.71875l2.21875 -0.203125l0 -12.4375l-2.21875 -0.203125l0 -0.71875l0.09375 -0.09375l3.34375 0.1875l-0.078125 7.0625l0.078125 7.046875l-3.34375 0.1875l-0.09375 -0.109375z" fill-rule="nonzero"/>
<path fill="#000000" fill-opacity="0.0" d="m290.26773 73.283646l0 0c0 -4.943123 4.0071716 -8.95031 8.950287 -8.95031l91.233246 0c2.3737793 0 4.6503296 0.9429703 6.328827 2.6214828c1.6784973 1.678505 2.6214905 3.9550552 2.6214905 6.328827l0 35.800163c0 4.943123 -4.007202 8.95031 -8.950317 8.95031l-91.233246 0c-4.943115 0 -8.950287 -4.007187 -8.950287 -8.95031z" fill-rule="evenodd"/>
<path stroke="#000000" stroke-width="1.0" stroke-linejoin="round" stroke-linecap="butt" d="m290.26773 73.283646l0 0c0 -4.943123 4.0071716 -8.95031 8.950287 -8.95031l91.233246 0c2.3737793 0 4.6503296 0.9429703 6.328827 2.6214828c1.6784973 1.678505 2.6214905 3.9550552 2.6214905 6.328827l0 35.800163c0 4.943123 -4.007202 8.95031 -8.950317 8.95031l-91.233246 0c-4.943115 0 -8.950287 -4.007187 -8.950287 -8.95031z" fill-rule="evenodd"/>
<path fill="#000000" d="m312.96384 88.5256q1.90625 0 2.9375 1.234375q1.046875 1.234375 1.046875 3.5q0 2.375 -1.125 3.703125q-1.125 1.3125 -3.171875 1.3125q-1.90625 0 -2.953125 -1.234375q-1.03125 -1.234375 -1.03125 -3.453125q0 -2.40625 1.125 -3.734375q1.125 -1.328125 3.171875 -1.328125zm-0.1875 2.296875q-1.109375 0 -1.65625 0.625q-0.53125 0.625 -0.53125 1.96875q0 1.359375 0.53125 2.015625q0.546875 0.65625 1.6875 0.65625q1.109375 0 1.65625 -0.625q0.5625 -0.625 0.5625 -1.9375q0 -1.375 -0.5625 -2.03125q-0.5625 -0.671875 -1.6875 -0.671875zm7.5480347 4.8125q0.828125 0.671875 0.828125 1.609375q0 1.625 -2.25 3.234375l-0.796875 -0.71875l-0.03125 -0.1875q1.03125 -0.859375 1.03125 -1.828125q0 -0.625 -0.59375 -1.109375l0.015625 -0.203125q0.859375 -0.53125 1.796875 -0.796875zm10.870605 -3.875q2.0 2.78125 4.609375 5.828125l-0.03125 0.265625q-1.046875 0.265625 -2.46875 0.4375l-0.296875 -0.15625q-2.4375 -3.234375 -4.34375 -6.21875l-0.046875 2.375l0.09375 3.8125l-2.515625 0l0.125 -3.578125l-0.125 -8.4375l2.5625 0l-0.09375 5.578125l3.375 -4.421875l0.6875 -1.15625l2.71875 0l-4.25 5.671875z" fill-rule="nonzero"/>
<path fill="#000000" d="m343.98102 86.1506l0 0.703125l-2.21875 0.203125l0 12.453125l2.21875 0.203125l0 0.703125l-0.09375 0.09375l-3.34375 -0.1875l0.078125 -7.046875l-0.078125 -7.0625l3.34375 -0.1875l0.09375 0.125zm13.685486 11.953125l-1.34375 0l-0.28125 -3.640625l-0.625 -5.953125l-0.140625 0l-3.3125 8.21875l-1.234375 0l-3.15625 -8.21875l-0.140625 0l-0.640625 5.84375l-0.3125 3.75l-1.265625 0l1.25 -11.953125l1.6875 0l0.625 1.828125l2.59375 6.796875l0.109375 0l2.71875 -6.921875l0.625 -1.703125l1.671875 0l1.171875 11.953125zm3.5166626 -1.625q0.78125 0.484375 0.78125 1.25q0 1.21875 -1.75 2.5l-0.515625 -0.46875l0 -0.125q1.03125 -0.796875 1.03125 -1.640625q0 -0.5 -0.578125 -0.875l0 -0.15625q0.515625 -0.359375 1.03125 -0.484375zm8.14743 -2.109375l0.0625 3.734375l-1.484375 0l0.0625 -3.59375l-0.0625 -8.359375l1.5 0l-0.078125 8.21875zm1.84375 -2.609375q2.484375 3.203125 5.1875 6.03125l-0.015625 0.1875q-0.5625 0.15625 -1.40625 0.296875l-0.1875 -0.09375q-2.9375 -3.34375 -5.140625 -6.34375l0 -0.109375l3.984375 -4.53125l0.75 -1.046875l1.6875 0l-4.859375 5.609375zm5.2987366 8.640625l0 -0.71875l2.21875 -0.203125l0 -12.4375l-2.21875 -0.203125l0 -0.71875l0.09375 -0.09375l3.34375 0.1875l-0.078125 7.0625l0.078125 7.046875l-3.34375 0.1875l-0.09375 -0.109375z" fill-rule="nonzero"/>
<path fill="#000000" fill-opacity="0.0" d="m130.26772 217.28365l0 0c0 -4.943115 4.007187 -8.950317 8.950317 -8.950317l91.23323 0c2.373764 0 4.6503143 0.9429779 6.328827 2.6214905c1.6785126 1.6785126 2.6214905 3.9550476 2.6214905 6.328827l0 35.80017c0 4.9431 -4.007202 8.950302 -8.950317 8.950302l-91.23323 0c-4.9431305 0 -8.950317 -4.007202 -8.950317 -8.950302z" fill-rule="evenodd"/>
<path stroke="#000000" stroke-width="1.0" stroke-linejoin="round" stroke-linecap="butt" d="m130.26772 217.28365l0 0c0 -4.943115 4.007187 -8.950317 8.950317 -8.950317l91.23323 0c2.373764 0 4.6503143 0.9429779 6.328827 2.6214905c1.6785126 1.6785126 2.6214905 3.9550476 2.6214905 6.328827l0 35.80017c0 4.9431 -4.007202 8.950302 -8.950317 8.950302l-91.23323 0c-4.9431305 0 -8.950317 -4.007202 -8.950317 -8.950302z" fill-rule="evenodd"/>
<path fill="#000000" d="m161.72408 242.10373l-2.234375 0l-0.1875 -3.640625l-0.421875 -4.625l-0.1875 0l-2.6875 7.0l-1.96875 0l-2.4375 -7.0l-0.171875 0l-0.46875 4.390625l-0.25 3.875l-2.125 0l1.203125 -12.015625l2.90625 0l0.484375 1.765625l1.90625 5.703125l0.171875 0l2.0 -5.703125l0.53125 -1.765625l2.859375 0l1.078125 12.015625zm3.64682 -2.46875q0.828125 0.671875 0.828125 1.609375q0 1.625 -2.25 3.234375l-0.796875 -0.71875l-0.03125 -0.1875q1.03125 -0.859375 1.03125 -1.828125q0 -0.625 -0.59375 -1.109375l0.015625 -0.203125q0.859375 -0.53125 1.796875 -0.796875zm10.870621 -3.875q2.0 2.78125 4.609375 5.828125l-0.03125 0.265625q-1.046875 0.265625 -2.46875 0.4375l-0.296875 -0.15625q-2.4375 -3.234375 -4.34375 -6.21875l-0.046875 2.375l0.09375 3.8125l-2.515625 0l0.125 -3.578125l-0.125 -8.4375l2.5625 0l-0.09375 5.578125l3.375 -4.421875l0.6875 -1.15625l2.71875 0l-4.25 5.671875z" fill-rule="nonzero"/>
<path fill="#000000" d="m189.02754 230.1506l0 0.703125l-2.21875 0.203125l0 12.453125l2.21875 0.203125l0 0.703125l-0.09375 0.09375l-3.34375 -0.1875l0.078125 -7.046875l-0.078125 -7.0625l3.34375 -0.1875l0.09375 0.125zm5.138626 2.546875q1.890625 0 2.921875 1.234375q1.046875 1.21875 1.046875 3.46875q0 2.3125 -1.09375 3.578125q-1.078125 1.25 -3.0625 1.25q-1.890625 0 -2.9375 -1.203125q-1.03125 -1.21875 -1.03125 -3.4375q0 -2.34375 1.078125 -3.609375q1.09375 -1.28125 3.078125 -1.28125zm-0.125 1.359375q-1.421875 0 -2.140625 0.859375q-0.71875 0.859375 -0.71875 2.5625q0 1.71875 0.71875 2.59375q0.734375 0.859375 2.15625 0.859375q1.421875 0 2.15625 -0.84375q0.734375 -0.859375 0.734375 -2.515625q0 -1.734375 -0.734375 -2.625q-0.734375 -0.890625 -2.171875 -0.890625zm7.3192596 6.421875q0.78125 0.484375 0.78125 1.25q0 1.21875 -1.75 2.5l-0.515625 -0.46875l0 -0.125q1.03125 -0.796875 1.03125 -1.640625q0 -0.5 -0.578125 -0.875l0 -0.15625q0.515625 -0.359375 1.03125 -0.484375zm8.14743 -2.109375l0.0625 3.734375l-1.484375 0l0.0625 -3.59375l-0.0625 -8.359375l1.5 0l-0.078125 8.21875zm1.84375 -2.609375q2.484375 3.203125 5.1875 6.03125l-0.015625 0.1875q-0.5625 0.15625 -1.40625 0.296875l-0.1875 -0.09375q-2.9375 -3.34375 -5.140625 -6.34375l0 -0.109375l3.984375 -4.53125l0.75 -1.046875l1.6875 0l-4.859375 5.609375zm5.2987213 8.640625l0 -0.71875l2.21875 -0.203125l0 -12.4375l-2.21875 -0.203125l0 -0.71875l0.09375 -0.09375l3.34375 0.1875l-0.078125 7.0625l0.078125 7.046875l-3.34375 0.1875l-0.09375 -0.109375z" fill-rule="nonzero"/>
<path fill="#000000" fill-opacity="0.0" d="m450.26773 217.28365l0 0c0 -4.943115 4.0071716 -8.950317 8.950287 -8.950317l91.23328 0c2.3737183 0 4.6502686 0.9429779 6.3287964 2.6214905c1.6785278 1.6785126 2.62146 3.9550476 2.62146 6.328827l0 35.80017c0 4.9431 -4.007141 8.950302 -8.950256 8.950302l-91.23328 0c-4.943115 0 -8.950287 -4.007202 -8.950287 -8.950302z" fill-rule="evenodd"/>
<path stroke="#000000" stroke-width="1.0" stroke-linejoin="round" stroke-linecap="butt" d="m450.26773 217.28365l0 0c0 -4.943115 4.0071716 -8.950317 8.950287 -8.950317l91.23328 0c2.3737183 0 4.6502686 0.9429779 6.3287964 2.6214905c1.6785278 1.6785126 2.62146 3.9550476 2.62146 6.328827l0 35.80017c0 4.9431 -4.007141 8.950302 -8.950256 8.950302l-91.23328 0c-4.943115 0 -8.950287 -4.007202 -8.950287 -8.950302z" fill-rule="evenodd"/>
<path fill="#000000" d="m481.7241 242.10373l-2.234375 0l-0.1875 -3.640625l-0.421875 -4.625l-0.1875 0l-2.6875 7.0l-1.96875 0l-2.4375 -7.0l-0.171875 0l-0.46875 4.390625l-0.25 3.875l-2.125 0l1.203125 -12.015625l2.90625 0l0.484375 1.765625l1.90625 5.703125l0.171875 0l2.0 -5.703125l0.53125 -1.765625l2.859375 0l1.078125 12.015625zm3.64682 -2.46875q0.828125 0.671875 0.828125 1.609375q0 1.625 -2.25 3.234375l-0.796875 -0.71875l-0.03125 -0.1875q1.03125 -0.859375 1.03125 -1.828125q0 -0.625 -0.59375 -1.109375l0.015625 -0.203125q0.859375 -0.53125 1.796875 -0.796875zm10.870605 -3.875q2.0 2.78125 4.609375 5.828125l-0.03125 0.265625q-1.046875 0.265625 -2.46875 0.4375l-0.296875 -0.15625q-2.4375 -3.234375 -4.34375 -6.21875l-0.046875 2.375l0.09375 3.8125l-2.515625 0l0.125 -3.578125l-0.125 -8.4375l2.5625 0l-0.09375 5.578125l3.375 -4.421875l0.6875 -1.15625l2.71875 0l-4.25 5.671875z" fill-rule="nonzero"/>
<path fill="#000000" d="m509.02756 230.1506l0 0.703125l-2.21875 0.203125l0 12.453125l2.21875 0.203125l0 0.703125l-0.09375 0.09375l-3.34375 -0.1875l0.078125 -7.046875l-0.078125 -7.0625l3.34375 -0.1875l0.09375 0.125zm5.1386414 2.546875q1.890625 0 2.921875 1.234375q1.046875 1.21875 1.046875 3.46875q0 2.3125 -1.09375 3.578125q-1.078125 1.25 -3.0625 1.25q-1.890625 0 -2.9375305 -1.203125q-1.03125 -1.21875 -1.03125 -3.4375q0 -2.34375 1.078125 -3.609375q1.0937805 -1.28125 3.0781555 -1.28125zm-0.125 1.359375q-1.421875 0 -2.1406555 0.859375q-0.71875 0.859375 -0.71875 2.5625q0 1.71875 0.71875 2.59375q0.7344055 0.859375 2.1562805 0.859375q1.421875 0 2.15625 -0.84375q0.734375 -0.859375 0.734375 -2.515625q0 -1.734375 -0.734375 -2.625q-0.734375 -0.890625 -2.171875 -0.890625zm7.319214 6.421875q0.78125 0.484375 0.78125 1.25q0 1.21875 -1.75 2.5l-0.515625 -0.46875l0 -0.125q1.03125 -0.796875 1.03125 -1.640625q0 -0.5 -0.578125 -0.875l0 -0.15625q0.515625 -0.359375 1.03125 -0.484375zm8.147461 -2.109375l0.0625 3.734375l-1.484375 0l0.0625 -3.59375l-0.0625 -8.359375l1.5 0l-0.078125 8.21875zm1.84375 -2.609375q2.484375 3.203125 5.1875 6.03125l-0.015625 0.1875q-0.5625 0.15625 -1.40625 0.296875l-0.1875 -0.09375q-2.9375 -3.34375 -5.140625 -6.34375l0 -0.109375l3.984375 -4.53125l0.75 -1.046875l1.6875 0l-4.859375 5.609375zm5.298706 8.640625l0 -0.71875l2.21875 -0.203125l0 -12.4375l-2.21875 -0.203125l0 -0.71875l0.09375 -0.09375l3.34375 0.1875l-0.078125 7.0625l0.078125 7.046875l-3.34375 0.1875l-0.09375 -0.109375z" fill-rule="nonzero"/>
<path fill="#000000" fill-opacity="0.0" d="m450.26773 73.283646l0 0c0 -4.943123 4.0071716 -8.95031 8.950287 -8.95031l91.23328 0c2.3737183 0 4.6502686 0.9429703 6.3287964 2.6214828c1.6785278 1.678505 2.62146 3.9550552 2.62146 6.328827l0 35.800163c0 4.943123 -4.007141 8.95031 -8.950256 8.95031l-91.23328 0c-4.943115 0 -8.950287 -4.007187 -8.950287 -8.95031z" fill-rule="evenodd"/>
<path stroke="#000000" stroke-width="1.0" stroke-linejoin="round" stroke-linecap="butt" d="m450.26773 73.283646l0 0c0 -4.943123 4.0071716 -8.95031 8.950287 -8.95031l91.23328 0c2.3737183 0 4.6502686 0.9429703 6.3287964 2.6214828c1.6785278 1.678505 2.62146 3.9550552 2.62146 6.328827l0 35.800163c0 4.943123 -4.007141 8.95031 -8.950256 8.95031l-91.23328 0c-4.943115 0 -8.950287 -4.007187 -8.950287 -8.95031z" fill-rule="evenodd"/>
<path fill="#000000" d="m472.96384 88.5256q1.90625 0 2.9375 1.234375q1.046875 1.234375 1.046875 3.5q0 2.375 -1.125 3.703125q-1.125 1.3125 -3.171875 1.3125q-1.90625 0 -2.953125 -1.234375q-1.03125 -1.234375 -1.03125 -3.453125q0 -2.40625 1.125 -3.734375q1.125 -1.328125 3.171875 -1.328125zm-0.1875 2.296875q-1.109375 0 -1.65625 0.625q-0.53125 0.625 -0.53125 1.96875q0 1.359375 0.53125 2.015625q0.546875 0.65625 1.6875 0.65625q1.109375 0 1.65625 -0.625q0.5625 -0.625 0.5625 -1.9375q0 -1.375 -0.5625 -2.03125q-0.5625 -0.671875 -1.6875 -0.671875zm7.5480347 4.8125q0.828125 0.671875 0.828125 1.609375q0 1.625 -2.25 3.234375l-0.796875 -0.71875l-0.03125 -0.1875q1.03125 -0.859375 1.03125 -1.828125q0 -0.625 -0.59375 -1.109375l0.015625 -0.203125q0.859375 -0.53125 1.796875 -0.796875zm10.870605 -3.875q2.0 2.78125 4.609375 5.828125l-0.03125 0.265625q-1.046875 0.265625 -2.46875 0.4375l-0.296875 -0.15625q-2.4375 -3.234375 -4.34375 -6.21875l-0.046875 2.375l0.09375 3.8125l-2.515625 0l0.125 -3.578125l-0.125 -8.4375l2.5625 0l-0.09375 5.578125l3.375 -4.421875l0.6875 -1.15625l2.71875 0l-4.25 5.671875z" fill-rule="nonzero"/>
<path fill="#000000" d="m503.98102 86.1506l0 0.703125l-2.21875 0.203125l0 12.453125l2.21875 0.203125l0 0.703125l-0.09375 0.09375l-3.34375 -0.1875l0.078125 -7.046875l-0.078125 -7.0625l3.34375 -0.1875l0.09375 0.125zm13.685486 11.953125l-1.34375 0l-0.28125 -3.640625l-0.625 -5.953125l-0.140625 0l-3.3125 8.21875l-1.234375 0l-3.15625 -8.21875l-0.140625 0l-0.640625 5.84375l-0.3125 3.75l-1.265625 0l1.25 -11.953125l1.6875 0l0.625 1.828125l2.59375 6.796875l0.109375 0l2.71875 -6.921875l0.625 -1.703125l1.671875 0l1.171875 11.953125zm3.5166626 -1.625q0.78125 0.484375 0.78125 1.25q0 1.21875 -1.75 2.5l-0.515625 -0.46875l0 -0.125q1.03125 -0.796875 1.03125 -1.640625q0 -0.5 -0.578125 -0.875l0 -0.15625q0.515625 -0.359375 1.03125 -0.484375zm8.147461 -2.109375l0.0625 3.734375l-1.484375 0l0.0625 -3.59375l-0.0625 -8.359375l1.5 0l-0.078125 8.21875zm1.84375 -2.609375q2.484375 3.203125 5.1875 6.03125l-0.015625 0.1875q-0.5625 0.15625 -1.40625 0.296875l-0.1875 -0.09375q-2.9375 -3.34375 -5.140625 -6.34375l0 -0.109375l3.984375 -4.53125l0.75 -1.046875l1.6875 0l-4.859375 5.609375zm5.298706 8.640625l0 -0.71875l2.21875 -0.203125l0 -12.4375l-2.21875 -0.203125l0 -0.71875l0.09375 -0.09375l3.34375 0.1875l-0.078125 7.0625l0.078125 7.046875l-3.34375 0.1875l-0.09375 -0.109375z" fill-rule="nonzero"/>
<path fill="#000000" fill-opacity="0.0" d="m55.060368 162.42783l75.2126 -71.2441" fill-rule="evenodd"/>
<path stroke="#000000" stroke-width="1.0" stroke-linejoin="round" stroke-linecap="butt" d="m55.060368 162.42783l70.8566 -67.11793" fill-rule="evenodd"/>
<path fill="#000000" stroke="#000000" stroke-width="1.0" stroke-linecap="butt" d="m127.05285 96.50905l2.1587753 -4.319977l-4.4305496 1.9216614z" fill-rule="evenodd"/>
<path fill="#000000" fill-opacity="0.0" d="m239.40158 91.18372l50.86615 0" fill-rule="evenodd"/>
<path stroke="#000000" stroke-width="1.0" stroke-linejoin="round" stroke-linecap="butt" d="m239.40158 91.18372l44.86615 0" fill-rule="evenodd"/>
<path fill="#000000" stroke="#000000" stroke-width="1.0" stroke-linecap="butt" d="m284.26773 92.83546l4.538086 -1.6517334l-4.538086 -1.6517258z" fill-rule="evenodd"/>
<path fill="#000000" fill-opacity="0.0" d="m399.40158 91.18372l50.86615 0" fill-rule="evenodd"/>
<path stroke="#000000" stroke-width="1.0" stroke-linejoin="round" stroke-linecap="butt" d="m399.40158 91.18372l44.86615 0" fill-rule="evenodd"/>
<path fill="#000000" stroke="#000000" stroke-width="1.0" stroke-linecap="butt" d="m444.26773 92.83546l4.538086 -1.6517334l-4.538086 -1.6517258z" fill-rule="evenodd"/>
<path fill="#000000" fill-opacity="0.0" d="m239.40158 235.18373l210.86615 0" fill-rule="evenodd"/>
<path stroke="#000000" stroke-width="1.0" stroke-linejoin="round" stroke-linecap="butt" d="m239.40158 235.18373l204.86615 0" fill-rule="evenodd"/>
<path fill="#000000" stroke="#000000" stroke-width="1.0" stroke-linecap="butt" d="m444.26773 236.83546l4.538086 -1.6517334l-4.538086 -1.6517334z" fill-rule="evenodd"/>
<path fill="#000000" fill-opacity="0.0" d="m55.060368 162.42783l75.2126 72.755905" fill-rule="evenodd"/>
<path stroke="#000000" stroke-width="1.0" stroke-linejoin="round" stroke-linecap="butt" d="m55.060368 162.42783l70.900116 68.584274" fill-rule="evenodd"/>
<path fill="#000000" stroke="#000000" stroke-width="1.0" stroke-linecap="butt" d="m124.81208 232.19928l4.4101486 1.9680328l-2.11335 -4.342392z" fill-rule="evenodd"/>
<path fill="#000000" fill-opacity="0.0" d="m130.26772 0l80.97638 0l0 36.09449l-80.97638 0z" fill-rule="evenodd"/>
<path fill="#000000" d="m144.25209 14.873123q1.28125 0 2.671875 0.546875l-0.25 1.4375l-0.265625 0.078125q-0.53125 -0.40625 -1.140625 -0.625q-0.609375 -0.234375 -1.203125 -0.234375q-0.6875 0 -1.203125 0.28125q-0.515625 0.28125 -0.796875 0.734375q-0.265625 0.4375 -0.265625 0.921875q0 0.515625 0.3125 0.90625q0.3125 0.375 0.765625 0.65625q0.46875 0.265625 1.28125 0.671875q1.0 0.453125 1.609375 0.859375q0.625 0.40625 1.0625 1.03125q0.4375 0.609375 0.4375 1.484375q0 0.921875 -0.5 1.71875q-0.5 0.78125 -1.421875 1.25q-0.90625 0.453125 -2.078125 0.453125q-0.796875 0 -1.640625 -0.203125q-0.84375 -0.21875 -1.578125 -0.625l0.1875 -1.53125l0.203125 -0.09375q0.65625 0.625 1.453125 0.9375q0.796875 0.3125 1.515625 0.3125q0.703125 0 1.25 -0.28125q0.5625 -0.28125 0.84375 -0.75q0.296875 -0.46875 0.296875 -0.984375q0 -0.78125 -0.609375 -1.25q-0.59375 -0.484375 -1.796875 -1.046875q-0.984375 -0.484375 -1.59375 -0.859375q-0.59375 -0.390625 -1.03125 -1.0q-0.421875 -0.609375 -0.421875 -1.484375q0 -0.90625 0.5 -1.671875q0.5 -0.765625 1.375 -1.203125q0.890625 -0.4375 2.03125 -0.4375zm5.3995056 8.421875l-0.046875 -4.828125l1.4375 -0.125l-0.046875 4.828125l0.046875 3.75l-1.453125 0l0.0625 -3.625zm0.703125 -8.75q0.359375 0 0.578125 0.234375q0.234375 0.21875 0.234375 0.609375q0 0.40625 -0.234375 0.65625q-0.234375 0.234375 -0.640625 0.234375q-0.390625 0 -0.609375 -0.21875q-0.21875 -0.21875 -0.21875 -0.640625q0 -0.390625 0.234375 -0.625q0.234375 -0.25 0.65625 -0.25zm14.421066 6.578125q0.015625 -0.765625 -0.359375 -1.1875q-0.375 -0.4375 -1.09375 -0.4375q-1.203125 0 -2.484375 1.28125l-0.046875 2.421875l0.078125 3.71875l-1.515625 0l0.125 -5.796875q0.015625 -0.765625 -0.359375 -1.1875q-0.375 -0.4375 -1.109375 -0.4375q-1.171875 0 -2.53125 1.328125l0 2.34375l0.078125 3.75l-1.46875 0l0.078125 -3.625l-0.0625 -4.828125l1.390625 -0.125l0 1.25l0.109375 0q1.15625 -1.03125 2.375 -1.328125q1.046875 0 1.71875 0.375q0.671875 0.359375 0.953125 1.046875l0.109375 0q1.171875 -1.0625 2.515625 -1.421875q1.421875 0 2.140625 0.65625q0.734375 0.65625 0.71875 1.859375l-0.046875 2.421875l0.078125 3.71875l-1.484375 0l0.09375 -5.796875zm4.909378 4.15625q0.359375 0 0.578125 0.234375q0.234375 0.21875 0.234375 0.609375q0 0.40625 -0.234375 0.65625q-0.234375 0.234375 -0.640625 0.234375q-0.390625 0 -0.609375 -0.21875q-0.21875 -0.234375 -0.21875 -0.640625q0 -0.390625 0.234375 -0.625q0.234375 -0.25 0.65625 -0.25zm9.381805 -7.578125l-0.09375 5.359375l0.0625 3.859375l-1.484375 0l0.09375 -3.671875l-0.015625 -4.0l-2.40625 0.671875l-0.15625 -0.078125l0 -1.171875l3.765625 -1.109375l0.234375 0.140625z" fill-rule="nonzero"/>
<path fill="#000000" fill-opacity="0.0" d="m290.26773 0l80.97635 0l0 36.09449l-80.97635 0z" fill-rule="evenodd"/>
<path fill="#000000" d="m304.2521 14.873123q1.28125 0 2.671875 0.546875l-0.25 1.4375l-0.265625 0.078125q-0.53125 -0.40625 -1.140625 -0.625q-0.609375 -0.234375 -1.203125 -0.234375q-0.6875 0 -1.203125 0.28125q-0.515625 0.28125 -0.796875 0.734375q-0.265625 0.4375 -0.265625 0.921875q0 0.515625 0.3125 0.90625q0.3125 0.375 0.765625 0.65625q0.46875 0.265625 1.28125 0.671875q1.0 0.453125 1.609375 0.859375q0.625 0.40625 1.0625 1.03125q0.4375 0.609375 0.4375 1.484375q0 0.921875 -0.5 1.71875q-0.5 0.78125 -1.421875 1.25q-0.90625 0.453125 -2.078125 0.453125q-0.796875 0 -1.640625 -0.203125q-0.84375 -0.21875 -1.578125 -0.625l0.1875 -1.53125l0.203125 -0.09375q0.65625 0.625 1.453125 0.9375q0.796875 0.3125 1.515625 0.3125q0.703125 0 1.25 -0.28125q0.5625 -0.28125 0.84375 -0.75q0.296875 -0.46875 0.296875 -0.984375q0 -0.78125 -0.609375 -1.25q-0.59375 -0.484375 -1.796875 -1.046875q-0.984375 -0.484375 -1.59375 -0.859375q-0.59375 -0.390625 -1.03125 -1.0q-0.421875 -0.609375 -0.421875 -1.484375q0 -0.90625 0.5 -1.671875q0.5 -0.765625 1.375 -1.203125q0.890625 -0.4375 2.03125 -0.4375zm5.3995056 8.421875l-0.046875 -4.828125l1.4375 -0.125l-0.046875 4.828125l0.046875 3.75l-1.453125 0l0.0625 -3.625zm0.703125 -8.75q0.359375 0 0.578125 0.234375q0.234375 0.21875 0.234375 0.609375q0 0.40625 -0.234375 0.65625q-0.234375 0.234375 -0.640625 0.234375q-0.390625 0 -0.609375 -0.21875q-0.21875 -0.21875 -0.21875 -0.640625q0 -0.390625 0.234375 -0.625q0.234375 -0.25 0.65625 -0.25zm14.421051 6.578125q0.015625 -0.765625 -0.359375 -1.1875q-0.375 -0.4375 -1.09375 -0.4375q-1.203125 0 -2.484375 1.28125l-0.046875 2.421875l0.078125 3.71875l-1.515625 0l0.125 -5.796875q0.015625 -0.765625 -0.359375 -1.1875q-0.375 -0.4375 -1.109375 -0.4375q-1.171875 0 -2.53125 1.328125l0 2.34375l0.078125 3.75l-1.46875 0l0.078125 -3.625l-0.0625 -4.828125l1.390625 -0.125l0 1.25l0.109375 0q1.15625 -1.03125 2.375 -1.328125q1.046875 0 1.71875 0.375q0.671875 0.359375 0.953125 1.046875l0.109375 0q1.171875 -1.0625 2.515625 -1.421875q1.421875 0 2.140625 0.65625q0.734375 0.65625 0.71875 1.859375l-0.046875 2.421875l0.078125 3.71875l-1.484375 0l0.09375 -5.796875zm4.9093933 4.15625q0.359375 0 0.578125 0.234375q0.234375 0.21875 0.234375 0.609375q0 0.40625 -0.234375 0.65625q-0.234375 0.234375 -0.640625 0.234375q-0.390625 0 -0.609375 -0.21875q-0.21875 -0.234375 -0.21875 -0.640625q0 -0.390625 0.234375 -0.625q0.234375 -0.25 0.65625 -0.25zm9.475555 -7.71875q1.109375 0 1.828125 0.5625q0.71875 0.5625 0.71875 1.5q0 0.75 -0.53125 1.671875q-0.515625 0.921875 -1.640625 2.25l-1.890625 2.140625l0.03125 0.078125l3.453125 0l1.5625 -0.109375l0.125 0.09375l-0.203125 1.25l-1.90625 -0.078125l-4.1875 0l-0.6875 0.03125l-0.09375 -0.0625l0 -0.9375l2.59375 -2.8125q1.015625 -1.125 1.46875 -1.859375q0.46875 -0.734375 0.46875 -1.296875q0 -0.5 -0.375 -0.828125q-0.359375 -0.328125 -1.03125 -0.328125q-0.671875 0 -1.34375 0.34375q-0.65625 0.328125 -1.203125 0.921875l-0.1875 0l-0.234375 -1.140625q0.71875 -0.671875 1.5625 -1.03125q0.84375 -0.359375 1.703125 -0.359375z" fill-rule="nonzero"/>
<path fill="#000000" fill-opacity="0.0" d="m442.26773 0l80.97635 0l0 36.09449l-80.97635 0z" fill-rule="evenodd"/>
<path fill="#000000" d="m457.8771 22.091873q1.21875 2.03125 3.09375 4.46875l-0.015625 0.1875q-0.671875 0.265625 -1.390625 0.34375l-0.203125 -0.140625q-1.53125 -2.28125 -2.734375 -4.46875l-2.1875 0l0 0.703125l0.078125 3.734375l-1.515625 0l0.09375 -3.59375l-0.09375 -8.359375l3.65625 -0.03125q1.75 0 2.703125 0.890625q0.96875 0.890625 0.96875 2.53125q0 1.234375 -0.640625 2.21875q-0.640625 0.96875 -1.8125 1.515625zm-3.421875 -0.875l2.234375 0.046875q2.09375 -0.59375 2.09375 -2.78125q0 -1.140625 -0.625 -1.703125q-0.609375 -0.5625 -1.828125 -0.609375l-1.8125 0.0625l-0.0625 4.984375zm8.666595 1.375l0 0.03125q0 1.578125 0.703125 2.421875q0.703125 0.828125 2.015625 0.828125q0.65625 0 1.296875 -0.203125q0.65625 -0.21875 1.203125 -0.609375l0.140625 0.109375l-0.21875 1.265625q-1.3125 0.640625 -2.609375 0.640625q-1.828125 0 -2.875 -1.171875q-1.046875 -1.171875 -1.046875 -3.203125q0 -2.046875 1.046875 -3.25q1.046875 -1.21875 2.8125 -1.21875q1.5 0 2.3125 0.84375q0.828125 0.84375 0.828125 2.359375q0 0.5 -0.078125 0.953125l-0.1875 0.1875l-5.34375 0.015625zm2.46875 -3.328125q-0.921875 0 -1.546875 0.609375q-0.625 0.609375 -0.84375 1.71875l4.125 -0.078125l0.046875 -0.265625q0 -0.953125 -0.46875 -1.46875q-0.453125 -0.515625 -1.3125 -0.515625zm8.2724 -1.03125q1.203125 0 1.90625 0.59375q0.703125 0.59375 0.703125 1.640625q0 0.3125 -0.109375 2.296875q-0.109375 1.96875 -0.109375 2.3125q0 0.40625 0.125 0.5625q0.125 0.15625 0.453125 0.15625q0.28125 0 0.625 -0.109375l0.15625 0.140625l-0.15625 0.828125q-0.53125 0.328125 -1.0625 0.375q-0.578125 -0.046875 -0.921875 -0.40625q-0.34375 -0.359375 -0.359375 -0.96875l-0.140625 0q-1.0 1.015625 -1.984375 1.390625q-1.25 0 -1.953125 -0.625q-0.703125 -0.625 -0.703125 -1.703125q0 -0.8125 0.4375 -1.296875q0.4375 -0.5 1.296875 -0.703125l3.03125 -0.703125l0.015625 -1.21875q0 -0.71875 -0.390625 -1.109375q-0.375 -0.390625 -1.09375 -0.390625q-0.625 0 -1.265625 0.296875q-0.625 0.296875 -1.359375 0.921875l-0.140625 -0.0625l-0.1875 -1.0625q1.546875 -1.15625 3.1875 -1.15625zm-1.5 5.390625q-0.375 0.109375 -0.546875 0.3125q-0.15625 0.1875 -0.15625 0.578125q0 0.609375 0.34375 0.96875q0.34375 0.34375 0.921875 0.34375q1.03125 0 2.09375 -1.328125l0.046875 -1.53125l-2.703125 0.65625zm8.759827 -0.453125l0.09375 3.75l-1.484375 0l0.109375 -3.640625l-0.09375 -9.328125l1.484375 -0.125l-0.109375 9.34375zm3.2330933 0.125l-0.046875 -4.828125l1.4375 -0.125l-0.046875 4.828125l0.046875 3.75l-1.453125 0l0.0625 -3.625zm0.703125 -8.75q0.359375 0 0.578125 0.234375q0.234375 0.21875 0.234375 0.609375q0 0.40625 -0.234375 0.65625q-0.234375 0.234375 -0.640625 0.234375q-0.390625 0 -0.609375 -0.21875q-0.21875 -0.21875 -0.21875 -0.640625q0 -0.390625 0.234375 -0.625q0.234375 -0.25 0.65625 -0.25zm6.405426 11.234375q0.625 0 1.34375 -0.375l0.171875 0.109375l-0.140625 0.984375q-1.015625 0.453125 -1.875 0.53125q-1.0625 -0.203125 -1.53125 -0.78125q-0.46875 -0.578125 -0.46875 -1.671875l0.0625 -1.328125l-0.015625 -3.859375l-1.171875 0l-0.078125 -0.125l0.09375 -0.875l1.140625 0l0 -1.375l1.28125 -0.5625l0.15625 0.140625l-0.046875 1.796875l2.546875 0l0.0625 0.125l-0.09375 0.875l-2.515625 0l-0.09375 5.046875q-0.015625 0.75 0.265625 1.046875q0.28125 0.296875 0.90625 0.296875zm3.62323 -6.359375l2.0625 5.890625l0.125 0l2.125 -5.765625l0.28125 -1.171875l1.328125 0l-3.0625 8.25q-0.46875 1.265625 -1.0 1.984375q-0.53125 0.734375 -1.296875 1.15625q-0.75 0.421875 -1.984375 0.71875l-0.125 0q-0.390625 -0.390625 -0.609375 -0.875l0.03125 -0.171875q1.421875 -0.28125 2.171875 -0.796875q0.765625 -0.5 1.40625 -1.71875l-3.125 -8.5l1.390625 -0.09375l0.28125 1.09375z" fill-rule="nonzero"/>
<path fill="#000000" fill-opacity="0.0" d="m59.695538 96.01837l36.094486 0l0 36.094482l-36.094486 0z" fill-rule="evenodd"/>
<path fill="#000000" d="m70.77367 123.453995l10.03125 -14.09375l1.421875 0l-10.03125 14.09375l-1.421875 0zm0.9375 -7.1875l0 -5.046875q-0.921875 0.734375 -2.046875 1.03125l0 -1.125q0.578125 -0.1875 1.265625 -0.671875q0.6875 -0.484375 1.125 -1.03125l0.984375 0l0 6.84375l-1.328125 0zm6.546875 6.953125q0.046875 -0.515625 0.4375 -1.03125q0.5625 -0.796875 1.875 -1.734375q1.3125 -0.953125 1.546875 -1.234375q0.34375 -0.375 0.34375 -0.765625q0 -0.4375 -0.3125 -0.703125q-0.3125 -0.28125 -0.953125 -0.28125q-0.59375 0 -0.890625 0.21875q-0.296875 0.203125 -0.484375 0.796875l-1.375 -0.125q0.265625 -1.03125 0.921875 -1.5q0.65625 -0.484375 1.859375 -0.484375q1.375 0 2.015625 0.546875q0.640625 0.53125 0.640625 1.3125q0 0.765625 -0.53125 1.4375q-0.40625 0.515625 -1.859375 1.578125q-0.765625 0.546875 -1.078125 0.875l3.515625 0l0 1.09375l-5.671875 0z" fill-rule="nonzero"/>
<path fill="#000000" fill-opacity="0.0" d="m59.695538 184.01837l36.094486 0l0 36.094482l-36.094486 0z" fill-rule="evenodd"/>
<path fill="#000000" d="m70.77367 211.454l10.03125 -14.09375l1.421875 0l-10.03125 14.09375l-1.421875 0zm0.9375 -7.1875l0 -5.046875q-0.921875 0.734375 -2.046875 1.03125l0 -1.125q0.578125 -0.1875 1.265625 -0.671875q0.6875 -0.484375 1.125 -1.03125l0.984375 0l0 6.84375l-1.328125 0zm6.546875 6.953125q0.046875 -0.515625 0.4375 -1.03125q0.5625 -0.796875 1.875 -1.734375q1.3125 -0.953125 1.546875 -1.234375q0.34375 -0.375 0.34375 -0.765625q0 -0.4375 -0.3125 -0.703125q-0.3125 -0.28125 -0.953125 -0.28125q-0.59375 0 -0.890625 0.21875q-0.296875 0.203125 -0.484375 0.796875l-1.375 -0.125q0.265625 -1.03125 0.921875 -1.5q0.65625 -0.484375 1.859375 -0.484375q1.375 0 2.015625 0.546875q0.640625 0.53125 0.640625 1.3125q0 0.765625 -0.53125 1.4375q-0.40625 0.515625 -1.859375 1.578125q-0.765625 0.546875 -1.078125 0.875l3.515625 0l0 1.09375l-5.671875 0z" fill-rule="nonzero"/>
<path fill="#000000" fill-opacity="0.0" d="m130.26772 110.90874l59.2126 0l0 53.70079l-59.2126 0z" fill-rule="evenodd"/>
<path fill="#000000" d="m143.53334 124.92499q1.109375 0 2.296875 0.484375l-0.203125 1.234375l-0.234375 0.0625q-0.46875 -0.34375 -0.984375 -0.53125q-0.515625 -0.203125 -1.03125 -0.203125q-0.59375 0 -1.03125 0.25q-0.4375 0.234375 -0.671875 0.625q-0.234375 0.375 -0.234375 0.796875q0 0.421875 0.265625 0.75q0.265625 0.328125 0.65625 0.578125q0.40625 0.234375 1.09375 0.5625q0.875 0.40625 1.390625 0.75q0.53125 0.34375 0.90625 0.875q0.375 0.53125 0.375 1.28125q0 0.796875 -0.4375 1.46875q-0.421875 0.671875 -1.203125 1.078125q-0.78125 0.390625 -1.796875 0.390625q-0.671875 0 -1.40625 -0.171875q-0.71875 -0.1875 -1.34375 -0.546875l0.15625 -1.3125l0.1875 -0.078125q0.546875 0.546875 1.21875 0.8125q0.6875 0.265625 1.3125 0.265625q0.609375 0 1.078125 -0.234375q0.484375 -0.25 0.71875 -0.640625q0.25 -0.40625 0.25 -0.859375q0 -0.671875 -0.515625 -1.078125q-0.5 -0.40625 -1.546875 -0.890625q-0.84375 -0.421875 -1.359375 -0.75q-0.515625 -0.328125 -0.890625 -0.84375q-0.359375 -0.515625 -0.359375 -1.265625q0 -0.78125 0.421875 -1.4375q0.4375 -0.65625 1.1875 -1.03125q0.765625 -0.390625 1.734375 -0.390625zm7.552017 0q1.109375 0 2.296875 0.484375l-0.203125 1.234375l-0.234375 0.0625q-0.46875 -0.34375 -0.984375 -0.53125q-0.515625 -0.203125 -1.03125 -0.203125q-0.59375 0 -1.03125 0.25q-0.4375 0.234375 -0.671875 0.625q-0.234375 0.375 -0.234375 0.796875q0 0.421875 0.265625 0.75q0.265625 0.328125 0.65625 0.578125q0.40625 0.234375 1.09375 0.5625q0.875 0.40625 1.390625 0.75q0.53125 0.34375 0.90625 0.875q0.375 0.53125 0.375 1.28125q0 0.796875 -0.4375 1.46875q-0.421875 0.671875 -1.203125 1.078125q-0.78125 0.390625 -1.796875 0.390625q-0.671875 0 -1.40625 -0.171875q-0.71875 -0.1875 -1.34375 -0.546875l0.15625 -1.3125l0.1875 -0.078125q0.546875 0.546875 1.21875 0.8125q0.6875 0.265625 1.3125 0.265625q0.609375 0 1.078125 -0.234375q0.484375 -0.25 0.71875 -0.640625q0.25 -0.40625 0.25 -0.859375q0 -0.671875 -0.515625 -1.078125q-0.5 -0.40625 -1.546875 -0.890625q-0.84375 -0.421875 -1.359375 -0.75q-0.515625 -0.328125 -0.890625 -0.84375q-0.359375 -0.515625 -0.359375 -1.265625q0 -0.78125 0.421875 -1.4375q0.4375 -0.65625 1.1875 -1.03125q0.765625 -0.390625 1.734375 -0.390625zm12.177017 10.34375l-1.3125 0l-0.234375 -1.03125l-0.578125 -1.75l-2.296875 -0.046875l-2.078125 0.046875l-0.625 1.6875l-0.265625 1.09375l-1.28125 0l3.765625 -10.25l1.484375 0l3.421875 10.25zm-2.46875 -3.875l-1.671875 -5.046875l-0.109375 0l-1.859375 5.046875l1.796875 0.03125l1.84375 -0.03125zm4.6230164 2.46875q0.3125 0 0.5 0.1875q0.203125 0.1875 0.203125 0.53125q0 0.34375 -0.203125 0.5625q-0.203125 0.203125 -0.546875 0.203125q-0.34375 0 -0.53125 -0.1875q-0.171875 -0.1875 -0.171875 -0.546875q0 -0.328125 0.1875 -0.53125q0.203125 -0.21875 0.5625 -0.21875zm0.0625 -4.0q0.3125 0 0.5 0.1875q0.203125 0.1875 0.203125 0.53125q0 0.34375 -0.203125 0.5625q-0.203125 0.203125 -0.546875 0.203125q-0.34375 0 -0.53125 -0.1875q-0.171875 -0.1875 -0.171875 -0.546875q0 -0.328125 0.1875 -0.53125q0.203125 -0.21875 0.5625 -0.21875z" fill-rule="nonzero"/>
<path fill="#000000" d="m143.53334 146.92499q1.109375 0 2.296875 0.484375l-0.203125 1.234375l-0.234375 0.0625q-0.46875 -0.34375 -0.984375 -0.53125q-0.515625 -0.203125 -1.03125 -0.203125q-0.59375 0 -1.03125 0.25q-0.4375 0.234375 -0.671875 0.625q-0.234375 0.375 -0.234375 0.796875q0 0.421875 0.265625 0.75q0.265625 0.328125 0.65625 0.578125q0.40625 0.234375 1.09375 0.5625q0.875 0.40625 1.390625 0.75q0.53125 0.34375 0.90625 0.875q0.375 0.53125 0.375 1.28125q0 0.796875 -0.4375 1.46875q-0.421875 0.671875 -1.203125 1.078125q-0.78125 0.390625 -1.796875 0.390625q-0.671875 0 -1.40625 -0.171875q-0.71875 -0.1875 -1.34375 -0.546875l0.15625 -1.3125l0.1875 -0.078125q0.546875 0.546875 1.21875 0.8125q0.6875 0.265625 1.3125 0.265625q0.609375 0 1.078125 -0.234375q0.484375 -0.25 0.71875 -0.640625q0.25 -0.40625 0.25 -0.859375q0 -0.671875 -0.515625 -1.078125q-0.5 -0.40625 -1.546875 -0.890625q-0.84375 -0.421875 -1.359375 -0.75q-0.515625 -0.328125 -0.890625 -0.84375q-0.359375 -0.515625 -0.359375 -1.265625q0 -0.78125 0.421875 -1.4375q0.4375 -0.65625 1.1875 -1.03125q0.765625 -0.390625 1.734375 -0.390625zm6.005142 7.140625l0.0625 3.203125l-1.296875 0l0.078125 -3.09375l-0.078125 -7.15625l1.3125 0l-0.078125 7.046875zm10.459885 3.203125l-1.3125 0l-0.234375 -1.03125l-0.578125 -1.75l-2.296875 -0.046875l-2.078125 0.046875l-0.625 1.6875l-0.265625 1.09375l-1.28125 0l3.765625 -10.25l1.484375 0l3.421875 10.25zm-2.46875 -3.875l-1.671875 -5.046875l-0.109375 0l-1.859375 5.046875l1.796875 0.03125l1.84375 -0.03125zm4.6230164 2.46875q0.3125 0 0.5 0.1875q0.203125 0.1875 0.203125 0.53125q0 0.34375 -0.203125 0.5625q-0.203125 0.203125 -0.546875 0.203125q-0.34375 0 -0.53125 -0.1875q-0.171875 -0.1875 -0.171875 -0.546875q0 -0.328125 0.1875 -0.53125q0.203125 -0.21875 0.5625 -0.21875zm0.0625 -4.0q0.3125 0 0.5 0.1875q0.203125 0.1875 0.203125 0.53125q0 0.34375 -0.203125 0.5625q-0.203125 0.203125 -0.546875 0.203125q-0.34375 0 -0.53125 -0.1875q-0.171875 -0.1875 -0.171875 -0.546875q0 -0.328125 0.1875 -0.53125q0.203125 -0.21875 0.5625 -0.21875z" fill-rule="nonzero"/>
<path fill="#000000" fill-opacity="0.0" d="m203.30708 110.908134l36.094498 0l0 53.700783l-36.094498 0z" fill-rule="evenodd"/>
<path fill="#000000" d="m222.76096 125.43751q0.296875 -0.375 0.375 -0.75l0.734375 0l0 6.6093826l-1.1875 0l0 -5.0468826l-1.109375 0l0 -0.515625l0.296875 0q0.515625 0 0.890625 -0.296875zm0.078125 12.312508l-0.53125 -0.296875l5.640625 -12.765633l0.53125 0.21875l-5.640625 12.843758zm3.03125 -3.40625q0 -2.234375 0.96875 -3.203125q0.453125 -0.4375 1.265625 -0.4375q0.8125 0 1.265625 0.4375q0.515625 0.53125 0.59375 1.421875l-1.046875 0q0 -0.53125 -0.21875 -0.75q-0.296875 -0.21875 -0.59375 -0.21875q-0.375 0 -0.59375 0.21875q-0.515625 0.453125 -0.59375 2.078125q0.4375 -0.890625 1.265625 -0.890625q0.734375 0 1.1875 0.453125q0.65625 0.65625 0.65625 1.921875q0 1.265625 -0.65625 1.9375q-0.453125 0.4375 -1.421875 0.4375q-0.8125 0 -1.328125 -0.515625q-0.75 -0.75 -0.75 -2.890625zm2.15625 -0.453125q-0.296875 0 -0.515625 0.21875q-0.515625 0.453125 -0.515625 1.1875q0 0.96875 0.359375 1.34375q0.296875 0.21875 0.671875 0.21875q0.296875 0 0.59375 -0.21875q0.296875 -0.296875 0.296875 -1.265625q0 -0.890625 -0.375 -1.265625q-0.296875 -0.21875 -0.515625 -0.21875z" fill-rule="nonzero"/>
<path fill="#000000" d="m226.1047 156.48439l0.21875 -3.34375l3.34375 0l0 0.890625l-2.453125 0l-0.15625 1.484375q0.53125 -0.359375 1.046875 -0.359375q0.8125 0 1.265625 0.4375q0.515625 0.515625 0.515625 1.5625q0 1.265625 -0.671875 1.921875q-0.515625 0.53125 -1.40625 0.53125q-0.8125 0 -1.265625 -0.453125q-0.515625 -0.515625 -0.59375 -1.625l1.046875 0q0.0625 0.734375 0.359375 1.03125q0.15625 0.15625 0.53125 0.15625q0.359375 0 0.59375 -0.234375q0.359375 -0.359375 0.359375 -1.1875q0 -0.890625 -0.296875 -1.1875q-0.21875 -0.21875 -0.59375 -0.21875q-0.59375 0 -0.953125 0.59375l-0.890625 0zm-3.203125 3.265625l-0.515625 -0.296875l5.640625 -12.765625l0.515625 0.21875l-5.640625 12.84375zm-0.140625 -12.3125q0.296875 -0.375 0.375 -0.75l0.734375 0l0 6.609375l-1.1875 0l0 -5.046875l-1.109375 0l0 -0.515625l0.296875 0q0.515625 0 0.890625 -0.296875z" fill-rule="nonzero"/>
<path fill="#000000" fill-opacity="0.0" d="m363.3071 110.908134l36.094482 0l0 53.700783l-36.094482 0z" fill-rule="evenodd"/>
<path fill="#000000" d="m382.76096 125.43751q0.296875 -0.375 0.375 -0.75l0.734375 0l0 6.6093826l-1.1875 0l0 -5.0468826l-1.109375 0l0 -0.515625l0.296875 0q0.515625 0 0.890625 -0.296875zm0.078125 12.312508l-0.53125 -0.296875l5.640625 -12.765633l0.53125 0.21875l-5.640625 12.843758zm3.03125 -3.40625q0 -2.234375 0.96875 -3.203125q0.453125 -0.4375 1.265625 -0.4375q0.8125 0 1.265625 0.4375q0.515625 0.53125 0.59375 1.421875l-1.046875 0q0 -0.53125 -0.21875 -0.75q-0.296875 -0.21875 -0.59375 -0.21875q-0.375 0 -0.59375 0.21875q-0.515625 0.453125 -0.59375 2.078125q0.4375 -0.890625 1.265625 -0.890625q0.734375 0 1.1875 0.453125q0.65625 0.65625 0.65625 1.921875q0 1.265625 -0.65625 1.9375q-0.453125 0.4375 -1.421875 0.4375q-0.8125 0 -1.328125 -0.515625q-0.75 -0.75 -0.75 -2.890625zm2.15625 -0.453125q-0.296875 0 -0.515625 0.21875q-0.515625 0.453125 -0.515625 1.1875q0 0.96875 0.359375 1.34375q0.296875 0.21875 0.671875 0.21875q0.296875 0 0.59375 -0.21875q0.296875 -0.296875 0.296875 -1.265625q0 -0.890625 -0.375 -1.265625q-0.296875 -0.21875 -0.515625 -0.21875z" fill-rule="nonzero"/>
<path fill="#000000" d="m386.1047 156.48439l0.21875 -3.34375l3.34375 0l0 0.890625l-2.453125 0l-0.15625 1.484375q0.53125 -0.359375 1.046875 -0.359375q0.8125 0 1.265625 0.4375q0.515625 0.515625 0.515625 1.5625q0 1.265625 -0.671875 1.921875q-0.515625 0.53125 -1.40625 0.53125q-0.8125 0 -1.265625 -0.453125q-0.515625 -0.515625 -0.59375 -1.625l1.046875 0q0.0625 0.734375 0.359375 1.03125q0.15625 0.15625 0.53125 0.15625q0.359375 0 0.59375 -0.234375q0.359375 -0.359375 0.359375 -1.1875q0 -0.890625 -0.296875 -1.1875q-0.21875 -0.21875 -0.59375 -0.21875q-0.59375 0 -0.953125 0.59375l-0.890625 0zm-3.203125 3.265625l-0.515625 -0.296875l5.640625 -12.765625l0.515625 0.21875l-5.640625 12.84375zm-0.140625 -12.3125q0.296875 -0.375 0.375 -0.75l0.734375 0l0 6.609375l-1.1875 0l0 -5.046875l-1.109375 0l0 -0.515625l0.296875 0q0.515625 0 0.890625 -0.296875z" fill-rule="nonzero"/>
<path fill="#000000" fill-opacity="0.0" d="m523.30707 110.908134l36.094482 0l0 53.700783l-36.094482 0z" fill-rule="evenodd"/>
<path fill="#000000" d="m542.7609 125.43751q0.296875 -0.375 0.375 -0.75l0.734375 0l0 6.6093826l-1.1875 0l0 -5.0468826l-1.109375 0l0 -0.515625l0.296875 0q0.515625 0 0.890625 -0.296875zm0.078125 12.312508l-0.53125 -0.296875l5.640625 -12.765633l0.53125 0.21875l-5.640625 12.843758zm3.03125 -3.40625q0 -2.234375 0.96875 -3.203125q0.453125 -0.4375 1.265625 -0.4375q0.8125 0 1.265625 0.4375q0.515625 0.53125 0.59375 1.421875l-1.046875 0q0 -0.53125 -0.21875 -0.75q-0.296875 -0.21875 -0.59375 -0.21875q-0.375 0 -0.59375 0.21875q-0.515625 0.453125 -0.59375 2.078125q0.4375 -0.890625 1.265625 -0.890625q0.734375 0 1.1875 0.453125q0.65625 0.65625 0.65625 1.921875q0 1.265625 -0.65625 1.9375q-0.453125 0.4375 -1.421875 0.4375q-0.8125 0 -1.328125 -0.515625q-0.75 -0.75 -0.75 -2.890625zm2.15625 -0.453125q-0.296875 0 -0.515625 0.21875q-0.515625 0.453125 -0.515625 1.1875q0 0.96875 0.359375 1.34375q0.296875 0.21875 0.671875 0.21875q0.296875 0 0.59375 -0.21875q0.296875 -0.296875 0.296875 -1.265625q0 -0.890625 -0.375 -1.265625q-0.296875 -0.21875 -0.515625 -0.21875z" fill-rule="nonzero"/>
<path fill="#000000" d="m546.1047 156.48439l0.21875 -3.34375l3.34375 0l0 0.890625l-2.453125 0l-0.15625 1.484375q0.53125 -0.359375 1.046875 -0.359375q0.8125 0 1.265625 0.4375q0.515625 0.515625 0.515625 1.5625q0 1.265625 -0.671875 1.921875q-0.515625 0.53125 -1.40625 0.53125q-0.8125 0 -1.265625 -0.453125q-0.515625 -0.515625 -0.59375 -1.625l1.046875 0q0.0625 0.734375 0.359375 1.03125q0.15625 0.15625 0.53125 0.15625q0.359375 0 0.59375 -0.234375q0.359375 -0.359375 0.359375 -1.1875q0 -0.890625 -0.296875 -1.1875q-0.21875 -0.21875 -0.59375 -0.21875q-0.59375 0 -0.953125 0.59375l-0.890625 0zm-3.203125 3.265625l-0.515625 -0.296875l5.640625 -12.765625l0.515625 0.21875l-5.640625 12.84375zm-0.140625 -12.3125q0.296875 -0.375 0.375 -0.75l0.734375 0l0 6.609375l-1.1875 0l0 -5.046875l-1.109375 0l0 -0.515625l0.296875 0q0.515625 0 0.890625 -0.296875z" fill-rule="nonzero"/>
<path fill="#000000" fill-opacity="0.0" d="m130.26772 254.90874l59.2126 0l0 53.70079l-59.2126 0z" fill-rule="evenodd"/>
<path fill="#000000" d="m143.53334 268.925q1.109375 0 2.296875 0.484375l-0.203125 1.234375l-0.234375 0.0625q-0.46875 -0.34375 -0.984375 -0.53125q-0.515625 -0.203125 -1.03125 -0.203125q-0.59375 0 -1.03125 0.25q-0.4375 0.234375 -0.671875 0.625q-0.234375 0.375 -0.234375 0.796875q0 0.421875 0.265625 0.75q0.265625 0.328125 0.65625 0.578125q0.40625 0.234375 1.09375 0.5625q0.875 0.40625 1.390625 0.75q0.53125 0.34375 0.90625 0.875q0.375 0.53125 0.375 1.28125q0 0.796875 -0.4375 1.46875q-0.421875 0.671875 -1.203125 1.078125q-0.78125 0.390625 -1.796875 0.390625q-0.671875 0 -1.40625 -0.171875q-0.71875 -0.1875 -1.34375 -0.546875l0.15625 -1.3125l0.1875 -0.078125q0.546875 0.546875 1.21875 0.8125q0.6875 0.265625 1.3125 0.265625q0.609375 0 1.078125 -0.234375q0.484375 -0.25 0.71875 -0.640625q0.25 -0.40625 0.25 -0.859375q0 -0.671875 -0.515625 -1.078125q-0.5 -0.40625 -1.546875 -0.890625q-0.84375 -0.421875 -1.359375 -0.75q-0.515625 -0.328125 -0.890625 -0.84375q-0.359375 -0.515625 -0.359375 -1.265625q0 -0.78125 0.421875 -1.4375q0.4375 -0.65625 1.1875 -1.03125q0.765625 -0.390625 1.734375 -0.390625zm7.552017 0q1.109375 0 2.296875 0.484375l-0.203125 1.234375l-0.234375 0.0625q-0.46875 -0.34375 -0.984375 -0.53125q-0.515625 -0.203125 -1.03125 -0.203125q-0.59375 0 -1.03125 0.25q-0.4375 0.234375 -0.671875 0.625q-0.234375 0.375 -0.234375 0.796875q0 0.421875 0.265625 0.75q0.265625 0.328125 0.65625 0.578125q0.40625 0.234375 1.09375 0.5625q0.875 0.40625 1.390625 0.75q0.53125 0.34375 0.90625 0.875q0.375 0.53125 0.375 1.28125q0 0.796875 -0.4375 1.46875q-0.421875 0.671875 -1.203125 1.078125q-0.78125 0.390625 -1.796875 0.390625q-0.671875 0 -1.40625 -0.171875q-0.71875 -0.1875 -1.34375 -0.546875l0.15625 -1.3125l0.1875 -0.078125q0.546875 0.546875 1.21875 0.8125q0.6875 0.265625 1.3125 0.265625q0.609375 0 1.078125 -0.234375q0.484375 -0.25 0.71875 -0.640625q0.25 -0.40625 0.25 -0.859375q0 -0.671875 -0.515625 -1.078125q-0.5 -0.40625 -1.546875 -0.890625q-0.84375 -0.421875 -1.359375 -0.75q-0.515625 -0.328125 -0.890625 -0.84375q-0.359375 -0.515625 -0.359375 -1.265625q0 -0.78125 0.421875 -1.4375q0.4375 -0.65625 1.1875 -1.03125q0.765625 -0.390625 1.734375 -0.390625zm12.177017 10.34375l-1.3125 0l-0.234375 -1.03125l-0.578125 -1.75l-2.296875 -0.046875l-2.078125 0.046875l-0.625 1.6875l-0.265625 1.09375l-1.28125 0l3.765625 -10.25l1.484375 0l3.421875 10.25zm-2.46875 -3.875l-1.671875 -5.046875l-0.109375 0l-1.859375 5.046875l1.796875 0.03125l1.84375 -0.03125zm4.6230164 2.46875q0.3125 0 0.5 0.1875q0.203125 0.1875 0.203125 0.53125q0 0.34375 -0.203125 0.5625q-0.203125 0.203125 -0.546875 0.203125q-0.34375 0 -0.53125 -0.1875q-0.171875 -0.1875 -0.171875 -0.546875q0 -0.328125 0.1875 -0.53125q0.203125 -0.21875 0.5625 -0.21875zm0.0625 -4.0q0.3125 0 0.5 0.1875q0.203125 0.1875 0.203125 0.53125q0 0.34375 -0.203125 0.5625q-0.203125 0.203125 -0.546875 0.203125q-0.34375 0 -0.53125 -0.1875q-0.171875 -0.1875 -0.171875 -0.546875q0 -0.328125 0.1875 -0.53125q0.203125 -0.21875 0.5625 -0.21875z" fill-rule="nonzero"/>
<path fill="#000000" d="m143.53334 290.925q1.109375 0 2.296875 0.484375l-0.203125 1.234375l-0.234375 0.0625q-0.46875 -0.34375 -0.984375 -0.53125q-0.515625 -0.203125 -1.03125 -0.203125q-0.59375 0 -1.03125 0.25q-0.4375 0.234375 -0.671875 0.625q-0.234375 0.375 -0.234375 0.796875q0 0.421875 0.265625 0.75q0.265625 0.328125 0.65625 0.578125q0.40625 0.234375 1.09375 0.5625q0.875 0.40625 1.390625 0.75q0.53125 0.34375 0.90625 0.875q0.375 0.53125 0.375 1.28125q0 0.796875 -0.4375 1.46875q-0.421875 0.671875 -1.203125 1.078125q-0.78125 0.390625 -1.796875 0.390625q-0.671875 0 -1.40625 -0.171875q-0.71875 -0.1875 -1.34375 -0.546875l0.15625 -1.3125l0.1875 -0.078125q0.546875 0.546875 1.21875 0.8125q0.6875 0.265625 1.3125 0.265625q0.609375 0 1.078125 -0.234375q0.484375 -0.25 0.71875 -0.640625q0.25 -0.40625 0.25 -0.859375q0 -0.671875 -0.515625 -1.078125q-0.5 -0.40625 -1.546875 -0.890625q-0.84375 -0.421875 -1.359375 -0.75q-0.515625 -0.328125 -0.890625 -0.84375q-0.359375 -0.515625 -0.359375 -1.265625q0 -0.78125 0.421875 -1.4375q0.4375 -0.65625 1.1875 -1.03125q0.765625 -0.390625 1.734375 -0.390625zm6.005142 7.140625l0.0625 3.203125l-1.296875 0l0.078125 -3.09375l-0.078125 -7.15625l1.3125 0l-0.078125 7.046875zm10.459885 3.203125l-1.3125 0l-0.234375 -1.03125l-0.578125 -1.75l-2.296875 -0.046875l-2.078125 0.046875l-0.625 1.6875l-0.265625 1.09375l-1.28125 0l3.765625 -10.25l1.484375 0l3.421875 10.25zm-2.46875 -3.875l-1.671875 -5.046875l-0.109375 0l-1.859375 5.046875l1.796875 0.03125l1.84375 -0.03125zm4.6230164 2.46875q0.3125 0 0.5 0.1875q0.203125 0.1875 0.203125 0.53125q0 0.34375 -0.203125 0.5625q-0.203125 0.203125 -0.546875 0.203125q-0.34375 0 -0.53125 -0.1875q-0.171875 -0.1875 -0.171875 -0.546875q0 -0.328125 0.1875 -0.53125q0.203125 -0.21875 0.5625 -0.21875zm0.0625 -4.0q0.3125 0 0.5 0.1875q0.203125 0.1875 0.203125 0.53125q0 0.34375 -0.203125 0.5625q-0.203125 0.203125 -0.546875 0.203125q-0.34375 0 -0.53125 -0.1875q-0.171875 -0.1875 -0.171875 -0.546875q0 -0.328125 0.1875 -0.53125q0.203125 -0.21875 0.5625 -0.21875z" fill-rule="nonzero"/>
<path fill="#000000" fill-opacity="0.0" d="m203.30708 254.90814l36.094498 0l0 53.700775l-36.094498 0z" fill-rule="evenodd"/>
<path fill="#000000" d="m222.14468 269.065l-0.03125 3.828125l0.03125 2.171875l-1.140625 0l0.03125 -2.015625l0 -2.875l-1.28125 0.34375l-0.09375 -0.046875l0 -0.90625l2.34375 -0.65625l0.140625 0.15625zm5.734375 -0.328125l0.109375 0.234375l-7.5625 11.75l-0.5625 -0.3125l-0.078125 -0.203125l7.546875 -11.78125l0.546875 0.3125zm1.921875 8.96875l-0.9375 -0.046875l0.0625 1.515625l-0.0625 0.09375l-0.890625 0.046875l-0.0625 -0.0625l0.015625 -1.59375l-1.828125 0l-0.8125 0.046875l-0.078125 -0.125l0 -0.703125l2.625 -3.671875l1.03125 0l-0.015625 3.609375l0.921875 0l0.09375 0.09375l-0.0625 0.796875zm-3.609375 -0.890625l1.75 0l0 -2.34375l-0.078125 0l-1.671875 2.34375z" fill-rule="nonzero"/>
<path fill="#000000" d="m226.1047 300.48438l0.21875 -3.34375l3.34375 0l0 0.890625l-2.453125 0l-0.15625 1.484375q0.53125 -0.359375 1.046875 -0.359375q0.8125 0 1.265625 0.4375q0.515625 0.515625 0.515625 1.5625q0 1.265625 -0.671875 1.921875q-0.515625 0.53125 -1.40625 0.53125q-0.8125 0 -1.265625 -0.453125q-0.515625 -0.515625 -0.59375 -1.625l1.046875 0q0.0625 0.734375 0.359375 1.03125q0.15625 0.15625 0.53125 0.15625q0.359375 0 0.59375 -0.234375q0.359375 -0.359375 0.359375 -1.1875q0 -0.890625 -0.296875 -1.1875q-0.21875 -0.21875 -0.59375 -0.21875q-0.59375 0 -0.953125 0.59375l-0.890625 0zm-3.203125 3.265625l-0.515625 -0.296875l5.640625 -12.765625l0.515625 0.21875l-5.640625 12.84375zm-0.140625 -12.3125q0.296875 -0.375 0.375 -0.75l0.734375 0l0 6.609375l-1.1875 0l0 -5.046875l-1.109375 0l0 -0.515625l0.296875 0q0.515625 0 0.890625 -0.296875z" fill-rule="nonzero"/>
<path fill="#000000" fill-opacity="0.0" d="m523.30707 254.90814l36.094482 0l0 53.700775l-36.094482 0z" fill-rule="evenodd"/>
<path fill="#000000" d="m542.14465 269.065l-0.03125 3.828125l0.03125 2.171875l-1.140625 0l0.03125 -2.015625l0 -2.875l-1.28125 0.34375l-0.09375 -0.046875l0 -0.90625l2.34375 -0.65625l0.140625 0.15625zm5.734375 -0.328125l0.109375 0.234375l-7.5625 11.75l-0.5625 -0.3125l-0.078125 -0.203125l7.546875 -11.78125l0.546875 0.3125zm1.921875 8.96875l-0.9375 -0.046875l0.0625 1.515625l-0.0625 0.09375l-0.890625 0.046875l-0.0625 -0.0625l0.015625 -1.59375l-1.828125 0l-0.8125 0.046875l-0.078125 -0.125l0 -0.703125l2.625 -3.671875l1.03125 0l-0.015625 3.609375l0.921875 0l0.09375 0.09375l-0.0625 0.796875zm-3.609375 -0.890625l1.75 0l0 -2.34375l-0.078125 0l-1.671875 2.34375z" fill-rule="nonzero"/>
<path fill="#000000" d="m546.1047 300.48438l0.21875 -3.34375l3.34375 0l0 0.890625l-2.453125 0l-0.15625 1.484375q0.53125 -0.359375 1.046875 -0.359375q0.8125 0 1.265625 0.4375q0.515625 0.515625 0.515625 1.5625q0 1.265625 -0.671875 1.921875q-0.515625 0.53125 -1.40625 0.53125q-0.8125 0 -1.265625 -0.453125q-0.515625 -0.515625 -0.59375 -1.625l1.046875 0q0.0625 0.734375 0.359375 1.03125q0.15625 0.15625 0.53125 0.15625q0.359375 0 0.59375 -0.234375q0.359375 -0.359375 0.359375 -1.1875q0 -0.890625 -0.296875 -1.1875q-0.21875 -0.21875 -0.59375 -0.21875q-0.59375 0 -0.953125 0.59375l-0.890625 0zm-3.203125 3.265625l-0.515625 -0.296875l5.640625 -12.765625l0.515625 0.21875l-5.640625 12.84375zm-0.140625 -12.3125q0.296875 -0.375 0.375 -0.75l0.734375 0l0 6.609375l-1.1875 0l0 -5.046875l-1.109375 0l0 -0.515625l0.296875 0q0.515625 0 0.890625 -0.296875z" fill-rule="nonzero"/>
</g>
</svg></p>
<p>At each stage you see the content of the boxes, so if it is M, K (without loss of generality), you’re either (1) in the bold sim. 1 case at the top, (2) the bracketed sim. 2 case at the top, (3) the bracketed reality case at the top, (4) the bold sim. 1 case at the bottom, or (5) the bold reality case at the bottom. (Hence the <span class="caps">SIA</span> probabilities of <span class="math">\(\frac{1}{5}\)</span> each.)</p>
<p>A few reservations and simplifications:</p>
<ol>
<li>I ignore “rebel” agents in this example as they could know for certain that they’re not in one of Omega’s simulations if they see 0, 0. That would make things more complicated.</li>
<li>What would also make things complicated are agents who have a strong inclination to one- or two-box and are aware of it. For example, an agent with a strong inclination to one-box may be surprised to see 0, K and so conclude that it’s even more likely to be in a simulation than its <span class="caps">SSA</span> or <span class="caps">SIA</span> prior implies. Conversely, having a strong inclination and seeing a configuration that is consistent with it may boost once confidence that one is not in a simulation. If the inclination is to one-box (<em>with</em> loss of generality), that might make two-boxing more attractive again, thereby undermining the very inclination to one-box. This seems complicated. Agents with no particular initial inclination (or that are not aware of it) seem to be a simpler case.<sup id="fnref:2"><a class="footnote-ref" href="#fn:2">2</a></sup>
Adding up the fractions yields an <span class="caps">SSA</span> probability of <span class="math">\(\frac{7}{12}\)</span> and an <span class="caps">SIA</span> probability of <span class="math">\(\frac{3}{5}\)</span> in favor of being in a simulation.</li>
</ol>
<p>I’ve wondered whether there is a way to tweak the payoffs of two-boxing such that an agent that doesn’t take into account that they can’t act differently depending on which stage they’re at can be tricked into making a nonoptimal choice, but I haven’t found one yet.</p>
<p>My main caveat, as mentioned above, is that I’m uncertain over whether <span class="caps">CDT</span> and <span class="caps">EDT</span> realize that they’ll make the same recommendation in the same situation regardless of the stage. But if that’s the case (and maybe also if not), it seems to me that this argument checks out and indexical (or anthropic) uncertainty really lets <span class="caps">CDT</span> and <span class="caps">EDT</span> one-box in the transparent Newcomb’s problem.</p>
<h2 id="notes"><a class="toclink" href="#notes">Notes</a></h2>
<div class="footnote">
<hr>
<ol>
<li id="fn:1">
<p>Is there a difference? I use “indexical” because it seems more self-explanatory to me. <a class="footnote-backref" href="#fnref:1" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
<li id="fn:2">
<p>This might be related to the <a href="https://philpapers.org/rec/EELCDT">metatickle defence</a>. I haven’t read the paper, though. <a class="footnote-backref" href="#fnref:2" title="Jump back to footnote 2 in the text">↩</a></p>
</li>
</ol>
</div>
<script type="text/javascript">if (!document.getElementById('mathjaxscript_pelican_#%@#$@#')) {
var align = "center",
indent = "0em",
linebreak = "false";
if (false) {
align = (screen.width < 768) ? "left" : align;
indent = (screen.width < 768) ? "0em" : indent;
linebreak = (screen.width < 768) ? 'true' : linebreak;
}
var mathjaxscript = document.createElement('script');
mathjaxscript.id = 'mathjaxscript_pelican_#%@#$@#';
mathjaxscript.type = 'text/javascript';
mathjaxscript.src = 'https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.3/latest.js?config=TeX-AMS-MML_HTMLorMML';
var configscript = document.createElement('script');
configscript.type = 'text/x-mathjax-config';
configscript[(window.opera ? "innerHTML" : "text")] =
"MathJax.Hub.Config({" +
" config: ['MMLorHTML.js']," +
" TeX: { extensions: ['AMSmath.js','AMSsymbols.js','noErrors.js','noUndefined.js'], equationNumbers: { autoNumber: 'none' } }," +
" jax: ['input/TeX','input/MathML','output/HTML-CSS']," +
" extensions: ['tex2jax.js','mml2jax.js','MathMenu.js','MathZoom.js']," +
" displayAlign: '"+ align +"'," +
" displayIndent: '"+ indent +"'," +
" showMathMenu: true," +
" messageStyle: 'normal'," +
" tex2jax: { " +
" inlineMath: [ ['\\\\(','\\\\)'] ], " +
" displayMath: [ ['$$','$$'] ]," +
" processEscapes: true," +
" preview: 'TeX'," +
" }, " +
" 'HTML-CSS': { " +
" availableFonts: ['STIX', 'TeX']," +
" preferredFont: 'STIX'," +
" styles: { '.MathJax_Display, .MathJax .mo, .MathJax .mi, .MathJax .mn': {color: 'inherit ! important'} }," +
" linebreaks: { automatic: "+ linebreak +", width: '90% container' }," +
" }, " +
"}); " +
"if ('default' !== 'default') {" +
"MathJax.Hub.Register.StartupHook('HTML-CSS Jax Ready',function () {" +
"var VARIANT = MathJax.OutputJax['HTML-CSS'].FONTDATA.VARIANT;" +
"VARIANT['normal'].fonts.unshift('MathJax_default');" +
"VARIANT['bold'].fonts.unshift('MathJax_default-bold');" +
"VARIANT['italic'].fonts.unshift('MathJax_default-italic');" +
"VARIANT['-tex-mathit'].fonts.unshift('MathJax_default-italic');" +
"});" +
"MathJax.Hub.Register.StartupHook('SVG Jax Ready',function () {" +
"var VARIANT = MathJax.OutputJax.SVG.FONTDATA.VARIANT;" +
"VARIANT['normal'].fonts.unshift('MathJax_default');" +
"VARIANT['bold'].fonts.unshift('MathJax_default-bold');" +
"VARIANT['italic'].fonts.unshift('MathJax_default-italic');" +
"VARIANT['-tex-mathit'].fonts.unshift('MathJax_default-italic');" +
"});" +
"}";
(document.body || document.getElementsByTagName('head')[0]).appendChild(configscript);
(document.body || document.getElementsByTagName('head')[0]).appendChild(mathjaxscript);
}
</script>Frequently Asked Questions2020-08-10T21:00:00+00:002020-08-10T21:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2020-08-10:/faq.html<p>A few frequently asked questions and my tentative and preliminary responses to them. I also try to give an approximate lower bound on how frequently the questions were asked.</p><h2 id="who-are-you"><a class="toclink" href="#who-are-you">Who are you?</a></h2>
<p><strong>Frequency</strong>: Approximately 708,000,000 Google results by 2020-08-02. The interest is currently highest in the Philippines (among countries where English queries are likely to be used) according to Google Trends.</p>
<p>I’m a bit fuzzy on this one. It seems broadly defensible to identify with the continuous stream of person moments that is experienced by a particular biological body. A major advantage of doing so is that it’s probably the default assumption that others are going to make when it comes to guessing what you might mean by “I.” Thus it aids in communication.</p>
<p>But I find it somewhat unsatisfying. An exact copy of mine would, for some time, react identically to the same stimuli. This seems like a clear case where <em>I</em> am implemented in two bodies. The same goes for digitized bodies of mine, or at least I’m rather confident in physicalism so that I think that they can be treated the same. (<a href="https://casparoesterheld.com/2018/12/07/not-your-grandmas-phenomenal-idealism/">Some friends of mine might disagree about the specifics.</a>)</p>
<p>After a little while, these two bodies will diverge because of the different experiences they have. (I’ve conducted an experiment to get a bit of a quantitative idea of how much I’ve diverged from my version of 2015.) But it still seems like a stretch to me not to identify with that version of me.</p>
<p>Likewise, if the universe is infinite (or just sufficiently large) in some temporal, spatial, or other direction, if there is infinite matter, and if it’s not distributed regularly, there are many exact and near copies of me just like my past and future mes. If these didn’t want to identify with me, in the sense of this physical body, then our interests may clash a bit here. But as it stands, I take my readiness to identify with them as evidence that they’re fine with my identifying with them.</p>
<p>But the fuzziness doesn’t end there. Brian Tomasik has addressed the question of <a href="https://reducing-suffering.org/interpret-physical-system-mind/">what constitutes the same algorithm</a>. It’s tricky. I suppose nature wasn’t built with such concepts as identity, algorithm, or computation in mind, so when we try to impose them onto reality, the abstraction remains leaky.</p>
<p>All in all, I’m left with the intuition that there likely is some sort of notion of similarity, be it one that is imposed, hard to measure, and often undefined. That would imply that I’m more some systems than others. But it probably doesn’t put any hard limits on who I am unless you introduce an arbitrary threshold. So I can’t even exclude that I’m everyone, just to vastly different degrees.</p>
<p>So apologies if this is a bit of an unsatisfying answer to a seemingly mundane question. I’ll update this entry when I learn more.</p>
<h2 id="what-is-the-purpose-of-life"><a class="toclink" href="#what-is-the-purpose-of-life">What is the purpose of life?</a></h2>
<p><strong>Frequency</strong>: Approximately 124,000,000 Google results by 2020-08-02. Much fewer than for “Who are you?” But then again it’s also a longer question, so fewer people have enough time to ask all of it. The interest is currently highest in India (among countries where English queries are likely to be used) according to Google Trends.</p>
<p>I figured I would make it about minimizing suffering. Not selecting any purpose for myself made me feel lost, and I didn’t like that. But your mileage may vary. It’s not like I’m a normative authority on the issue.</p>
<p>Some children seem to have this phase where they ask “Why?,” listen to your answer, and then follow it up with “Why?” I’ve certainly had this phase, lasting all my life up to now, but being introverted I just asked myself. At some point you arrive at something terminal, something <em>ipso facto</em>, like minimizing suffering or maximizing well-being because that’s what reinforcement learners are all about.</p>
<p>I don’t mean “minimizing suffering” in any exclusive sense. If you want to focus on improving your spelling, you don’t start writing just “tapir tapir tapir tapir tapir” to no end because it’s the one word that you’re most confident you’re spelling correctly. Rather minimizing suffering is something that I feel strongly about, which is motivating, and is shared among many value systems so that there are many uncontroversial ways to achieve it. That makes it attractive as a purpose. It may even be the case that the best way to minimize suffering is to do wildly unrelated things if <a href="https://longtermrisk.org/msr">evidential cooperation in large worlds</a> works out. Others who are in an even better position to reduce suffering will then “reciprocate” (though potentially nonreciprocatorily) by reducing even more suffering.</p>
<p>Finally, the reason I say “<em>minimize</em>” rather than “<em>reduce</em> suffering” is that we may be in the sad position of only being able to, say, slow the increase of suffering or slow the acceleration of the increase of suffering. I feel like the m-term gets that across better than the r-term. But in either case, I suppose most people will know what is meant.</p>
<h2 id="are-you-nice"><a class="toclink" href="#are-you-nice">Are you nice?</a></h2>
<p><strong>Frequency</strong>: Approximately 1,210,000 Google results by 2020-08-02. The interest is currently highest in South Africa (among countries where English queries are likely to be used) according to Google Trends.</p>
<p>(Well, that’s a bit awkward to answer oneself, but maybe it’s less awkward after acknowledging its awkwardness.)</p>
<p>Hopefully, because I’m trying! I’ve also fallen short of my own expectations before, but I hope I’ve learned and won’t make the same mistakes twice.</p>
<p>The reason is that I imagine the natural state as one where every individual (in the physical body over time sense) fends for the survival of themselves and their offspring (if they’re a <a href="https://en.wikipedia.org/wiki/R/K_selection_theory">K-strategist</a>, otherwise it’s even worse). You live a short life at subsistence level. Just rummaging, hunting, and fighting nonstop or predation and starvation put an early end to you. Meanwhile disease and the cold may still kill you any day. I might be able to <a href="https://www.instagram.com/telofy/">climb</a> to safety from some predators, but not from others, and I can’t escape viruses and the cold that way. So all in all, my life in nature would be stressful, painful, and brutal. At least it would be short.</p>
<p>As it happens, I was born into a species in a country at a time where that’s not the case. The only explanation that I can think of for why it’s not the case is <a href="https://impartial-priorities.org/friends-of-the-civilization.html">civilization</a>.</p>
<p>Over the millenia, moral circles emanated outward from each of us, rippling like ripples, pulsing like a pulse. Care and trust were central to them. Going outward, trust became more guarded and care more conditional, but it took a few more ripples before it faded entirely. Over time, the circles became wider and wider. And that reduced transaction costs.</p>
<p>Imagine you want to buy a product online:</p>
<ol>
<li>It costs a couple dollars, comes in the mail the next day, and if it’s the wrong size, you return it, otherwise you pay the bill that’s attached.</li>
<li>Or it costs a fortune. You wait for a few weeks until the seller’s brother has time to fly to your neck of the woods. He arrives together with his bodyguard, who has one hand on her gun while you try on the product. You find that it fits. You take it off again, careful not to pull out the tubes. The bodyguard waits a few meters away with the product while you pay the brother. The brother meticulously checks the genuineness of the bills, then makes a sign to the bodyguard. Both take off quickly and you run over to your product so it won’t get stolen.</li>
</ol>
<p>If the store trusts with sufficiently high probability that you’ll pay, it can afford to implement the first procedure, which is much more efficient as it’s faster, cheaper, and doesn’t require special, unusually trusted staff like a brother. That also makes it more robust as business doesn’t break down when the brother gets sick.</p>
<p>There’s even this pretty decent correlation between <a href="https://ourworldindata.org/trust">trust and <span class="caps">GDP</span></a>, but I suppose there the causation likely runs both ways.</p>
<p>So my take-away has been that whenever we behave less nicely than our cooperation partners would’ve expected us to, there’s a risk that they’ll notice and update downward on how nicely they can expect someone like you (not just you) to act. That gradually erodes trust and increases <a href="https://en.wikipedia.org/wiki/Anomie">anomie</a>. It doesn’t always backfire, but the expected backfiring is nonzero. In the limit, it destroys civilizations.</p>
<p>There’s not just one civilization but many nested quasi-civilizations: families, relationships, friend groups, companies, communities, countries, etc. If they are small or young, or if you’re in a powerful position in them, the risk of destroying them is high. If the opposite is the case, the risk is lower.</p>
<p>Conversely, whenever someone asks more nicely than their cooperation partner would’ve expected them too, and the cooperation partner doesn’t exploit or free-ride on that niceness, you have a chance to improve civilization or your part of it. This is a bit risky, too, in that you open yourself up to exploitation and the other may overgeneralize from the experience and may be exploited by someone else later. But whenever one can afford to, it seems superior to me to err slightly in the direction of greater niceness than warranted than less niceness than warranted.</p>
<p>It may also be that you have to trade off great gains (say, suffering reduction) against niceness or even niceness against niceness, say, if you can only be nice to one of two parties who consider each other enemies. Here I warn of four dangers: (1) underestimating the value of civilization for you and your goal; (2) underestimating your expected negative influence especially on small civilizations like communities, companies, and relationships; (3) falling prey to a <a href="https://www.lesswrong.com/posts/erGipespbbzdG5zYb/the-third-alternative">false dilemma</a>, because usually there is a third alternative better than either of the first two; (4) not taking into account what <a href="https://longtermrisk.org/msr">evidence this gives you of the behavior of those others</a> who think similarly to you but may have opposite goals from you.</p>
<p>So that’s why I want to be nice.</p>
<p>Finally, note that there are <a href="https://impartial-priorities.org/self-study-directions-2020.html#tit-for-tat-or-pavlovian-cultures-for-safe-agi">at least two different kinds of niceness</a>, so depending on which one comes naturally to you, you need to take special care when interacting with those who are used to the other kind.</p>
<h2 id="can-i-ask-you-something"><a class="toclink" href="#can-i-ask-you-something">Can I ask you something?</a></h2>
<p><strong>Frequency</strong>: Approximately 1,440,000 Google results by 2020-08-02. The interest is currently highest in Pakistan (among countries where English queries are likely to be used) according to Google Trends.</p>
<p>Sure. I’m a bit annoyed with the usual process:</p>
<ol>
<li>Try to think of an excuse why you’re contacting the person.</li>
<li>Try to think of social padding to cushion the question.</li>
<li>“Can you imagine smells?”</li>
</ol>
<p>Stage 1 kills off a felt 99% of contact attempts, stage 2 is time consuming, and so a lot of interesting questions don’t get asked.</p>
<p>Maybe the world would be an ever so slightly better place if more people had an entry in their FAQs to license padding-less questions. This is mine. The same goes for non-question statements, e.g., “I can’t imagine smells.”</p>Self-Similarity Experiment2020-08-10T20:00:00+00:002020-08-10T20:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2020-08-10:/self-similarity-experiment.html<p>Some of the people on earth who are most similar to you are likely your own person moments from other points in time. Your degree of similarity to them informs (though I haven’t worked out how) what the density is of you-like computation in the universe. This question is interesting for <a href="https://longtermrisk.org/msr">evidential cooperation</a> as I hope that it can help to disentangle evidential cooperation from <a href="https://www.nickbostrom.com/ethics/infinite.pdf">infinite ethics</a>. Here I tested how similar my decisions in the board game Othello are today compared to 2015. The result was that I chose the same move in 57% of positions for a peculiarity of 0.41 (explained below). The 2015 move was among the 2020 plausible moves in 76% of positions for a plausibility of 0.52 (explained below).</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#motivation">Motivation</a></li>
<li><a href="#another-world">Another World</a></li>
<li><a href="#the-real-large-world">The Real, Large World</a></li>
<li><a href="#self-similarity-experiment">Self-Similarity Experiment</a></li>
<li><a href="#experiment-setup">Experiment Setup</a></li>
<li><a href="#evaluation">Evaluation</a></li>
<li><a href="#results">Results</a></li>
<li><a href="#reservations">Reservations</a></li>
<li><a href="#further-research">Further Research</a></li>
<li><a href="#acknowledgements">Acknowledgements</a></li>
</ul>
</div>
<h2 id="motivation"><a class="toclink" href="#motivation">Motivation</a></h2>
<p>I’ve become interested in what algorithms or computations are and what constitutes their similarity to understand better how large a universe would have to be for evidential cooperation in large worlds (<span class="caps">ECL</span>, or formally “multiverse-wide superrationality,” <span class="caps">MSR</span>) to be interesting. <a href="https://reducing-suffering.org/interpret-physical-system-mind/">Here is an insightful article on the topic.</a> This post only documents a self-experiment I conducted to form some intuitions for the topic. It doesn’t contain any qualitative insights.</p>
<p>In particular it seems to me that either (1) the universe is infinite in some way or (2) it is not. If it’s not, (a) it may be too small for <span class="caps">ECL</span> to be relevant or (b) it may be big enough. In case ii.a, <span class="caps">ECL</span> does not go through; in case ii.b, <span class="caps">ECL</span> is fine. But in case i, which is maybe most likely, there’s a sense in which it is more exposed to infinite ethics than all the rest of aggregative consequentialism in that it both depends and is threatened by infinity. Aggregative consequentialism is only threatened by infinity. This could be alleviated if it turned out that the expected gains from trade from <span class="caps">ECL</span> are considerable even in a universe that is only, say, “<a href="https://www.forbes.com/sites/startswithabang/2018/07/14/ask-ethan-how-large-is-the-entire-unobservable-universe/">15 million times as large as the volume we can observe</a>.” The article suggests that this size and even much greater finite sizes are plausible. Unfortunately, this is mostly about the volume of space and not about the amount of energy in it.</p>
<p>Please see the summaries by Caspar Oesterheld or Lukas Gloor for an introduction to <a href="https://longtermrisk.org/msr">evidential cooperation in large worlds</a>. I’ll give a quick, narrative introduction in the following.</p>
<h2 id="another-world"><a class="toclink" href="#another-world">Another World</a></h2>
<p>In the following, I’ll try to sketch the idea behind <span class="caps">ECL</span> in intuitive terms. This attempt may easily be flawed. So please see Caspar Oesterheld’s paper before deriving strong conclusions about <span class="caps">ECL</span>. I wouldn’t want you to dismiss <span class="caps">ECL</span> because I made a mistake in my presentation or endorse it for reasons it can’t deliver on.</p>
<p>That said, imagine a world that is different from ours in that superrationality is somehow much more widely known and considered than in ours. Maybe Douglas Hofstadter, Gary Drescher,<sup id="fnref:1"><a class="footnote-ref" href="#fn:1">1</a></sup> or someone from the Center on Long-Term Risk who has worked on <span class="caps">ECL</span> has become very famous in that world. (Even more famous than Douglas Hofstadter is in ours already.)</p>
<p>Further imagine, <a href="https://blog.givewell.org/2014/12/02/donor-coordination-and-the-givers-dilemma/">much like in our world</a>, that there are different camps of conscientious altruists – some highly skeptical of all interventions that don’t have a long, well-studied track record of having any effect, and others highly sceptical of such effects. The first group is enthusiastic about GiveWell’s top charities and the second group is enthusiastic about more research into long-term effects. Both of them, however, are even more enthusiastic about the Uncontroversial Missions Force (<span class="caps">UMF</span>), which somehow combines the best of both worlds. These facts are known to both camps.</p>
<p>All of these altruists are very busy, so they just have time to keep well-networked within their intellectual community and don’t engage much with the other community.</p>
<p>In this world the coordination problem goes through roughly two stages:</p>
<ol>
<li>Much like in our world, the altruists plan to donate purely within respective camps and hope that the other will fill the funding gap of the <span class="caps">UMF</span>. Sadly, this gives them evidence that the other altruists will do the same, and <span class="caps">UMF</span> will go underfunded. This is in neither of their interests.</li>
<li>But then they realize that if their acting this way gives them evidence that the other altruists will do the same, then this should hold also in cases where they act differently. “Differently” is quite a wide spectrum, though, so that alone is unhelpful to select the optimal alternative action. But a Schelling point comes to mind: They can fill 50% for the <span class="caps">UMF</span> funding gap. This action seems like enough of a Schelling point that they believe that the other group is fairly likely to choose the same.</li>
</ol>
<p>This might’ve worked without the premise of superrationality being well known, but then the other group is a lot less likely to think of this solution, and that reduces the first group’s confidence that the plan will work.</p>
<h2 id="the-real-large-world"><a class="toclink" href="#the-real-large-world">The Real, Large World</a></h2>
<p><span class="caps">ECL</span> makes one additional crucial assumption: that the universe is infinite in time or space (i.e. that it has infinite volume and matter, not merely without edge, and that the matter is not arranged in some repetitive way), and that any solution to the problem of infinite ethics that we may find or settle on doesn’t upset it more than it will upset all of aggregative consequentialism.</p>
<p>This assumption has two related convenient implications:</p>
<ol>
<li>Not everyone needs to have thought of <span class="caps">ECL</span>. You can just cooperate with those who have.</li>
<li>It doesn’t matter in <span class="caps">ECL</span> whether anyone tries to freeride on the compromise. You can just cooperate with those who don’t.</li>
</ol>
<h2 id="self-similarity-experiment"><a class="toclink" href="#self-similarity-experiment">Self-Similarity Experiment</a></h2>
<p>Going into this, I didn’t know how similar people tend to be or how to measure that, but three years ago, <a href="https://www.facebook.com/Xuenay/posts/10156170094573662">Tobias Baumann</a> came up with an operationalization that at least allowed him to approximate an upper bound on the similarity of two people by comparing different person moments of his own person:</p>
<blockquote>
<p>I just cooked up and carried out a small experiment: I went through archives of online chess games I played last month and sample random positions (filtering those where the best move is obvious) from these games. Then, I think about what move to make, and check whether my favored move matches the move my past self actually played.</p>
<p>The results surprised me a lot. My prior was that I’d play the same move as my past self most of the time (maybe 80% or so). But actually, I chose the same move <em>only 40% of the time</em> (10 out of 25)! Sometimes my past self played “surprising” moves that didn’t even cross my (present) mind.</p>
<p>Of course, I’m not claiming that this is any serious evidence about correlations, but I still found it interesting. Contextual differences (like not <em>actually</em> being in a game, time of day, how exactly the browser window looks, …) are the most plausible explanation for this effect – it’s unlikely that my chess play has radically changed since then. Still, I would have expected my choices to be more “convergent” or stable, and not depend so much on random stuff.</p>
</blockquote>
<p>So I also wanted to test how much you could know about me today based on information about a past version of me.</p>
<p>I don’t know chess, so I couldn’t reproduce this experiment exactly, but I have been playing <a href="https://en.wikipedia.org/wiki/Reversi">Othello</a> occasionally over the past years and know enough strategy to have some seemingly nonrandom feeling about most situations. In particular I was able to find a few dozen games from September 2015, which I used as a basis for my experiment.</p>
<p>My main training in Othello happened in 2005–2006. Since then I’ve played it again during phases of about a month each in 2015, 2018, and now, 2020. There might’ve been more such phases between 2006 and 2015. My impression was generally that I started from a similar level each time, improved slightly over the course of a month, but then lost some of that ability again over the intervening years. So I surely haven’t, across the board (no pun intended), improved since 2015.</p>
<h2 id="experiment-setup"><a class="toclink" href="#experiment-setup">Experiment Setup</a></h2>
<p>I selected 20 games from 2015 in which I played black and an <span class="caps">AI</span> (DroidZebra, today <a href="https://play.google.com/store/apps/details?id=de.earthlingz.oerszebra&hl=en_US">Reversatile</a>) played white. I divided them into segments of first 10 moves, then 11 times 4 moves, and finally the rest of the moves for a total of 240 positions. The rationale was that we played the same openings repeatedly, which would’ve made many of the positions redundant, so I skipped many moves at the beginning until all games were distinct. I also skipped the final moves because I’d often be able to count out the last few moves precisely, which I could’ve also done in 2015, so that differences in performance here would indicate only that I must’ve blitzed the game in 2015 or been too sleepy to care. In one case, I chose a position two moves later than the preordained cut because my next move would’ve been a pass.</p>
<p>I pasted the game transcripts (to me unintelligible alphanumeric strings like f5d6c3d3c4f4c5b4d2e3) into a spreadsheet and set it up such that for each position all further moves were hidden from me. <a href="https://docs.google.com/spreadsheets/d/1e422vRJZzqEJgerGsVytaKnH5xSyTENRI1j8AJiiHr0/edit?usp=sharing">You can find the spreadsheet here.</a> Then I pasted the start of the move sequence that I could see into WZebra and tried to find the best move or moves.</p>
<p>For each position, I recorded one move that I would choose today, all moves that I considered plausible today, the number of all available moves, and afterwards also the perfect moves according to WZebra’s <span class="caps">AI</span>. (AIs are vastly superior to all humans in Othello.) The plausible moves were moves where I imagined I’d be almost unsurprised if they turned out as good or better than the move that I had chosen. I sometimes picked an arbitrary chosen move from among the plausible moves. Originally, I had planned to record the number of all moves that were not obviously terrible (the intersection of the set of all moves that I found obviously terrible and the worst moves according to the <span class="caps">AI</span>), but I couldn’t find any principled threshold for what moves I should consider obviously terrible according to the <span class="caps">AI</span>.</p>
<p>I skipped 90 positions to save time and skipped 6 more because of misclicks that led me to see the perfect moves before I had made my choice. The final sample was one of 144 positions.</p>
<h2 id="evaluation"><a class="toclink" href="#evaluation">Evaluation</a></h2>
<p>The last sheet in the spreadsheet calculates the number of times the 2015 move coincides with the 2020 move; the number of times the 2015 move was among the 2020 plausible moves, and two custom scores. In the following, I’ll call it a hit if the 2015 move coincided with the 2020 move or if the 2015 move was among the 2020 plausible moves.</p>
<p>I called the scores “peculiarity” and “plausibility.” They range from −1 to +1. Peculiarity is just a special case of plausibility, so I’ll focus on explaining the latter. The idea is to weigh a miss by the (negative) additive inverse of the probability of it randomly happening and a hit by the (positive) probability of it randomly not happening. This way, I don’t have to exclude any “obvious” or forced moves from the start but can penalize them later. A forced move has no influence on the score, a hit among 10 possible moves has a strong positive influence on the score, and a miss among 2 possible moves a medium strong negative influence on the score.</p>
<p><span class="math">\(All\)</span>: set of all legal moves at a position</p>
<p><span class="math">\(Plausible\)</span>: set of 2020 plausible moves at a position</p>
<p><span class="math">\(old\)</span>: 2015 chosen move</p>
<div class="math">$$
plausibility =
\begin{cases}
1 - \frac{|Plausible|}{|All|} & \text{if } old \in Plausible \\
- \frac{|Plausible|}{|All|} & \text{if } old \notin Plausible \\
\end{cases}
$$</div>
<p>The peculiarity is just the plausibility where <span class="math">\(Plausible = \{new\}\)</span>.</p>
<h2 id="results"><a class="toclink" href="#results">Results</a></h2>
<ol>
<li>The positions had 1–20 legal moves, 9 on average.</li>
<li>I chose the same move in 57% of positions for a peculiarity of 0.41. (As opposed to 40% in Tobias’s experiment.)</li>
<li>The 2015 move was among the 2020 plausible moves in 76% of positions for a plausibility of 0.52.</li>
<li>I got somewhat worse at the game, picking one of the perfect moves 54% as opposed to 62% of the time.</li>
</ol>
<h2 id="reservations"><a class="toclink" href="#reservations">Reservations</a></h2>
<ol>
<li>I’m quite enthusiastic about <span class="caps">ECL</span>: insofar as unknown biases have crept into my experiment, they more likely overstate my similarity to my 2015 version.</li>
<li>There seems to be less controversy over what makes an Othello move better than there is over what makes a decision theory better. The only disagreements that I could see in the first case are that there may be moves A and B, where A is better than B under perfect play such that:<ol>
<li>The sequence after A may be less intuitive for the player or more intuitive for the opponent than the sequence after B, making B better in some sense, or</li>
<li>A and B are both winning moves under perfect play but the sequence after B allows for fewer losing mistakes than the sequence after A. </li>
</ol>
</li>
<li>This experiment took longer than I thought it was worth, so I resorted to solving positions also when I was somewhat tired. These are unlikely to coincide with the times when I was tired while playing in 2015, a source of lower similarity. This lower similarity is likely uninformative as <span class="caps">ECL</span>-relevant decisions are probably made over the course of months and years rather than hours.</li>
<li>There may be differences between playing a whole game in one go (or rather Othello) compared to solving a particular position, but I don’t know how informative these differences are.<ol>
<li>I sometimes made mistakes because I overlooked pieces in key positions, which I imagine is less likely if my mental picture of the board forms step by step throughout the game.</li>
<li>I may have more time to notice mistakes if I think about a future move or a move sequence over the course of several moves. On the other hand, I also have a less clear picture of the future board setup when it’s still several moves away.</li>
<li>I may overlook mistakes I made visualizing a future board position if I just carry out a preordained move sequence, so just solving one position may solve this issue.</li>
</ol>
</li>
<li>I sometimes got curious about the reasons for why a particular move of mine was a mistake and experimented with different moves to try to understand it. I didn’t realize that I may thus see WZebra’s evaluation of future positions that I would later have to solve. This happened rarely, I avoided it when I noticed the failure mode, and I don’t think I ever directly remembered a perfect move, but I may have remembered general ideas like “Diagonal control is important in this position.” Since my 2015 gameplay was better than today’s, this might push in the direction of overstating similarity.</li>
<li>I noticed thinking patterns along the lines of “I’m unsure between moves A and B. Today, I’d play the more risky one to potentially learn something interesting, but I was more risk averse in 2015 and so probably played the safer one.” This might push in the direction of overstating similarity since I likely remember more about 2015 than I know about most near-copies of mine, but I never tested whether I was even right about my 2015 version.</li>
<li>I tried to find the best moves in every position even if all options were obviously terrible. Back in 2015, I probably gave up on a game at some point and hardly thought about the moves anymore when I became clear that I had lost by a large margin. This may understate my similarity.</li>
</ol>
<h2 id="further-research"><a class="toclink" href="#further-research">Further Research</a></h2>
<p>I’m not planning to prioritize this investigation again any time soon (within a year or so), but if I do, I’d like to:</p>
<ol>
<li>Compare a future version of myself to my today’s version, ideally a shorter time in the future than 5 years but enough to forget the 144 positions.<ol>
<li>This would be more interesting as an upper bound on the similarity between two people at the cost of a bit more risk of overfitting. (Even if I don’t remember the particular positions that I’m solving, I may have learned particular principles this time that I’ll still apply next time. A milder form of overfitting. But I think the risk is low.)</li>
</ol>
</li>
<li>Compare that future version of myself on the basis of solving positions only, not continuous game play vs. solving positions.<ol>
<li>This will happen automatically if I use the same positions.</li>
</ol>
</li>
<li>Select games of people who are around my level and solve positions from their games.<ol>
<li>This way, I can test how different the similarity between my own person moments is compared to interpersonal similarities.</li>
</ol>
</li>
<li>Try to play intentionally similarly to a near-copy of myself.<ol>
<li>In <span class="caps">ECL</span>, the cooperation partners try to agree on a compromise utility function without being able to communicate. They do this not by selecting one action that is the perfect Schelling point because that action would likely be morally irrelevant for most of them. Rather they do this by selecting a bargaining solution that is a Schelling point and apply it to their best guess of the distribution of goals.</li>
<li>If Othello is to serve as a model for that, we can hold constant the equivalent of the bargaining solution and will, by necessity, hold constant the goals. As a result is should be very easy to cooperate with a near-copy by always playing the same moves.</li>
<li>I’d like to test how well this works by thinking about my general strategy, solving some positions with this acausal cooperation goal in mind, waiting a year for my memory to fade, then trying again, and finally comparing the results.</li>
<li>The scoring could be based on points that I get when I play the perfect move or the same move with some downward adjustment when there are several perfect moves.</li>
</ol>
</li>
</ol>
<h2 id="acknowledgements"><a class="toclink" href="#acknowledgements">Acknowledgements</a></h2>
<p>Thanks for feedback and suggestions to Max Daniel and Daniel Kokotajlo. No chiasmus intended.</p>
<div class="footnote">
<hr>
<ol>
<li id="fn:1">
<p>His last name is strangely easy to type for me. <a class="footnote-backref" href="#fnref:1" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
</ol>
</div>
<script type="text/javascript">if (!document.getElementById('mathjaxscript_pelican_#%@#$@#')) {
var align = "center",
indent = "0em",
linebreak = "false";
if (false) {
align = (screen.width < 768) ? "left" : align;
indent = (screen.width < 768) ? "0em" : indent;
linebreak = (screen.width < 768) ? 'true' : linebreak;
}
var mathjaxscript = document.createElement('script');
mathjaxscript.id = 'mathjaxscript_pelican_#%@#$@#';
mathjaxscript.type = 'text/javascript';
mathjaxscript.src = 'https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.3/latest.js?config=TeX-AMS-MML_HTMLorMML';
var configscript = document.createElement('script');
configscript.type = 'text/x-mathjax-config';
configscript[(window.opera ? "innerHTML" : "text")] =
"MathJax.Hub.Config({" +
" config: ['MMLorHTML.js']," +
" TeX: { extensions: ['AMSmath.js','AMSsymbols.js','noErrors.js','noUndefined.js'], equationNumbers: { autoNumber: 'none' } }," +
" jax: ['input/TeX','input/MathML','output/HTML-CSS']," +
" extensions: ['tex2jax.js','mml2jax.js','MathMenu.js','MathZoom.js']," +
" displayAlign: '"+ align +"'," +
" displayIndent: '"+ indent +"'," +
" showMathMenu: true," +
" messageStyle: 'normal'," +
" tex2jax: { " +
" inlineMath: [ ['\\\\(','\\\\)'] ], " +
" displayMath: [ ['$$','$$'] ]," +
" processEscapes: true," +
" preview: 'TeX'," +
" }, " +
" 'HTML-CSS': { " +
" availableFonts: ['STIX', 'TeX']," +
" preferredFont: 'STIX'," +
" styles: { '.MathJax_Display, .MathJax .mo, .MathJax .mi, .MathJax .mn': {color: 'inherit ! important'} }," +
" linebreaks: { automatic: "+ linebreak +", width: '90% container' }," +
" }, " +
"}); " +
"if ('default' !== 'default') {" +
"MathJax.Hub.Register.StartupHook('HTML-CSS Jax Ready',function () {" +
"var VARIANT = MathJax.OutputJax['HTML-CSS'].FONTDATA.VARIANT;" +
"VARIANT['normal'].fonts.unshift('MathJax_default');" +
"VARIANT['bold'].fonts.unshift('MathJax_default-bold');" +
"VARIANT['italic'].fonts.unshift('MathJax_default-italic');" +
"VARIANT['-tex-mathit'].fonts.unshift('MathJax_default-italic');" +
"});" +
"MathJax.Hub.Register.StartupHook('SVG Jax Ready',function () {" +
"var VARIANT = MathJax.OutputJax.SVG.FONTDATA.VARIANT;" +
"VARIANT['normal'].fonts.unshift('MathJax_default');" +
"VARIANT['bold'].fonts.unshift('MathJax_default-bold');" +
"VARIANT['italic'].fonts.unshift('MathJax_default-italic');" +
"VARIANT['-tex-mathit'].fonts.unshift('MathJax_default-italic');" +
"});" +
"}";
(document.body || document.getElementsByTagName('head')[0]).appendChild(configscript);
(document.body || document.getElementsByTagName('head')[0]).appendChild(mathjaxscript);
}
</script>Self-Study Directions 20202020-06-27T20:00:00+00:002020-06-27T20:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2020-06-27:/self-study-directions-2020.html<p>Over the past three years, I’ve collected some 60-odd questions that I now finally have the time to investigate further. I summarize some of them here. This post may be helpful for you if you want to snatch one of these from me and investigate it yourself and helpful for me if you have pointers for any of them.</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#top-priority">Top Priority</a><ul>
<li><a href="#evidential-cooperation-in-large-worlds">Evidential Cooperation in Large Worlds</a><ul>
<li><a href="#summary">Summary</a></li>
<li><a href="#approach">Approach</a></li>
</ul>
</li>
<li><a href="#common-views-on-the-shape-of-the-universe">Common Views on the Shape of the Universe</a><ul>
<li><a href="#summary_1">Summary</a></li>
<li><a href="#approach_1">Approach</a></li>
</ul>
</li>
<li><a href="#considerations-for-independent-researchers">Considerations for Independent Researchers</a><ul>
<li><a href="#summary_2">Summary</a></li>
<li><a href="#approach_2">Approach</a></li>
</ul>
</li>
<li><a href="#working-toward-the-long-reflection-today">Working Toward the Long Reflection Today</a><ul>
<li><a href="#summary_3">Summary</a></li>
<li><a href="#approach_3">Approach</a></li>
</ul>
</li>
<li><a href="#deconfusion-about-trajectories">Deconfusion About Trajectories</a><ul>
<li><a href="#summary_4">Summary</a></li>
<li><a href="#approach_4">Approach</a></li>
</ul>
</li>
<li><a href="#cluelessness-robustness-and-ensembles">Cluelessness, Robustness, and Ensembles</a></li>
<li><a href="#value-of-information-of-priorities-research">Value of Information of Priorities Research</a><ul>
<li><a href="#summary_5">Summary</a></li>
<li><a href="#approach_5">Approach</a></li>
</ul>
</li>
</ul>
</li>
<li><a href="#medium-priority">Medium Priority</a><ul>
<li><a href="#moral-trade-compared-to-regression-analysis">Moral trade compared to regression analysis</a></li>
<li><a href="#effects-of-importance-first-selection">Effects of Importance-First Selection</a></li>
<li><a href="#space-governance">Space Governance</a></li>
<li><a href="#speed-reading-comprehension-rates">Speed Reading Comprehension Rates</a></li>
<li><a href="#green-houses">Green Houses</a></li>
<li><a href="#controlling-evolution">Controlling Evolution</a></li>
<li><a href="#importance-of-the-valley-of-death-effect">Importance of the Valley of Death Effect</a></li>
<li><a href="#classification-systems-for-sandboxing-task-agi">Classification Systems for Sandboxing Task AGI</a></li>
<li><a href="#importance-of-prediction">Importance of Prediction</a></li>
<li><a href="#pros-and-cons-of-funding-mechanisms">Pros and Cons of Funding Mechanisms</a></li>
<li><a href="#tit-for-tat-or-pavlovian-cultures-for-safe-agi">Tit-for-Tat or Pavlovian Cultures for Safe AGI</a></li>
</ul>
</li>
<li><a href="#low-priority">Low Priority</a><ul>
<li><a href="#prevent-stable-totalitarian-regimes">Prevent Stable Totalitarian Regimes</a></li>
<li><a href="#measure-in-hiring">Measure in Hiring</a></li>
<li><a href="#fruit-flies">Fruit Flies</a></li>
<li><a href="#losses-from-taxes">Losses from Taxes</a></li>
</ul>
</li>
<li><a href="#footnotes">Footnotes</a></li>
</ul>
</div>
<h2 id="top-priority"><a class="toclink" href="#top-priority">Top Priority</a></h2>
<h3 id="evidential-cooperation-in-large-worlds"><a class="toclink" href="#evidential-cooperation-in-large-worlds">Evidential Cooperation in Large Worlds</a></h3>
<h4 id="summary"><a class="toclink" href="#summary">Summary</a></h4>
<p><a href="https://longtermrisk.org/msr">Evidential cooperation in large worlds</a> (<span class="caps">ECL</span>, because W is annoying to pronounce) is a form of acausal trade between agents who are sufficiently relevantly similar to one another that their own behavior gives them evidence of the behavior of the others. They don’t need to simulate each other and of course they don’t need to communicate.</p>
<p>If they do it right, they can cooperate with one another and realize gains from trade, for example because different regions of space may lend themselves to realizing particular moral goals particularly cheaply. These gains from trade are maximized by a particular universal compromise utility function. So for all cooperators it’s rational to maximize one joint combination of all their moral goals.</p>
<p>One upshot is that <span class="caps">ECL</span> may significantly limit what moral goals it is rational for an actor to pursue. If this turns out to be sufficiently knowable and important, it’ll be a significant contribution to moral philosophy and priorities research.</p>
<h4 id="approach"><a class="toclink" href="#approach">Approach</a></h4>
<ol>
<li>Understand the existing research by Caspar Oesterheld and Johannes Treutlein.</li>
<li>Understand the requisite decision theory and bargaining theory.</li>
<li>Understand the importance of updatelessness for successful cooperation.<ol>
<li>If we can notice comparative advantages for realizing others’ moral preferences, we can also notice comparative advantages for realizing our own moral preferences. That should be fine?</li>
<li>If <span class="caps">EDT</span> agents are <a href="https://casparoesterheld.com/2016/11/21/thoughts-on-updatelessness/">updateless by implication if that’s what gives them superior news</a> and <span class="caps">UDT</span>/<span class="caps">FDT</span>/<span class="caps">LDT</span> are updateless by design, that doesn’t leave any major <span class="caps">DT</span> that can cooperate evidentially and yet might fail to be updateless, right?</li>
</ol>
</li>
<li>Understand the interplay with infinite ethics. We can’t simultaneously hope/assume that the universe is finite for standard aggregative consequentialism to work and that it’s infinite for us to realize gains from evidential trade. </li>
<li>Estimate the magnitude of the gains from trade, maybe in the ideal case and in the realistic case where we implement the existing robust cooperation heuristics. (Others are already planning to do some version of this.)</li>
<li>Publicize the results and recommendations or make sure particular decision-makers take them into account.</li>
<li>Somewhat improve the allocation of enormous amounts of time and capital that is subject to the decisions of people in or close to the <span class="caps">EA</span> community.</li>
</ol>
<h3 id="common-views-on-the-shape-of-the-universe"><a class="toclink" href="#common-views-on-the-shape-of-the-universe">Common Views on the Shape of the Universe</a></h3>
<h4 id="summary_1"><a class="toclink" href="#summary_1">Summary</a></h4>
<p>Common models in prioritization seem to either be based on the assumption that the universe is finite (standard <a href="https://www.nickbostrom.com/ethics/infinite.pdf">aggregative consequentialism</a>) or infinite (<a href="https://longtermrisk.org/msr">evidential cooperation in large worlds</a>). It is frequently mentioned that an infinite universe – one with infinite volume, not just circular – is the more common assumption among physical cosmologists. But I haven’t double-checked that or tried to understand their reasons for the assumption.</p>
<p>It is also important to make sure that the universe is no perfectly regular beyond a certain point as that would interfere with arguments to the effect that every permutation of particles is found somewhere in the universe (infinitely many times). (H/t Robert Harling.)</p>
<p>Further, it is important that not only space is infinite but also the number of atoms or the amount of energy within. (H/t Inga Großmann.)</p>
<p>Friends of mine also sometimes mention that <a href="https://www.fhi.ox.ac.uk/wp-content/uploads/Anthropic_Decision_Theory_Tech_Report.pdf">Anthropic Decision Theory</a> bears on this assumption. I want to understand this too.</p>
<h4 id="approach_1"><a class="toclink" href="#approach_1">Approach</a></h4>
<ol>
<li>Read up on some physical cosmology.</li>
<li>Refresh myself on Anthropic Decision Theory.</li>
<li>Refresh myself on what Amanda Askell has already said or written on the topic.</li>
</ol>
<h3 id="considerations-for-independent-researchers"><a class="toclink" href="#considerations-for-independent-researchers">Considerations for Independent Researchers</a></h3>
<h4 id="summary_2"><a class="toclink" href="#summary_2">Summary</a></h4>
<p>If you inadvertently duplicate work that produces an easily verifiable result, your work was wasted (apart from skill-building etc.). But if you inadvertently duplicate work that is not verifiable, your result informs the aggregate assessment of the issue: If it is similar, the result more likely to be correct than one would’ve thought based on either data point alone; if it is different, neither data point is quite as likely to be the full answer.</p>
<p>This pushes researchers with a vast network and close collaborators to pursue research that is verifiable, because they can make sure to coordinate with most other people who might duplicate it. It pushes independent researchers without such a network to pursue research that is not easily verifiable.</p>
<p>There are likely to be more such considerations.</p>
<h4 id="approach_2"><a class="toclink" href="#approach_2">Approach</a></h4>
<ol>
<li>Collect considerations for independent researchers: articles, social media posts, interviews, or thinking really hard.</li>
<li>Publish them.</li>
<li>Reprioritize my work accordingly.</li>
</ol>
<h3 id="working-toward-the-long-reflection-today"><a class="toclink" href="#working-toward-the-long-reflection-today">Working Toward the Long Reflection Today</a></h3>
<h4 id="summary_3"><a class="toclink" href="#summary_3">Summary</a></h4>
<p>The <a href="https://forum.effectivealtruism.org/posts/H2zno3ggRJaph9P6c/quotes-about-the-long-reflection">Long Reflection</a> appears to me to be robustly desirable. It only suffers from being more or less unrealistic depending on how it is construed.</p>
<p>In particular, I feel that two aspects of it are in tension: (1) delaying important, risky, and irreversible decisions until after we’ve arrived at a Long Reflection–facilitated consensus on them, and (2) waiting with the Long Reflection itself until after we’ve achieved existential security.</p>
<p>I would expect, as a prior, that most things happen because of economic or political necessity, which is very hard to influence. Hence the Long Reflection either has to ramp up early enough that we can arrive at consensus conclusions and then engage in the advocacy efforts that’ll be necessary to improve over the default outcomes or else risk that flawed solutions get locked in forever. But the first comes at the risk of diverting resources from existential security. This indicates that there is some optimal trade-off point between existential security and timely conclusions.</p>
<p>I’m unsure whether I’m right to think that this is (fairly) urgent from an impartial perspective because I may well be unusually concerned with even moderate risks of lock-ins of terrible futures. This is something that I’ll seek to determine before I’ll try to grapple with the object-level question.</p>
<p>There’s also another sense in which it may not be urgent as the major lock-in that I can foresee concerns settlements beyond our solar system, which may only be attempted many centuries from now or, if sooner, then only after transitions (ems or <span class="caps">AGI</span>) that we should probably rather prepare for explicitly.</p>
<h4 id="approach_3"><a class="toclink" href="#approach_3">Approach</a></h4>
<ol>
<li>Determine whether I have a personal bias in the matter.</li>
<li>Consider whether there are strategies to facilitate the Long Reflection that are robust against major transitions like ems or <span class="caps">AGI</span>.</li>
<li>Consider the next steps.</li>
</ol>
<h3 id="deconfusion-about-trajectories"><a class="toclink" href="#deconfusion-about-trajectories">Deconfusion About Trajectories</a></h3>
<h4 id="summary_4"><a class="toclink" href="#summary_4">Summary</a></h4>
<p>I understand trajectories as projections of world history onto a plane of time and some dimension that we care about, such as total welfare. Whenever there is an indefinite superlinear increase in (say) total welfare, a mere one-time delay or speed-up leads to <a href="https://impartial-priorities.org/cause-selection-2018.html#influencing-the-future">the actual and the counterfactual world diverging increasingly rather than decreasingly over time</a>. I find this highly unintuitive.</p>
<p>Maybe any development is bottlenecked in so many ways that any speed-up (or analogously any delay) only has a minimal effect because some other series of bottlenecks will be hit earlier.</p>
<p>This may not generally be the case (the <a href="https://forum.effectivealtruism.org/posts/n8nXqqgbwp58wuo6a/how-fragile-was-history">fragility of history</a> is likely heterogeneous), and some interventions may affect coefficients other than the constant term, which will lead to even greater speed-ups or retardations if true.</p>
<h4 id="approach_4"><a class="toclink" href="#approach_4">Approach</a></h4>
<ol>
<li>Become less confused about trajectories.</li>
<li>Find out whether this way of thinking about trajectories maps to well-known considerations that I already take into account.</li>
<li>Or otherwise determine what consequences this has for prioritization.</li>
</ol>
<h3 id="cluelessness-robustness-and-ensembles"><a class="toclink" href="#cluelessness-robustness-and-ensembles">Cluelessness, Robustness, and Ensembles</a></h3>
<p>The response to <a href="http://users.ox.ac.uk/~mert2255/papers/cluelessness.pdf">complex cluelessness</a> that I’ve observed in the <span class="caps">EA</span> community over the past five years seems to me to be epitomized by a concentration on robust strategies, strategies that are particularly unlikely to have major problems, e.g., to backfire badly. The general approach is that researchers think about possible problems very long and hard, and then work out whether there are ways to prevent any problems they find.</p>
<p>Consequently, a well-developed proposal for an intervention looks more robust to me the more numerous and nonobvious the problems are (while also covering all the obvious ones) to which there are convincing preventive measures, and the fewer or less likely-seeming the failure modes are for which there are no preventive measures.</p>
<p>This seems almost satisfactory to me. There is the obvious drawback that of course we will have overlooked some problems with the proposals even after thinking about them long and hard. But then we’ll likely overlook problems with all proposals, including the proposal to think even longer and the proposal to give up. If the goal is to weigh the robustness of proposals, such incompleteness is unproblematic.</p>
<p>But it may also be the case that we’re biased in the way we look for problems, so that we may systematically notice the problems with one proposal and overlook the problems with another even though they are actually equally problematic.</p>
<p><a href="https://en.wikipedia.org/wiki/Ensemble_forecasting">Ensemble simulations</a> might help with this. (<a href="https://www.rand.org/content/dam/rand/pubs/monograph_reports/2007/MR1626.pdf"><span class="caps">RAND</span> applied them to environmental policy</a> as a proof of concept. H/t Christian Tarsney.)</p>
<ol>
<li>You put a lot of effort into modeling the mechanics of a toy model of the world.</li>
<li>You simulate hundreds of ways your proposed intervention might turn out.</li>
<li>Only then do you assess how likely and preventable the bad outcomes are.</li>
</ol>
<p>Doing the manual assessment last prevents a lot of unconscious biases from creeping in that would otherwise control which scenarios you’ll think of in the first place.</p>
<p>It seems interesting to apply this to <span class="caps">AI</span> policy.</p>
<h3 id="value-of-information-of-priorities-research"><a class="toclink" href="#value-of-information-of-priorities-research">Value of Information of Priorities Research</a></h3>
<h4 id="summary_5"><a class="toclink" href="#summary_5">Summary</a></h4>
<p>Philosophy has been an incubator of successful fields of science like physics. Maybe priorities research can follow a similar path. If there is any similarity there, a historical comparison may give us some hints of how much we’ll yet update away from our current best guesses as to our impartial priorities.</p>
<p>Our prospective updates may only be of the modest nature that we’ll be able to refute one of the currently plausible worldviews. A large <span class="caps">EA</span>-aligned funder like Open Phil is then likely to update on that. This scenario could serve as a lower bound on the value of information from further priorities research.</p>
<p>This can also inform our assessment of the relative importance of the Long Reflection.</p>
<h4 id="approach_5"><a class="toclink" href="#approach_5">Approach</a></h4>
<p>Very tentatively:</p>
<ol>
<li>Familiarize oneself with the history of philosophy or scholarship in general, select examples that may be representative of priorities research in some way or another, and make inferences from their development to the development of priorities research.</li>
<li>Decide on this basis what the probability distribution looks like over the range from <em>no change of our current idea of impartial priorities</em> to <em>a 180° change of priorities</em>.</li>
<li>Estimate the range of capital that is allocated by funders who are influenced by insights into impartial priorities.</li>
<li>Derive from that the value of information of further research into impartial priorities.</li>
<li>Decide whether to invest more into priorities research or to concentrate on the top priority.</li>
</ol>
<h2 id="medium-priority"><a class="toclink" href="#medium-priority">Medium Priority</a></h2>
<h3 id="moral-trade-compared-to-regression-analysis"><a class="toclink" href="#moral-trade-compared-to-regression-analysis">Moral trade compared to regression analysis</a></h3>
<p>Some clusters of people (epitomized, to me, by <a href="https://docs.google.com/viewer?a=v&pid=sites&srcid=ZGVmYXVsdGRvbWFpbnxuYmVja3N0ZWFkfGd4OjExNDBjZTcwNjMxMzRmZGE">Nick Beckstead’s thesis</a>) conceive of ethical frameworks as regression over an immense set of moral intuitions that we have for particular situations.</p>
<p>To me, it is much more intuitive to think in terms of moral markets where people trade to reap moral (and other) gains from trade. (For example, a person may accept a longer commute so not to work for a company whose business model they view as unethical, or an animal rights activist may support a half-hearted certification scheme if no better certification scheme would stand a chance of being widely adopted.)</p>
<p>It may be interesting in what ways these two approaches differ, and if one of them appears to be clearly more suitable than the other. For example, it may be that there’s some bargaining solution implicit in the first approach, which might be important to realize.</p>
<p>All in all, I don’t consider this a top priority because I’m unsure whether it’ll yield any action-relevant results.</p>
<h3 id="effects-of-importance-first-selection"><a class="toclink" href="#effects-of-importance-first-selection">Effects of Importance-First Selection</a></h3>
<p>I typically first consider the most important projects that I can find (such as suffering risk, permanent dystopian lock-ins, etc.), and then filter further by such criteria as tractability, neglectedness, option value, personal fit, transferable skills, etc.</p>
<p>This approach can easily miss projects whose slightly lacking importance is outweighed by (say) great tractability. Maybe a search pattern that first looks for the most tractable opportunities and then selects the most important ones from among them would produce new high-priority projects.</p>
<p>Some typical criteria, like neglectedness, are not suitable for this role, but a preselection based on tractability seems feasible and resembles the a more general case of the <a href="/prepared-opportunism.html">prepared opportunism</a> that I’ve advocated before.</p>
<p>I can’t currently think of a way of approaching this.</p>
<h3 id="space-governance"><a class="toclink" href="#space-governance">Space Governance</a></h3>
<p>Quoting <a href="https://forum.effectivealtruism.org/posts/QkRq6aRA84vv4xsu9/space-governance-is-important-tractable-and-neglected">Tobias Baumann</a>:</p>
<blockquote>
<p>I argue that space governance has been overlooked as a potentially promising cause area for longtermist effective altruists. While many uncertainties remain, there is a reasonably strong case that such work is important, time-sensitive, tractable and neglected, and should therefore be part of the longtermist <span class="caps">EA</span> portfolio.</p>
</blockquote>
<p>My reasons for not prioritizing it (even) more highly are that I see little personal fit for this work and Tobias’s argument for its urgency sounds urgent only on relatively long timescales:</p>
<blockquote>
<p>But I would, again, argue that establishing good space governance is plausibly time-sensitive. It seems at least possible that our civilisation will start settling other planets in the not-too-distant future – say, within the next centuries.</p>
</blockquote>
<p>But otherwise I find the topic brilliantly interesting!</p>
<h3 id="speed-reading-comprehension-rates"><a class="toclink" href="#speed-reading-comprehension-rates">Speed Reading Comprehension Rates</a></h3>
<p>I like to listen to texts at ~ 400 words per minute using <a href="https://play.google.com/store/apps/details?id=com.hyperionics.avar">text-to-speech software</a>. I seem to have no problem processing the meaning at that rate and like to eagerly recommend it around among friends.</p>
<p>But my ability to read visually is much more limited. I can read at ~ 150 words per minute without technical assistance and at 300–400 words per minute using apps that display only one (or a few) words at a time, but it’s very tiring. (I can do it for at most half an hour before I get tired.) If the text is written in a poetic style that makes it enjoyable on other levels than that of its meaning, then slow reading works fine. If it is like most nonfiction texts, though, reading at a rate of 150 words per minute is supremely boring. The result is that my thoughts drift off tangents all the time and I forget to pay attention to what I’m reading. These thoughts are perhaps sometimes interesting but, in effect, my reading rate across a whole book is lower still.</p>
<p>Naturally, I want to listen to everything now, but I worry that it may have unwanted side effects. I spend less time thinking about the text or thoughts evoked by the text, so:</p>
<ol>
<li>I may not attach it to as many associations and thereby might make it harder to remember,</li>
<li>I may accept it relatively uncritically, and</li>
<li>I may fail to have interesting new ideas inspired by the text.</li>
</ol>
<p>But I can easily reread a text slowly if it seems worth the time. So even though all of those seem quite testable, the result may just be that they are slight effects that may or may not be worth the saved time and effort. That trade-off seems less objectively testable.</p>
<h3 id="green-houses"><a class="toclink" href="#green-houses">Green Houses</a></h3>
<p>Traditional agricultural systems continually kill insects with insecticides which are not optimized to be humane. (<a href="https://reducing-suffering.org/humane-insecticides/">Brian Tomasik</a> and <a href="https://www.wildanimalinitiative.org/blog/humane-insecticides-four-month-update">Wild Animal Initiative</a> on the topic.) Insect populations are usually proportional to the <a href="https://en.wikipedia.org/wiki/Primary_production#GPP_and_NPP">net primary productivity</a>, but I would expect that this case is an exception because the insects are repeatedly killed before they can consume much of the biomass.</p>
<p>Insects, being short-lived, are more likely to have net negative lives than animals with longer lives and similarly painful deaths. I like to define that as a life that a sufficiently idealized version of the individual would not choose to (re-) live if they were given the choice. (Of course there are <a href="https://reducing-suffering.org/small-animals-clock-speed/">many</a> <a href="https://reducing-suffering.org/is-brain-size-morally-relevant/">further</a> <a href="https://casparoesterheld.com/2017/01/18/is-it-a-bias-or-just-a-preference-an-interesting-issue-in-preference-idealization/">complications</a>.) Therefore, insects in lawns and forests may already have net negative lives. But if these areas are used for fields instead, (1) the animals are killed much more often than otherwise, (2) the animals are not clearly fewer at any moment in time, and (3) their deaths are not clearly less painful.</p>
<p>Conversely, agricultural systems like <a href="https://squarerootsgrow.com/">Square Roots</a> or the <a href="https://www.nationalgeographic.com/magazine/2017/09/holland-agriculture-sustainable-farming/">systems used in the Netherlands and Japan</a> physically shield the crops from insects that would otherwise try to consume them. (H/t Inga Großmann.) Apart from accidents (gaps in the shielding) they can prevent the crops from having any influence on wild animal populations thus also rendering their continual killing unnecessary. If these systems are economically viable, they could displace some of the traditional agriculture, increasing total welfare.</p>
<p>The caveat that has led me to deprioritize this topic is that insecticide use <a href="https://en.wikipedia.org/wiki/Decline_in_insect_populations#Suggested_causes">has been suggested as a contributor to the long-term decline in insect populations</a>. If work on this topic would lead to an increase or slowed decline in insect populations, the case for this work becomes a lot less robust and hinges much more on difficult questions about the average welfare of insect populations in the different conditions.</p>
<h3 id="controlling-evolution"><a class="toclink" href="#controlling-evolution">Controlling Evolution</a></h3>
<p>Evolution creates a lot of suffering by creating a lot of feeling individuals in a try-and-error fashion for no particular end. The process is highly immoral and the lack of a goal maybe amoral or also, by omission, immoral.</p>
<p><a href="https://web.archive.org/web/20200618055858/https://slatestarcodex.com/2020/05/12/studies-on-slack/"><em>Slack</em></a><sup id="fnref:scott-alexander-disclaimer"><a class="footnote-ref" href="#fn:scott-alexander-disclaimer">1</a></sup> or the <a href="http://www.overcomingbias.com/2009/09/this-is-the-dream-time.html"><em>dream time</em></a><sup id="fnref:robin-hanson-disclaimer"><a class="footnote-ref" href="#fn:robin-hanson-disclaimer">2</a></sup> enable us to sidestep the worst of evolutionary pressures for a while. While we’re in a dream time–like period, we may have a chance to make it last. Or rather, after the dream time we’ll have no chance anymore.</p>
<p>A Leviathan, say in the shape of an <span class="caps">AGI</span>, might be the solution, but maybe there are other solutions that we could prioritize for redundancy and to allow more people with different strengths to work on different problems without incurring prohibitive coordination costs.</p>
<p>All sorts of selection mechanisms (intelligent foresight, competitions, etc.) that come without suffering may serve to accelerate progress sufficiently that no group that cooperates internally can be crushed by outside forces until all such groups can become one.</p>
<p>I’m not currently prioritizing this topic because its path to impact seems fuzzy to me and it has some risk of touching on controversial issues.</p>
<h3 id="importance-of-the-valley-of-death-effect"><a class="toclink" href="#importance-of-the-valley-of-death-effect">Importance of the Valley of Death Effect</a></h3>
<p>If you conceive of altruists and interventions as forming clusters according to human temperaments and fit for human temperaments, mismatches emerge.</p>
<p>For example, a lot of people will be happy to support Make-A-Wish-like organizations and a lot of people will be happy to support <span class="caps">MIRI</span>-like organizations but some organizations or interventions combine properties in such a way as to be the first, obvious choice for no cluster of altruists at all. (They may well be the second choice for all.) <span class="caps">ALLFED</span> has been cited as an example of such an organization.</p>
<p>It would be interesting to determine whether the resulting neglectedness – sometimes called the Valley of Death – may again make these organizations or interventions particularly attractive or whether it’s insufficient.</p>
<h3 id="classification-systems-for-sandboxing-task-agi"><a class="toclink" href="#classification-systems-for-sandboxing-task-agi">Classification Systems for Sandboxing Task <span class="caps">AGI</span></a></h3>
<p>In a <a href="https://www.fhi.ox.ac.uk/reframing/"><em>comprehensive <span class="caps">AI</span> services</em> (<span class="caps">CAIS</span>)</a> context, it may be important to shield task AIs from some others in systematic ways to prevent failure modes where an <span class="caps">AI</span> that can coordinate other AIs (or a human) gains too much power.</p>
<p>Intelligence agencies have developed classification systems that limit the risks from individuals exposing whatever information they had been authorized to see and sometimes allow them to trace leaks. It would be interesting to understand such systems and transfer them to the <span class="caps">CAIS</span> context.</p>
<p>For example, it may be advisable to (1) never deploy more than one exact copy of a system but always make alterations to the code such that the output is subtly different in ways that are not apparent without comparing two systems, (2) never allowing anyone access to two such near-identical systems once they’re running, and (3) only ever give system access to up to two other systems.</p>
<p>Rule 3 limits the power of any one system. This assumes that each system operates as a black box but that the communication between systems can be monitored. Rules 1 and 2 make sure the overseers can identify the origin of misused information. This should keep the system generally capable while allowing overseers to readily extricate parts of it that are not working their interest.</p>
<p>This feels vague and inchoate to me, but some literature research may shed light on whether ideas in this direction could work and be useful.</p>
<p>Finally, there should also always be encouraged cooperative routes for all systems that are agent-like to achieve their goals so that underhanded methods are both difficult and unnecessary.</p>
<h3 id="importance-of-prediction"><a class="toclink" href="#importance-of-prediction">Importance of Prediction</a></h3>
<p>Forecasting – through superforecasters or prediction markets – seems valuable and popular in the <span class="caps">EA</span> community. But I don’t know just how valuable it is, so some Fermi estimation may be valuable here to trade it off against other things people (like me) could be doing.</p>
<h3 id="pros-and-cons-of-funding-mechanisms"><a class="toclink" href="#pros-and-cons-of-funding-mechanisms">Pros and Cons of Funding Mechanisms</a></h3>
<p>Funding mechanisms such as Patreon/Ko-fi, impact purchases, grants, and different sources of salaries probably all come with advantages and disadvantages. Dimensions include the overhead for the fund managers or donors, overhead for the recipients, control over cannibalization of funding, and dependencies. I feel like I lack an overview of these considerations.</p>
<h3 id="tit-for-tat-or-pavlovian-cultures-for-safe-agi"><a class="toclink" href="#tit-for-tat-or-pavlovian-cultures-for-safe-agi">Tit-for-Tat or Pavlovian Cultures for Safe <span class="caps">AGI</span></a></h3>
<p><a href="https://www.alexandria.unisg.ch/43857/">Herrmann et al. (2008)</a> have found that in games that resemble collective prisoners dilemmas with punishment, cultures worldwide fall into different groups. Those with antisocial punishment fail to realize the gains from cooperation but two other cultures succeed: In the first (cities Boston, Copenhagen, and St. Gallen), participants cooperated at a high level from the start and used occasional punishments to keep it that way. In the second (cities Seoul, Melbourne, and Chengdu), the prior appeared to be low cooperation, but through punishment they achieved after a few rounds the same level of cooperation as the first group.</p>
<p>These two strategies appear to me to map (somewhat imperfectly) to the successful <a href="https://en.wikipedia.org/wiki/Prisoner's_dilemma#Strategy_for_the_iterated_prisoner's_dilemma">Tit for Tat</a> and <a href="https://www.lesswrong.com/posts/3rxMBRCYEmHCNDLhu/the-pavlov-strategy">Pavlov</a> strategies in iterated prisoner’s dilemmas. </p>
<p>Sarah Constantin writes:</p>
<blockquote>
<p>In <a href="https://www.pnas.org/content/pnas/93/7/2686.full.pdf">Wedekind and Milinski’s 1996 experiment</a> with human subjects, playing an iterated prisoner’s dilemma game, a full 70% of them engaged in Pavlov-like strategies. The human Pavlovians were smarter than a pure Pavlov strategy — they eventually recognized the DefectBots and stopped cooperating with them, while a pure-Pavlov strategy never would — but, just like Pavlov, the humans kept “pushing boundaries” when unopposed.</p>
</blockquote>
<p>As mentioned, I think these strategies map somewhat imperfectly to human behavior, but I feel that I can often classify the people around me as tending toward one or the other strategy.</p>
<p>Pavlovian behaviors:</p>
<ol>
<li>Break rules until the cost to yourself from punishments exceeds the profits from the rule-breaking.</li>
<li>View rules as rights for other people to request that you stop a behavior if they disapprove of it. Then stop if anyone invokes a rule.</li>
<li>Push boundaries, the Overton window, or unwritten social rules habitually or for fun, but then take note if someone looks hurt or complains. Someone else merely looking unhappy with the situation is a form of punishment for an empathetic person. (I’m thinking of things like “<a href="https://sanerthanlasagna.wordpress.com/2015/09/11/soft-vs-sharp-culture/">sharp culture</a>.”)</li>
<li>Don’t worry about etiquette because you expect others to give you frank feedback if they are annoyed/hurt/threatened by something you do. Don’t see it as a morally relevant mistake so long as you change your behavior in response to the feedback. (This seems to me like it might be associated with low agreeableness.)</li>
</ol>
<p>Tit for Tat behaviors:</p>
<ol>
<li>Try to anticipate the correct behavior in every situation. Feel remorse over any mistakes.</li>
<li>Attribute rule-breaking, boundary-pushing behaviors to malice.</li>
<li>Keep to very similar people to be able to anticipate the correct behaviors reliably and to avoid being exploited (if only for a short number of “rounds”).</li>
</ol>
<p>This way of categorizing behaviors has led me to think that there are forms of both strategies that seem perfectly nice to me. In particularly, I’ve met socially astute agents who noticed that I’m a “soft culture” tit-for-tat type of person and adjusted to my interaction style. I don’t think it would make sense for an empathetic tit-for-tat agent to adjust to a Pavlovian agent in such a way, but it’s a straightforward self-modification for an empathetic Pavlovian agent.</p>
<p>Further, Pavlovian agents probably have a much easier time navigating areas like entrepreneurship where you’re always moving in innovative areas that don’t have any hard and fast rules yet that you could anticipate. Rather they need to be renegotiated all the time.</p>
<p>Pavlov also seems more time-consuming and cognitively demanding, so it may be more attractive for socially astute agents and for situations where there are likely gains to be had as compared to a tit for tat approach.</p>
<p>The idea is that one type of culture may be safer than another for AIs to learn from through, e.g., inverse reinforcement learning. My tentative hypothesis is that the Pavlovian culture is safer because punishments are small and routine with little risk of ideological, fanatical retributivism emerging.</p>
<h2 id="low-priority"><a class="toclink" href="#low-priority">Low Priority</a></h2>
<h3 id="prevent-stable-totalitarian-regimes"><a class="toclink" href="#prevent-stable-totalitarian-regimes">Prevent Stable Totalitarian Regimes</a></h3>
<p><a href="https://impartial-priorities.org/cause-area-human-rights-in-north-korea.html">North Korea strikes me as a great example of a totalitarian regime straight out of 1984.</a> Its systematic oppression of its citizens is so sophisticated that I could well imagine a world-wide regime of this sort to be stable for a very long time. Even as it exists today, it’s remarkably stable.</p>
<p>The main source of instability is that there’s a world all around North Korea, and especially right to its south, that works so much better in terms of welfare, justice, prospecity, growth, and various moral preferences that are widely shared in the rest of the world.</p>
<p>There may be other sources of instability – for example, I don’t currently understand why North Korea’s currency is inflated to worthlessness – but if not, then we, today, are toward a hypothetical future global totalitarian state what the rest of the world is to North Korea.</p>
<p>Just like some organizations are trying to send leaflets with information about the outside world into North Korea, so we may need to try to send messages into the future just in case a totalitarian dystopia takes hold. These messages would need to be hard to censure and should not depend on people acting against their self-interest to distribute. (Information from most normal time capsules could easily be suppressed.) Maybe a satellite can be set on a course that takes it past earth every century and projects messages against the moon.</p>
<p>This probably not the most cost-effective method, so I’d first like to think about approaches to this more. I’ll leave this under <em>low priority</em> until I have some more realistic ideas.</p>
<h3 id="measure-in-hiring"><a class="toclink" href="#measure-in-hiring">Measure in Hiring</a></h3>
<p>Maybe <em>measure</em>, in the many-worlds interpretation sense, should play a role in hiring decisions, e.g., because if we hire people with relatively low measure, our actions as an organization tell us less about our actions in other branches. But then again we can just assume that everyone has rather high measure, and we’ll be able to cooperate just fine in most branches. Plus, everyone is, ipso facto, likely to have quite high measure unless they’re selected in a special way, for example, if you want to hire the top free solo climber in the world. So it probably makes hardly any difference.</p>
<h3 id="fruit-flies"><a class="toclink" href="#fruit-flies">Fruit Flies</a></h3>
<p>I’d like to know how much I should invest to prevent fruit flies from being born into potentially net-negative lives by leaving my bananas lying around. (E.g., by not leaving my bananas lying around.) I eat a lot of bananas.</p>
<h3 id="losses-from-taxes"><a class="toclink" href="#losses-from-taxes">Losses from Taxes</a></h3>
<p>Tax exemptions usually make a difference of a factor of two at most but some people are probably convince by them to donate to organizations that are likely one or several orders of magnitude less effective. Maybe someone should check how often such misallocations happen and whether a bit more awareness of that can make a difference.</p>
<p>Update: <a href="https://forum.effectivealtruism.org/posts/jq4ENfSYjuD5PbTCv/donating-effectively-does-not-necessarily-imply-donating-tax">This has now been done.</a></p>
<h2 id="footnotes"><a class="toclink" href="#footnotes">Footnotes</a></h2>
<div class="footnote">
<hr>
<ol>
<li id="fn:scott-alexander-disclaimer">
<p>Disclaimer: I don’t generally endorse the works of the author. Alexander originated a wealth of helpful ideas, so that I can’t help but cite him lest it seem that I plagiarize them. Unfortunately, (1) the community around his blog contains some insalubrious factions, and (2) until roughly 2016, he himself still published articles that presented issues in a skewed fashion reminiscent of the very dynamics he warns of in <a href="https://slatestarcodex.com/2014/12/17/the-toxoplasma-of-rage/">Toxoplasma of Rage</a>. I’m adding these disclaimers to avoid the impression that I accept such intellectual wantonness or that it is accepted in my circles. I don’t know whether he still endorses his old approaches. <a class="footnote-backref" href="#fnref:scott-alexander-disclaimer" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
<li id="fn:robin-hanson-disclaimer">
<p>Disclaimer: I don’t generally endorse the works of the author. Hanson originated many useful ideas, so that I can’t help but cite him frequently lest it seem that I plagiarize them. Unfortunately, he seems to be attracted to controversial ideas for controversiality’s sake. But controversial ideas are unusually often insensitive, inaccurate, misleading, or all of these. They need to be treated with special care to make them as uncontroversial as possible. I’m adding these disclaimers to avoid the impression that I accept such intellectual wantonness or that it is accepted in my circles. <a class="footnote-backref" href="#fnref:robin-hanson-disclaimer" title="Jump back to footnote 2 in the text">↩</a></p>
</li>
</ol>
</div>Thoughts: Dream Time2020-06-05T19:35:00+00:002020-06-05T19:35:00+00:00Dawn Dreschertag:impartial-priorities.org,2020-06-05:/thoughts-on-dream-time.html<p>Personal thoughts on the idea of the <em>dream time</em> and the related idea of <em>slack</em>, which describe environments with relatively low competitive pressures.</p><p>Over a decade ago, Robin Hanson wrote about the <a href="http://www.overcomingbias.com/2009/09/this-is-the-dream-time.html"><em>dream time</em></a>,<sup id="fnref:robin-hanson-disclaimer"><a class="footnote-ref" href="#fn:robin-hanson-disclaimer">2</a></sup> and I feel that Scott Alexander’s <a href="https://web.archive.org/web/20200618055858/https://slatestarcodex.com/2020/05/12/studies-on-slack/">“Studies on Slack”</a> has led me to understand the origin of the phenomenon much better.<sup id="fnref:scott-alexander-disclaimer"><a class="footnote-ref" href="#fn:scott-alexander-disclaimer">1</a></sup> The dream time, then, is a period of unusually abundant <em>slack</em>.</p>
<p>There have been studies on how software teams use Slack. Scott Alexander’s article “Studies on Slack” is not about that. Rather it describes the world as a garlic-like nesting of abstraction layers on which there are different degrees of competition vs. cooperation between actors; how they emerged (in some cases); and what their benefit is.</p>
<p>The idea, put simply, at least in my mind, is that in a fierce competition innovations need to prove beneficial immediately in logical time or the innovator will be outcompeted. But limiting innovations to only those that either consist of only one step or whose every step is individually beneficial is, well, limiting. The result are innovators stuck in local optima unable to reach more global optima.</p>
<p>Enter slack. Somehow you create a higher-order mechanism that alleviates the competition a bit. The result is that now innovators have the slack to try a lot of multi-step innovations despite any neutral or detrimental intermediate steps. The mechanisms are different ones in different areas. Scott describes mechanisms from human biology, society, ecology, business management, fictional history, etc. Hence the garlic-like nesting: It seems to me that these systems are nested within each other, and while Scott only ever describes two levels at a time, it’s clear enough that higher levels such as business management depend on lower levels such as those that enable human bodies to function.
This essay made a lot of things clearer to me that I had half intuited but never quite understood. In particular it made me update downward a bit on how much I expect <span class="caps">AGI</span> to outperform humans. One of mine reasons for thinking that human intelligence is vastly inferior to a theoretical optimum is that I thought evolution could almost only ever improve one step at a time – that it would take an extremely long time for a multi-step mutation with detrimental intermediate steps to happen through sheer luck. Since slack seems to be built into biological evolution to some extent, maybe it is not as inferior as
I thought to “intelligent design” like we’re attempting it now.</p>
<p>It would also be interesting to think about how slack affects zero-sum board games – simulations of fierce competition. In the only board game I know, Othello, you can thwart any plans the opponent might have with your next move in, like, 90+% of cases. Hence, I made a (small but noticeable) leap forward in my performance when I switched from analyzing my position through the lens of “What is a nice move I can play?” to “What is a nice move my opponent could now play if it were their turn and how can I prevent it?” A lot of perfect moves, especially early in the game, switch from looking surprising and grotesk to looking good once I viewed them through that lens. So it seems that in Othello there is rarely any Slack. (I’m not saying that you don’t plan multi-step strategies in Othello, but it’s rare that you can plan them such that actually get to carry them out. Robust strategies play a much greater role in my experience. Then again this may be different at higher levels of gameplay than mine.)</p>
<p>Perhaps that’s related to why I’ve seen not particularly smart people yet turning out to be shockingly efficient social manipulators, and why these people are usually found in low-slack fields. If your situation is so competitive that your opponent can never plan more than one step ahead anyway, you only need to do the equivalent of thinking “What is a nice move my opponent could now play if it were their turn and how can I prevent it?” to beat, like, 80% of them. No need for baroque and brittle stratagems like in Skyfall.
I wonder if Go is different? The board is so big that I’d expect there to be room to do whatever for a few moves from time to time? Very vague surface-level heuristic idea! I have no idea of Go strategy.</p>
<p>I’m a bit surprised that Scott didn’t draw parallels to his interest in cost disease, though. Not that I see any clear once, but there got to be some that are worth at least checking and debunking – innovation slowing so that you need more slack to innovate at the same rate, or increasing wealth creating more slack thereby decreasing competition that would’ve otherwise kept prices down, etc.</p>
<p>The article was very elucidating, but I’m not quite able to now look at a system and tell whether it needs more or less slack or how to establish a mechanism that could produce that slack. That would be important since I have a number of <span class="caps">EA</span> friends who could use some more slack to figure out psychological issues or skill up on some areas. <a href="https://forum.effectivealtruism.org/posts/7iptwuSyzDzxsEY5z/the-case-for-impact-purchase-or-part-1">The <span class="caps">EA</span> Funds aim to help here, but I feel like we need more of that.</a></p>
<div class="footnote">
<hr>
<ol>
<li id="fn:scott-alexander-disclaimer">
<p>Disclaimer: I don’t generally endorse the works of the author. Alexander originated a wealth of helpful ideas, so that I can’t help but cite him lest it seem that I plagiarize them. Unfortunately, (1) the community around his blog contains some insalubrious factions, and (2) until roughly 2016, he himself still published articles that presented issues in a skewed fashion reminiscent of the very dynamics he warns of in <a href="https://slatestarcodex.com/2014/12/17/the-toxoplasma-of-rage/">Toxoplasma of Rage</a>. I’m adding these disclaimers to avoid the impression that I accept such intellectual wantonness or that it is accepted in my circles. I don’t know whether he still endorses his old approaches. <a class="footnote-backref" href="#fnref:scott-alexander-disclaimer" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
<li id="fn:robin-hanson-disclaimer">
<p>Disclaimer: I don’t generally endorse the works of the author. Hanson originated many useful ideas, so that I can’t help but cite him frequently lest it seem that I plagiarize them. Unfortunately, he seems to be attracted to controversial ideas for controversiality’s sake. But controversial ideas are unusually often insensitive, inaccurate, misleading, or all of these. They need to be treated with special care to make them as uncontroversial as possible. I’m adding these disclaimers to avoid the impression that I accept such intellectual wantonness or that it is accepted in my circles. <a class="footnote-backref" href="#fnref:robin-hanson-disclaimer" title="Jump back to footnote 2 in the text">↩</a></p>
</li>
</ol>
</div>Thoughts: Free-Riding2020-06-04T18:02:00+00:002020-06-04T18:02:00+00:00Dawn Dreschertag:impartial-priorities.org,2020-06-04:/thoughts-on-free-riding.html<p>Personal thoughts on the article “<a href="https://forum.effectivealtruism.org/posts/54Cdt4BR84vDcki6i/effective-altruism-and-free-riding">Effective Altruism and Free Riding</a>” by Scott Behmer.</p><p>Coordination is an oft-discussed topic within <span class="caps">EA</span>, and people generally try hard to behave cooperatively toward other <span class="caps">EA</span> researchers, entrepreneurs, and donors present and future. But “Effective Altruism and Free Riding” makes the case that standard <span class="caps">EA</span> advice favors defection over cooperation in prisoner’s dilemmas (and stag hunts) with non-EAs. It poses the question whether this is good or bad, and what can be done about it.</p>
<p>I’ve had a few thoughts while reading the article but found that most of them were already covered in the most upvoted comment thread. I’ll still outline them in the following as a reference for myself, to add some references that weren’t mentioned, and to frame them a bit differently.</p>
<p>The project of maximizing gains from moral trade is one that I find very interesting and promising, and want to investigate further to better understand its relative importance and strategic implications.</p>
<p>Still, Scott’s perspective was a somewhat new one for me. He points out that in particular the neglectedness criterion encourages freeriding: Climate change is a terrible risk but we tend to be convinced by neglectedness considerations that additional work on it is not maximally pressing. In effect, we’re freeriding on the efforts of activists working on climate change mitigation.</p>
<p>What was new to me about that is that I’ve conceived of neglectedness as a cheap coordination heuristic. Cheap in that it doesn’t require a lot of communication with other cooperators; coordination in the sense that everyone is working towards a bunch of similar goals but need to distribute work among themselves optimally; and heuristic in that it falls short insofar as values are not perfectly aligned, momentum in capacity building is hard to anticipate, and the tradeoffs with tractability and importance are usually highly imprecise.</p>
<p>So in essence, my simplification was to conceive of the world as filled with agents like me in values that use neglectedness to coordinate their cooperative work, and Scott conceives of the world as filled with agents very much unlike me in values that use neglectedness to freeride off of each other’s work.</p>
<p>Obviously, neither is exactly true, but I don’t see an easy way to home in on which model is better: (1) I suppose most people are not centrally motivated by consequentialism in their work, and it may be impossible for us to benefit the motivations that are central to them. But then again there are probably consequentialist aspects to most people’s motivations. (2) Insofar as there are aspects to people’s motivations for their work that we can benefit, how would these people wish for their preferences to be idealized (if that is even the framing that they’d prefer to think about their behavior)? Caspar Oesterheld discusses the ins and outs of different forms of idealization in the eponymous section 3.3.1 of “Multiverse-wide Cooperation via Correlated Decision Making.” The upshot is, very roughly, that idealization through additional information seems less doubious than idealization through moral arguments (Scott’s article mentions advocacy for example). So would exposing non-EAs to information about the importance of <span class="caps">EA</span> causes lead them to agree that people should focus on them even at the expense of the cause that they chose? (3) What consequentialist preferences should be even take into account – only altruistic ones or also personal ones, since personal ones may be particularly strong? A lot of people have personal preferences not to die or suffer and for their children not to die or suffer, which may be (imperfectly) aligned with catastrophe prevention.</p>
<p>But the framing of the article and the comments was also different from the way I conceive of the world in that it framed the issue as a game between altruistic agents with different goals. I’ve so far seen all sorts of nonagents as being part of the game by dint of being moral patients. If instead we have a game between altruists who are stewards of the interests of other nonagent moral patients, it becomes clearer why everyone is part of the game, their power, but there are a few other aspects that elude me. Is there a risk of double-counting the interests of the nonagent moral patients if they have many altruist stewards – and does that make a difference if everyone does it? And should a bargaining solution only take the stewards’ power into account (perhaps the natural default, for better or worse) or also the number of moral patients they stand up for? The first falls short of my moral intuitions in the case. It may also cause Ben Todd and many others to leave the coalition because the gains from trade are not worth the sacrifice for them. Maybe we can do better. But the second option seems gameable (by pretending to see moral patienthood where one in fact does not see it) and may cause powerful cooperators to leave the coalition if they have a particularly narrow concept of moral patienthood. (Whatever the result, it seems like that this the portfolio that commenters mentioned, probably akin to the compromise utility function that you maximize in evidential cooperation – see <a href="https://longtermrisk.org/files/Multiverse-wide-Cooperation-via-Correlated-Decision-Making.pdf">Caspar Oesterheld’s paper</a>.)</p>
<p>Personally, I can learn a lot more about these questions by just reading up on more game theory research. More specifically, it’s probably smart to investigate what the gains from trade are that we could realize in the best case to see if all of this is even worth the coordination overhead.</p>
<p>But there are probably also a few ways forward for the community. Causal (as opposed to acausal) cooperation requires some trust, so maybe the signal that there is a community of altruists that cooperate particularly well internally can be good if paired with the option of others to join that community by proving themselves to be sufficiently trustworthy. (That community may be wider than <span class="caps">EA</span> and called differently.) That would probably take the shape of newcomers making the case for new cause areas not necessarily based on their appeal to utilitarian values but based on their appeal to the values of the newcomer – alongside an argument that those values wouldn’t just turn into some form of utilitarianism upon idealization. That way, more value systems could gradually join this coalition, and we’d promote cooperation the way Scott recommends in the article. It’ll probably make sense to have different nested spheres of trust, though, with <span class="caps">EA</span> orgs at the center, the wider community around that, new aligned cooperators further outside, occasional mainstream cooperators further outside yet, etc. That way, the more high-trust spheres remain even if sphere’s further on the outside fail.</p>
<p>Finally, a lot of these things are easier in the acausal case that evidential cooperation in large worlds (<span class="caps">ECL</span>) is based on (once again, see <a href="https://longtermrisk.org/files/Multiverse-wide-Cooperation-via-Correlated-Decision-Making.pdf">Caspar Oesterheld’s paper</a>). Perhaps <span class="caps">ECL</span> will turn out to make sufficiently strong recommendations that we’ll want to cooperate causally anyway despite any risk of causal defection against us. This stikes me as somewhat unlikely (e.g., many environmentalists may find <span class="caps">ECL</span> weird, so there may never be many evidential cooperators among them), but I still feel sufficiently confused about the implications of <span class="caps">ECL</span> that I find it at least worth mentioning.</p>Modelers and Indexers2020-05-10T20:00:00+00:002020-05-10T20:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2020-05-10:/modelers-and-indexers.html<p>These are some rather introspective worries of mine to the effect that the longtermist community may be missing out on people with a particular knack for finding counterexamples.</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#memory-and-indexation">Memory and Indexation</a><ul>
<li><a href="#modelers">Modelers</a></li>
<li><a href="#indexers">Indexers</a></li>
<li><a href="#correlation">Correlation</a></li>
</ul>
</li>
<li><a href="#perspective-taking">Perspective-taking</a><ul>
<li><a href="#level-1">Level 1</a></li>
<li><a href="#level-2">Level 2</a></li>
<li><a href="#level-3">Level 3</a></li>
</ul>
</li>
<li><a href="#noticing">Noticing</a></li>
<li><a href="#tying-it-all-together">Tying It All Together</a></li>
<li><a href="#miscellaneous-notes">Miscellaneous Notes</a></li>
<li><a href="#footnotes">Footnotes</a></li>
</ul>
</div>
<h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<p>My mental model of people, which is probably informed or biased by few extreme cases, suggests that people vary widely in the following abilities:</p>
<ol>
<li>
<p>Memory and “indexation”<sup id="fnref:indexation"><a class="footnote-ref" href="#fn:indexation">1</a></sup> or connectedness of memory. That is the degree to which they can immediately think of many relevant experiences and relevant insights to inform a decision. (Maybe they’re also good at inventing great metaphors?) To that end they need to remember the experiences or insights in the first place, but they also need to recall them in connection to the decision.</p>
</li>
<li>
<p>Theory of Mind or conceptual perspective-taking. It seems to me that there are several levels of sophistication here: Level 1 is interpreting a situation from the perspective of a person whose emotional reactions are systematically different from one’s own. Level 2 is interpreting a situation from the perspective of a person who knows more or less than oneself. Level 3 is interpreting a situation from the perspective of a person who has a different model of it, and you’ve not worked together on identifying your cruxes or the cruxes are so fundamental that their implications are likely very different along long inductive chains.<sup id="fnref:levels"><a class="footnote-ref" href="#fn:levels">2</a></sup></p>
</li>
<li>
<p>Noticing. Noticing contradictions is obvious if two contradictory statements follow in quick succession. But if a statement contradicts something that is not related to anything the statement highlights, something that is latent, say, because one has thought about it explicitly for the last time years ago, then it’s much harder to notice the contradiction.</p>
</li>
</ol>
<h2 id="memory-and-indexation"><a class="toclink" href="#memory-and-indexation">Memory and Indexation</a></h2>
<p>When I talk with my friends, who are largely <span class="caps">EA</span> or <span class="caps">EA</span>-adjacent, there is a tendency that they agree with me on the following approach to, let’s say, ontology: (1) You are surprised by something, (2) you update a relevant internal model and propagate updates to related internal models as well as possible, (3) you gradually forget the particular experience that surprised you.</p>
<p>I suspect that a lot of people follow an approach more like this: (1) You are surprised by something, (2) you distill the essence of the surprising situation, some sort of centroid or archetype, something that implies a set of references classes, (3) you file the surprising experience away in an indexed structure that makes it easy to retrieve it again when you’re in a situation that fits the same archetype.</p>
<p>In the end, most people probably do both, but it seems to me that everyone has a tendency in one direction or the other. I surely use both, but in the first case I feel like I have a routine grasp on what it’s like, while in the second case I’m gesturing at something elusive and half-forgotten.</p>
<h3 id="modelers"><a class="toclink" href="#modelers">Modelers</a></h3>
<p>People who excel at the first cognitive style have an easy time generating revolutionary new hypotheses: They take the model they already have and see what it implies. In a sizable minority of the cases, it’ll imply something they haven’t seen yet: They’ll find it obvious to have (cognitive) empathy with beings that they’ve never been, worry about risks of catastrophes that have never happened, and were probably worried about corona early on even without knowing about the Spanish flu.</p>
<p>But if you mix in a lot of openness and a bit of confirmation bias, they may become <a href="https://en.wikipedia.org/wiki/The_Hedgehog_and_the_Fox">hedgehogs</a> espousing theories that are simple, elegant, and wrong. (More wrong than would be a good price to pay for the simplicity.) In self-selecting communities, in particular, they may get continual social proof for such theories from people around them who share the same cognitive style. It is hard to distinguish people who are better than you at something from people who are vastly better than you at it, so decision-makers who specialize in the first cognitive style and are only slightly above average at the second style may undervalue people who are exceptional at it unless these people are <em>also</em> exceptional at other things that the decision-makers <em>can</em> recognize.</p>
<p>For example, if I’m tasked to learn something by heart, I rather try to find some underlying regularity, so I can just rederive it on demand rather than having to laboriously learn it. Interestingly, this feels like it’s high status in my circles, but I just do it because it’s easier. It’s probably good when the generating model is fairly regular and general but bad when the information seems uncompressable or I need to be fast and the rederiving takes too long. My reaction is usually to focus on areas where I can use my cognitive style well and avoid the ones where it is unhelpful. This may be a major part of self-selection in communities.</p>
<p>Quantum mechanics, especially the many-worlds interpretation, and various brands of utilitarianism probably suit this style well.</p>
<h3 id="indexers"><a class="toclink" href="#indexers">Indexers</a></h3>
<p>Meanwhile, people who instead excel at the second cognitive style excel in areas where the information is hard to compress – highly nonlinear or lacking in the sort of structure that lends itself to abstractions vis à vis human cognition (e.g., law, medicine, psychology, sociology, history). Dr. House may be an example. They are great at encountering a situation and immediately thinking of all the relevant associated bits of information – maybe the archetype that certain symptoms boil down to, what diagnoses are plausible given the symptom combination archetype, and how to test for them. Or if the situation is a hypothetical implication of someone else’s model, they are great at finding counterexamples: For every implication of a model, they can distill the archetype of the situation and retrieve countless examples of how such situations have played out in the past. If these experiences tend to contradict the implication, they can warn that something is off about the model. If the situation has never happened before, they may just have to distill a bit harder to find an archetype whose reference class does contain a sufficient number of experiences.</p>
<p>But they have no grand models that would allow them to derive implications that have never happened (though curiously they seem to be good at the deriving itself when you give them a model), and they may be unduly sceptical of models that have unprecedented implications in the rare cases where these are actually about to happen: They’ll increasingly widen the reference class that their archetype implies until it starts to contain instances, but then these are fairly unrepresentative so that the fraction of confirming and disconfirming instances is uninformative. If the implication is unprecedented, maybe wide reference classes that contain it are also still full of rare events, so maybe the disconfirming instances are even systematically overrepresented. (E.g., there has never been a virus like <span class="caps">COVID</span>-19, so they expand the reference class to <span class="caps">SARS</span>, <span class="caps">MERS</span>, ebola, and the seasonal flu, and get uninformative results.)</p>
<p>The field that I classified as “highly nonlinear or lacking in the sort of structure that lends itself to abstractions vis à vis human cognition (e.g., law, medicine, psychology, sociology, history)” are known for degrees that involve a lot of memorizing. People who excel in these fields probably excel at the second type of cognition. In these areas, the second cognitive style excels because it can actually work efficiently with all the chaotic information that is so hard to compress.</p>
<p>The areas that are attractive to them are probably necessarily named for the problems rather than the solution concepts because the solution concepts are so diverse. Above are some fields that seem like good examples to me.</p>
<p>I’m not calling these people “foxes,” because foxes seem to me more like the people who manage to combine the best of both worlds, not extremes in either direction.</p>
<h3 id="correlation"><a class="toclink" href="#correlation">Correlation</a></h3>
<p>Self-selection of people with one or the other cognitive style into communities like effective altruism or exogenous selection of them into <span class="caps">EA</span> organizations may be unproblematic if these styles are highly correlated with general competence and general competence is what determines the selection. Then most people should be about equally good at both with a few having a slight preference one way or the other.</p>
<p>But if (1) people really vary widely in this regard, or (2) the self-selection is heavily influenced by cognitive styles or (3) the exogenous selection is based on markers of a particular cognitive style, then one group will end up being vastly overrepresented.</p>
<p>I have the unsubstantiated hypothesis that risk factors 1 and 2 are likely true and risk factor 3 may be in a few cases:</p>
<p>Risk factor 1: My own thinking, noticeably in my childhood, has been rather hedgehog-like. It was only in my adolescence that I started seeing this tendency of myself as a bias and questioned it. Conversely, I’ve known something who, due to a brain injury, had sustained some disabilities and also scored around <span class="caps">IQ</span> 60–70 in some unnamed <span class="caps">IQ</span> test. At the time we met, she was doing her PhD in anthropology. What I remember of our extensive conversations was that I usually enjoyed them for bouncing new, weird hypotheses off of her, which she usually promptly destroyed with a wealth of counterexamples that cut at the core of my hypotheses. These counterexamples were what inspired my term “archetype” – they weren’t specific situations that could be dismissed as isolated exceptions to an otherwise sound rule but situations where it was clear from the structure of the situation that they contradicted my model. She couldn’t muster any interest for <span class="caps">EA</span>.</p>
<p>Risk factor 2: Here I’m drawing on my impressions of my friend circles. I can easily think of friends of mine who are strong modelers. I can only think of a few strong indexers, and they’re also pretty strong modelers. I don’t think I currently know anyone with a really pronounced indexer preference.</p>
<p>Risk factor 3: Here I’m just guessing that if there’s a preponderance of modelers, that may also be true of position where people influence hiring decisions, so that they may be less able to recognize exceptional indexers than they’re able to recognize exceptional modelers. But at the one organization where I know people better, at least a few top people are also really good indexers. Maybe indexers are more efficient in operations roles.</p>
<h2 id="perspective-taking"><a class="toclink" href="#perspective-taking">Perspective-taking</a></h2>
<p>Perspective-taking is a separate skill. I find that it comes in different difficulty levels. I’ll use the first person here, because I haven’t checked whether others would order these differently.</p>
<h3 id="level-1"><a class="toclink" href="#level-1">Level 1</a></h3>
<p>If I already know in what specific ways someone else is different, and the implications of these differences are straightforward, then I can adjust for them easily. Say, I’m concerned what others think about me and someone else is not concerned what others think about them. Now the other person is treated disrespectfully in front of a group. I might’ve been worried about how that’ll affect how the groups sees me, but I can infer that the other person, being less concerned about that, will be more straightforwardly cross with the perpetrator.</p>
<h3 id="level-2"><a class="toclink" href="#level-2">Level 2</a></h3>
<p>Level 2 resembles the marble test: The subject sees that someone puts a marble in a place and leaves, and that someone else then places the marble in a different place. The first person returns and the subject is asked they will look for the marble.</p>
<p>The more challenging version goes: You’ve studied computer science and worked as a software engineer for a decade; another smart person has studied psychology and worked as a psychology researcher for a decade. They ask you what database migrations are. How much do you have to simplify and analogize your explanation so that they can <em>just</em> understand it but <em>not</em> so much that they feel patronized?</p>
<p>A different version of this is more about expectations than about knowledge. I meet a new person, and don’t know whether they like small talk. They start small talk. Can I reciprocate the small talk? I want to reciprocate small talk iff they are more likely to enjoy than not to enjoy small talk.</p>
<p>My prior is that someone who likes small talk is more likely to open with small talk and someone who doesn’t is less likely to. Also someone who dislikes small talk may be less likely to assume that I enjoy small talk than someone who likes it, so if their prior of me is that I like small talk, they’re more likely to open with small talk. And if they realize this and think that I will model them as someone who likes small talk if they open with small talk, then they likely think that liking small talk is something good, so they probably either like or want to train small talk. So this points in the direction of updating toward them liking small talk. This consideration feels like a positive but asymptotic feedback loop at approaches some limit. Other people are perhaps much better at intuiting where that limit is.</p>
<p>But it’s only an implication, so they might also start small talk for other reasons than liking small talk. It might be that they only started small talk because they thought it was polite but don’t actually like it. This complicates the situation and leads to a much slighter update toward them liking small talk. (Am I right to think that the update is still positive?) Putting in an optional small talk escape hatch (“I’ve been really enjoying reading about what decision theories are implicit in the training of AIs. What have you been up to?”) seems to help to generate more evidence, but I suppose it might be taken as a cue for the other that you want to end the small talk, so it’s hard to make it truly neutral.<sup id="fnref:smalltalk"><a class="footnote-ref" href="#fn:smalltalk">3</a></sup></p>
<p>I suppose that indexers, and with a bit of practice about anyone, no longer think through all of these considerations but just act according to a behavior that fits the situational archetype, which they have stored.</p>
<p>But it gets more complicated when several people are involved. A simple, linear version is the joke: Three mathematicians enter a bar. The bartender asks, “Would you all like a beer?” The first mathematician: “I don’t know.” The second one: “I don’t know.” The third one: “Yes.” But that only works absent any noise from social customs, politeness, pragmatics, and reasoning styles. The joke would not work with “Three people who don’t know each other enter a bar.”</p>
<p>I found a more complex and noncooperative demonstration of this reasoning problem in the scene in Dumbledore’s office in <a href="https://www.hpmor.com/chapter/18">chapter 18 of Harry Potter and the Methods of Rationality</a> (but this chapter does not make sense in isolation) and later in the book in the complexities of the three-army game. It’s difficult to imagine that some people have been in all possible permutations of these situations to have meaningful achetypes stored.</p>
<p>A wholly different solution is the <a href="https://www.lesswrong.com/posts/3rxMBRCYEmHCNDLhu/the-pavlov-strategy">Pavlov strategy</a> where you just do what you want and watch carefully how the other person reacts. If they seem unhappy with the situation (the altruistic “punishment”), you adjust your behavior until you’re both happy.</p>
<h3 id="level-3"><a class="toclink" href="#level-3">Level 3</a></h3>
<p>Level 3 is similar to level 2 except that the difference is not so much one of knowledge than one of world models.</p>
<p>For example, someone makes a statement, even a really long and comprehensive one, that strikes me as somehow confused but in a confusing way that I can’t make sense of. What I don’t know is that they subscribe to <span class="caps">CDT</span>, dualism about consciousness, moral objectivism, and a sort of essentialism about personal identity, none of which I share. But they don’t know these terms and can’t tell me. From within the framework of these concepts, insofar as they form a coherent framework, their statement was straightforward. But to understand that, I first have to recognize all the above cruxes and then take the perspective of someone whose model of the world is so different.</p>
<p>Once we’ve (collaboratively) identified all the cruxes, this should resemble Level 1. But I’ve found that it can be much more difficult than Level 1 if I’m not practiced at using the different model. Over the years of using the model, the other person will have noticed various implications the model has for religion, population ethics, and social customs. But I, being new to it, may be either unaware of these implications or unaware that some of my previous conclusions in these matters don’t hold under the new framework.</p>
<h2 id="noticing"><a class="toclink" href="#noticing">Noticing</a></h2>
<p>There is something magical about noticing something without paying attention to it. Maybe some System 1 background process is constantly paying attention to thousands of things and notifies System 2 if something odd transpires? Or rather maybe there’s a hierachy of such processes that each pay attention to one more and more fine-grained type of oddity?</p>
<p>In any case, if you are working on a task, your whole attention is on the task, and you get completely immersed in the intricacies of the task, then it may sometimes be hard to notice when you start working on the task for a longer time than the result is worth. Unless you explicitly timeboxed it from the start.</p>
<p>Or in an interview, you might concentrate on the interview partner, on asking interesting questions, on avoiding questions that would force them to mention things they can’t talk about, on coming up with the best spontaneous follow-up questions, etc. It may be hard to notice when the interviewee uses a term the audience may not know or when you start to go more in depth on a topic than it deserves given the limited airtime you have. (I think the <span class="caps">TV</span> show <a href="https://en.wikipedia.org/wiki/Monk_(TV_series)">Monk</a> was largely about noticing completely ordinary things that in the particular situation under the generally accepted assumptions should or should not have happened.)</p>
<h2 id="tying-it-all-together"><a class="toclink" href="#tying-it-all-together">Tying It All Together</a></h2>
<p>All three of these abilities are crucial for collaborative truth-seeking – or in particular collaborative mistake-spotting.</p>
<p>Most situations are probably unlikely and uninformative, so it will be hard to derive truths from a model that it doesn’t imply or that it even contradicts. But if you have a repository of truths and you have a method of indexation that allows you to look up relevant ones at will, it becomes more likely that you can find the pertinent counterexamples.</p>
<p>If the person who has the model is not you yourself, you need to first understand another person’s model – e.g., find all the cruxes – and its implications before you can bring to bear your skill at finding pertinent counterexamples.</p>
<p>And most such models are never discussed as explicitly as during a double crux exercise. Rather they’re latent, and sometimes something as fleeting as the connotation of one word or other in a fast-paced conversation on some orthogonal topic can reveal an unknown difference between one’s own and the other person’s model. A difference that may result in a lot of wasted work if it goes unnoticed.</p>
<h2 id="miscellaneous-notes"><a class="toclink" href="#miscellaneous-notes">Miscellaneous Notes</a></h2>
<ol>
<li>
<p><span class="caps">AI</span> safety, priorities research still strike me as so fragmented that indexers may actually do well in them, but I suppose that’s more because of their relative novelty.</p>
</li>
<li>
<p>University friends of mine (none of who are today associated with <span class="caps">EA</span> or Less Wrong) reported different experiences reading the Haskell 98 Report (Haskell 2010 wasn’t out yet) compared to some other books on Haskell: I found it amazingly pleasant reading because they put the definition first and then added some examples to confirm that it applies as expected in some edge cases. Some university friends of mine found a book much more instructive that contained many examples but almost no definitions. It was almost as if they didn’t need the definitions to learn to write their code. When I had read the same book, I tried to interpolate the definition from the examples, which was amazingly cumbersome compared to just reading the Report.</p>
</li>
<li>
<p>I suppose this blog post is “self-anti-suggesting.” I should downgrade my confidence in it for the very reasons it lists. Confusingly, that means that I’m less obliged to take it seriously, so that I also don’t need to downgrade my confidence in it as much. But I suppose there’s some limit between my felt confidence and the confidence after the first update that this feedback cycles updates me towards?</p>
</li>
<li>
<p>There’s a trope of students learning things by heart for exams that they were supposed to derive during the exam. Is there a corresponding meme of students needlessly rederiving things during exams that could’ve just memorized in some other fields I’m not familiar with?</p>
</li>
<li>
<p>The <span class="caps">EA</span> type of thinking seems to be uncommon and highly attuned to modelers. Self-selection effects tend to be <a href="https://slatestarcodex.com/2014/09/30/i-can-tolerate-anything-except-the-outgroup/">surprisingly extreme</a> (see part <span class="caps">III</span>).<sup id="fnref:scott-alexander-disclaimer"><a class="footnote-ref" href="#fn:scott-alexander-disclaimer">4</a></sup> So a small self-selected group within a large population can probably still self-select strongly against the population base rate?</p>
</li>
<li>
<p>There’s also <a href="https://www.lesswrong.com/posts/dC7mP5nSwvpL65Qu5/why-the-tails-come-apart">this funny effect</a>, so maybe the selection effect is relevant even if indexer and modeler abilities are highly correlated, e.g., because the top people in either group will have an outsized impact, will be modelers, and will be (per this effect) highly preferential modelers?</p>
</li>
<li>
<p>It’s a bit hard to keep track of just how well-informed a model is if information only comes in once a year, and you don’t have exceptional memory. So maybe modelers have much higher variance in terms of over- or underconfidence in their models. If it is overconfidence, they may get stuck with some only-locally-optimal model. Indexers who also model may be much quicker to discard their models and start from scratch, also because they still have the almost-raw data that informed the model. So they may be more “conservative” in that they are less willing to jump to hypothetical conclusions but less “conservative” in that they are less likely to cling to a bad model.</p>
</li>
<li>
<p>A teacher of mine once implied that he assumed that everyone categorized everyone else within seconds of meeting them. That had seemed very foreign to me. Interestingly, he was a history and language teacher.</p>
</li>
<li>
<p>There is the structurally reminiscent distinction between the inside and the outside view. But it seems orthogonal to my distinction here.</p>
</li>
<li>
<p>Here is <a href="https://www.lesswrong.com/s/qqFS6Kw5fmPyzkLby/p/wP2ymm44kZZwaFPYh">Eliezer Yudkowsky working hard to understand someone with a funny world model</a>. So what I called level 3 of perspective taking.</p>
</li>
<li>
<p>Maybe people don’t start out with a strong indexer or modeler disposition but it’s a feedback loop that increasingly turns them into one. They start doing one or the other a bit more, and so, say, they have a model that works fairly well for them, so they use it, so it becomes even more useful, so they use it more. Meanwhile, they’ve always discarded their situational memories rather than memorizing them, so they don’t have much of a useful repository of those, so the ability to memorize these things atrophys, and so on. Well, and conversely for the indexer.</p>
</li>
<li>
<p>Or maybe there is some innate effect, say, because people with more intelligence than memory (according to some sort of population percentile let’s say) tend to become modelers while people with more memory than intelligence become indexers – well, and the majority of people have both in similar amounts and so don’t develop a particular disposition.</p>
</li>
</ol>
<h2 id="footnotes"><a class="toclink" href="#footnotes">Footnotes</a></h2>
<div class="footnote">
<hr>
<ol>
<li id="fn:indexation">
<p>This is a bit of a jargon term here. What I mean is the same principle that an index of a book employs. It tries to list every word anyone might likely look for and for each word lists the pages where it features. This way, you can easily find every occurrence of the word. The alternative would be to leave through and skim every page of the book. <a class="footnote-backref" href="#fnref:indexation" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
<li id="fn:levels">
<p>Please ignore the exact classification if it seems unhelpful. There are a few challenges that I can’t clearly classify either, so it’s probably not very good. <a class="footnote-backref" href="#fnref:levels" title="Jump back to footnote 2 in the text">↩</a></p>
</li>
<li id="fn:smalltalk">
<p>I somehow still feel a vague sense of embarrassment in situations that are clearly mutual, so this type of reasoning doesn’t seem to come naturally to my System 1, or I’m making a mistake somewhere. <a class="footnote-backref" href="#fnref:smalltalk" title="Jump back to footnote 3 in the text">↩</a></p>
</li>
<li id="fn:scott-alexander-disclaimer">
<p>Disclaimer: I don’t generally endorse the works of the author. Alexander originated a wealth of helpful ideas, so that I can’t help but cite him lest it seem that I plagiarize them. Unfortunately, (1) the community around his blog contains some insalubrious factions, and (2) until roughly 2016, he himself still published articles that presented issues in a skewed fashion reminiscent of the very dynamics he warns of in <a href="https://slatestarcodex.com/2014/12/17/the-toxoplasma-of-rage/">Toxoplasma of Rage</a>. I’m adding these disclaimers to avoid the impression that I accept such intellectual wantonness or that it is accepted in my circles. I don’t know whether he still endorses his old approaches. <a class="footnote-backref" href="#fnref:scott-alexander-disclaimer" title="Jump back to footnote 4 in the text">↩</a></p>
</li>
</ol>
</div>Thoughts: Future People and Near-Term Effects2020-04-22T06:24:00+00:002020-06-28T14:47:00+00:00Dawn Dreschertag:impartial-priorities.org,2020-04-22:/thoughts-on-future-people.html<p>Personal thoughts on the article “<a href="https://forum.effectivealtruism.org/posts/ajZ8AxhEtny7Hhbv7/if-you-value-future-people-why-do-you-consider-near-term">If you value future people, why do you consider near term effects?</a>” by Alex Holness-Tofts.</p><p>“If you value future people, why do you consider near term effects?” by Alex Holness-Tofts makes the case that a lot of reasons to focus on near-term effects fall short of being persuasive. The case is based centrally on complex cluelessness. It closes with a series of possible objections and why they are not persuasive. (Alex also cites the amazing article “<a href="https://forum.effectivealtruism.org/posts/bsE5t6qhGC65fEpzN/growth-and-the-case-against-randomista-development">Growth and the case against randomista development</a>”!)</p>
<p>I find it disconcerting that there are a lot of very smart people in the <span class="caps">EA</span> community who focus more on near-term effects than I currently find reasonable. It creates a tension between my assessment of the question before and after Aumann update, i.e. taking into account that I’m more likely to be wrong when a lot of smart people disagree with me.</p>
<p>The article invites a discussion, and Michael St. Jules responded by explaining the shape of a utility function (bounded above and below) that would lead to a near-term focus and why it is a sensible utility function to have. The number of upvotes lead me to believe that this is a common reason for near-term focus, but Michael notes in a <a href="https://www.facebook.com/drescher.denis/posts/10157314365728691?comment_id=10157317913063691">comment on my Facebook post</a>:</p>
<blockquote>
<p>I’d actually guess that a bounded utility function is an uncommon reason for neartermist focus, based on the other answers and the way shortermist <span class="caps">EA</span> orgs do cost-effectiveness analysis (risk-neutrally), so people are upvoting because they find my comment interesting. I think I’ve only heard it endorsed explicitly a few times by EAs.</p>
<p>I’d guess different person-affecting views might explain some neartermism, but except for those who actually think future/extra individuals, including those in pure misery (s-risks), at most barely matter or don’t matter at all, this wouldn’t be a good objection to longtermism on its own.</p>
<p>I think this is probably the closest to what neartermists would endorse (I left the comment late compared to others, so it might not have gotten much attention because of it):</p>
<p>“My guess is that people who support <span class="caps">AMF</span>, <span class="caps">SCI</span>, or GiveDirectly don’t think the negative long-term effects are significant compared to the benefits, compared to ‘doing nothing.’ These do more good than harm under a longtermist framework. Compared to ‘doing nothing,’ they might generally just be skeptical of the causal effects of any interventions primarily targeting growth and all other so far proposed longtermist interventions (the causal evidence is much weaker) or believe these aren’t robustly good because of complex cluelessness.</p>
<p>“I focus on animal welfare, and it’s basically the same argument for me. If I think doing X is robustly better than doing nothing in expectation, and no other action is robustly better than doing X in expectation, then I’m happy to do X.”</p>
</blockquote>
<p><a href="https://forum.effectivealtruism.org/posts/ajZ8AxhEtny7Hhbv7/if-you-value-future-people-why-do-you-consider-near-term?commentId=7zKXMAQ8zPicuSHc4">See also this conversation between Michael and Phil Trammell.</a> I revisit this conversation and <a href="https://philiptrammell.com/static/simplifying_cluelessness.pdf">Phil’s essay on the matter</a> and may then update this post.</p>
<p>There are also hints in the discussion of whether there may be a reason to focus on near-term effects as a Schelling point in coordination problem with future generations. But that point is not fully developed, and I don’t think I could steelman it.</p>
<p>I’ve heard smart people argue for the merits of bounded utility functions before. They have a number of merits – avoiding Pascal’s mugging, the St. Petersburg game, and more. (Are there maybe even some benefits for dealing with infinite ethics?) But they’re also very unintuitive to me.</p>
<p>Besides, I wouldn’t know how to select the right parameters for it. With some parameters, it’ll be nearly linear in a third-degree-polynomial increase in aggregate positive or negative valence over the coming millennium, and that may be enough to prefer current longtermist over current near-termist approaches.</p>
<p>Another related article is Brian Tomasik’s <a href="https://longtermrisk.org/how-the-simulation-argument-dampens-future-fanaticism">“How the Simulation Argument Dampens Future Fanaticism”</a>:</p>
<blockquote>
<p>There’s a non-trivial chance that most of the copies of ourselves are instantiated in relatively short-lived simulations run by superintelligent civilizations, and if so, when we act to help others in the short run, our good deeds are duplicated many times over. Notably, this reasoning dramatically upshifts the relative importance of short-term helping even if there’s only a small chance that Nick Bostrom’s basic simulation argument is correct.</p>
</blockquote>Donation Decisions of 20182018-11-10T05:30:00+00:002018-11-10T05:30:00+00:00Dawn Dreschertag:impartial-priorities.org,2018-11-10:/donation-decisions-2018.html<p>A summary of my donation decisions of 2018. Please see my previous posts from the year for more reasoning behind the tentative conclusions I draw here.</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#animal-rights">Animal Rights</a></li>
<li><a href="#suffering-risks">Suffering Risks</a></li>
<li><a href="#welfare-biology">Welfare Biology</a></li>
<li><a href="#prioritization-research">Prioritization Research</a></li>
<li><a href="#assorted-remarks">Assorted Remarks</a><ul>
<li><a href="#matching-fundraisers">Matching Fundraisers</a></li>
<li><a href="#donation-splitting">Donation Splitting</a></li>
<li><a href="#donation-budget">Donation Budget</a></li>
</ul>
</li>
</ul>
</div>
<h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<p>I’ve written about my <a href="https://claviger.net/cause-selection-2018.html">general thoughts on prioritization</a> in more depth earlier this year. This is a summary of the donations that have followed from that for me.</p>
<h2 id="animal-rights"><a class="toclink" href="#animal-rights">Animal Rights</a></h2>
<p>One of the themes of my in-depth prioritization post has been uncertainty over how optimistic we should be about being able to influence the future intentionally. And that analysis even explicitly ignored <a href="https://foundational-research.org/how-the-simulation-argument-dampens-future-fanaticism">implications of the simulation hypothesis</a>, namely, in slogan form, that most simulations may be run for a purpose and discontinued when they’ve served that purpose so that most of them are vastly more short-lived than the root universe. So if we’re in a simulation, then chances are that it is also comparatively short-lived.</p>
<p>If further research makes me more pessimistic about this potential, I expect to prioritize the animal rights space more highly since it seems that it is a space where short-term impact can be found and where a lot of very valuable preparatory work has already been invested. By “short term,” I mean in this context something like one or several centuries. I currently expect this short-term impact to be about two orders of magnitude more cost-effectively realizable than short-term impact through GiveWell top charities.<sup id="fnref:bases"><a class="footnote-ref" href="#fn:bases">1</a></sup></p>
<p>I continue to recommend Animal Equality by default to donors who had never been exposed to effective altruist thought before. It strikes a good balance between being high impact and convincing to pitch in ten seconds, and it has the added benefit that I can easily transition from the pitch to a conversation about how to select charities using the research of <a href="https://animalcharityevaluators.org/donation-advice/recommended-charities/">Animal Charity Evaluators</a> (<span class="caps">ACE</span>). Their Stuttgart office is also very conveniently located. I like to experiment with alternative formats – supporting Wild Animal Suffering Research (<span class="caps">WASR</span>) and the Foundational Research Institute (<span class="caps">FRI</span>) – at smaller conventions but don’t see sufficiently practical alternatives yet to be worth the risk of another switch. (We switched from my own charity to the Against Malaria Foundation and then to Animal Equality.)</p>
<p>Generally, I consider it a solid approach to short-term donation investment to follow the <a href="https://animalcharityevaluators.org/donation-advice/recommended-charities/">top charity recommendations of <span class="caps">ACE</span></a>, which will soon be renewed.</p>
<p>Among <span class="caps">ACE</span>’s standout charities, Open Cages strikes me as an unusually underfunded organization, though I don’t know whether that may have changed over the past year. I feel similarly positively about most other standout charities.</p>
<p><span class="caps">ACE</span> itself is also an organization that I prioritize highly within the space. They’ve received a lot of unwarranted public criticism over the past year, so that I followed their fundraisers more closely to see whether I need to help fill a gap. <a href="https://blog.givewell.org/2014/12/02/donor-coordination-and-the-givers-dilemma/">Such waiting is not something I recommend in general</a> since it makes fundraising unnecessarily slow and unpredictable for the charities. As detailed below, I’ve applied a better methodology when I supported <span class="caps">FRI</span> earlier this year.</p>
<p>Another great benefit of <span class="caps">ACE</span> is that they can investigate factors such as their <a href="https://animalcharityevaluators.org/charity-reviews/evaluating-charities/evaluation-criteria/">criteria 6 and 7</a>, “Does the charity have strong leadership and a well-developed strategic vision?” and “Does the charity have a healthy culture and a sustainable structure?” While the strategic vision may be similarly possible for outsiders to determine as most other of their criteria are (though also at great time cost), evaluating these two crucial criteria may require getting to know people at the organization personally or even working or volunteering for them. Such involvement would be prohibitively costly for most potential individual donors.</p>
<p>One organization that I wish to highlight that has not yet received the standout designation but that I’ve seen put out very interesting research over the past year or two is the Sentience Institute in the <span class="caps">US</span>, founded by Kelly Witwicki and Jacy Reese, the latter of whom has just <a href="https://smile.amazon.com/End-Animal-Farming-Entrepreneurs-Animal-Free/dp/0807019453">published a book</a>.<sup id="fnref:jacy-anthis-disclaimer"><a class="footnote-ref" href="#fn:jacy-anthis-disclaimer">2</a></sup></p>
<h2 id="suffering-risks"><a class="toclink" href="#suffering-risks">Suffering Risks</a></h2>
<p>But I’m not quite pessimistic enough to prioritize such short-term campaigns just yet, so I also support organizations that seek to have a more long-term impact, by which I mean millennia and more. (After this stage, if we continue to exist, our ability to communicate with other civilizational clusters will likely deteriorate because of the distances between them, so that we’ll increasingly lose the leverage over the future that we now still have.)</p>
<p>I count as suffering risks not only those that result from a catastrophic event but also gradual developments that involve an exacerbation of suffering. Therefore there is really a substantive overlap between all spaces in this point, for example due to an exacerbation of animal suffering from space colonization.</p>
<p>Suffering risks, according to this wide definition, are for <a href="https://claviger.net/robustness-through-suffering-focus.html">my moral system</a> almost ipso facto the top priority, so the more interesting question is whether the particular activities that I can fund in this space are worthwhile.</p>
<p>The Foundational Research Institute (<span class="caps">FRI</span>) has some exciting new ideas in its pipeline that I want to make sure will come to fruition. At this point, its funding situation seems more secure, but earlier in the year this was much less certain because it was contingent on a few events that are influenced by randomness.</p>
<p>These events were mediated by donors who wanted to fund <span class="caps">FRI</span>. If these events turn out one way, the donors would support <span class="caps">FRI</span> sufficiently. Otherwise they would not.</p>
<p>It would’ve been easy to wait for these donors to either make their donation or not and then to decide on the basis of that information whether I wanted to donate, but that would first be a defection against them because I’d be unfairly exploiting their informational generosity and would secondly, over a longer time, encourage mutual waiting and leaving charities unfunded. (I’d always get to see all my favorite charities funded while they may not care about the charity that I fund from the money I don’t need to donate to <span class="caps">FRI</span> thanks to their donation and they don’t get to fund their other favorites.)</p>
<p>Ideally, I would’ve liked to resolve that by donating as much as they did in expectation, but that would’ve been hard to determine, I wasn’t privy to much of the necessary data, and I’m probably not rich enough either, so I needed a different solution. It also needed to be simple to make it more likely that it represented a decent Shelling point because I couldn’t communicate with the other donors. (I don’t know who they are.) My solution was to determine how much I would donate if it turned out that the funding gap had remained, and then to donate half of that right away and only make the second half contingent on the donation decision of the others.</p>
<p>Hence my decision to fund <span class="caps">FRI</span> with another four-digit donation (in <span class="caps">CHF</span>) in early 2018.</p>
<h2 id="welfare-biology"><a class="toclink" href="#welfare-biology">Welfare Biology</a></h2>
<p>The charities in this space that I’m aware of are Wild Animal Suffering Research (<span class="caps">WASR</span>), Animal Ethics, and Utility Farm. <span class="caps">WASR</span> seems to be fairly well funded at the moment and I know relatively little about Utility Farm, so my attention has been mostly to Animal Ethics in recent months. Generally, they all seem to be so well coordinated that it may make sense to apply some lessons from my <a href="https://claviger.net/concept-for-donor-coordination.html">donor coordination methodology</a> and support the space rather than the individual charity by filling funding gaps wherever they open. When several people coordinate in this fashion, all organizations can flourish and the impact is fungible between all donors or employees.</p>
<p>The recent focus on incentivizing academic research into welfare biologoy seems valuable to me. Animal Ethics has spearheaded it. Don’t be fooled by their strangely anonymous and sometimes rather dry website – they used to prioritize outreach to animal activists also through leafletting but have updated away from that approach. The <a href="https://animalcharityevaluators.org/advocacy-interventions/interventions/leafleting/">leafletting report of Animal Charity Evaluators</a> (basically, there’s still no evidence that leafletting has a net positive effect) and the general trend toward reaching out to thought leaders may have been causal factors. Animal Ethics’ updated budget for 2018 also reflects a strong shift away from the probably ineffective campaigns.</p>
<p>I’m very happy about this development but I also worry that none of the organizations in the space seem to base their prioritization on the important criterion which interventions can plausibly continue to have a positive effect on wild animals after human civilizational collapse. This omission may be owed to the very early stage of their work where capacity building is still a much greater concern, but even so I’ve seen disproportionately specific work come out of these organizations already, so they could just as well have looked into the very important long-term effects of possible interventions instead.</p>
<p>That said, I’ve supported <span class="caps">WASR</span> and Animal Ethics with four-digit donations (in <span class="caps">CHF</span>). (The first of which I transferred last year, for tax reasons, but took from this year’s donation budget.) I don’t currently prioritize this area anymore, but that may easily change again.</p>
<h2 id="prioritization-research"><a class="toclink" href="#prioritization-research">Prioritization Research</a></h2>
<p>Over the course of the year, I’ve noticed that I have great difficulty deciding how to allocate my donations because the existing research on prioritization leaves me not only with as many <a href="https://claviger.net/prioritization-research-ideas-2019.html">preliminary</a> <a href="https://claviger.net/was-research-ideas-2019.html">questions</a> as preliminary answers but with rather disorganized questions too.</p>
<p>My personal focus in my free time has therefore been to develop structures to organize and trade off the most important considerations in the space. Whether this is the most important thing to do is itself an open question, but it seems at least plausible to me, so that I want to invest the remainder of my donation budget to help researchers in this area off the ground.</p>
<p>So far, I’m aware of two groups in this space that might accept donations, the Global Priorities Institute (<span class="caps">GPI</span>) and Rethink Priorities (<span class="caps">RP</span>). I’ll wait to learn more about their respective funding gaps and update this post. So far, my guess is that <span class="caps">GPI</span> is much better funded than <span class="caps">RP</span>.</p>
<h2 id="assorted-remarks"><a class="toclink" href="#assorted-remarks">Assorted Remarks</a></h2>
<h3 id="matching-fundraisers"><a class="toclink" href="#matching-fundraisers">Matching Fundraisers</a></h3>
<p>In most matching fundraisers it’s intransparent whether the matching donor or donors have a more or less explicit donation budget and will donate the same amount to a similar intervention whether or not the match ceiling is reached. Some matching fundraisers are set up differently so that the counterfactuals are transparent, but as a general rule, I ignore the matching aspect of matching fundraisers and advise charities and matching donors to champion transparency in this department. See also “<a href="https://forum.effectivealtruism.org/posts/a2gYyTnAP36TxqdQp/matching-donation-fundraisers-can-be-harmfully-dishonest">Matching-donation fundraisers can be harmfully dishonest</a>” and “<a href="https://www.benkuhn.net/matching-results">Donation matching survey results</a>.”</p>
<h3 id="donation-splitting"><a class="toclink" href="#donation-splitting">Donation Splitting</a></h3>
<p>I’ve often split my donations between two to four recipients, <a href="https://fragile-credences.github.io/splitting/">which may not make any sense</a>, but then again also doesn’t seem very costly.<sup id="fnref:splitting-costs"><a class="footnote-ref" href="#fn:splitting-costs">3</a></sup> This is of course only true if the alternative where to select one at random. If I actually had a preference between them, I should donate all to the top favorite organization because in that case the subjective expected costs would otherwise be significant.<sup id="fnref:exploratory-donations"><a class="footnote-ref" href="#fn:exploratory-donations">4</a></sup></p>
<p>The actual split above is because I don’t donate in any particular rhythm and changed my opinion a few times over the course of the year because I learned more and because the landscape changed. But even now I would probably select at random between two organizations if I don’t split my donation between them.<sup id="fnref:subjective-expected-value"><a class="footnote-ref" href="#fn:subjective-expected-value">5</a></sup></p>
<p>My reasons include improving coordination (see the link above) and signaling<sup id="fnref:signaling"><a class="footnote-ref" href="#fn:signaling">6</a></sup> (which I think is honest and transparent<sup id="fnref:transparency"><a class="footnote-ref" href="#fn:transparency">7</a></sup>), but I suspect that there is a strong emotional component where I’m too excited to resist supporting a very promising project. (But I would if I thought it came at a sufficient cost.)</p>
<h3 id="donation-budget"><a class="toclink" href="#donation-budget">Donation Budget</a></h3>
<p>I’m holding back on donating quite as much as I set out to donate because I don’t see any group that does <em>exactly</em> the things that I think it would be <a href="https://claviger.net/prioritization-research-ideas-2019.html">most important to do</a>. If it emerges, I want to be able to support it, or otherwise found it myself.</p>
<div class="footnote">
<hr>
<ol>
<li id="fn:bases">
<p>That’s in base ten. In base two it’s six to seven orders of magnitude. <a class="footnote-backref" href="#fnref:bases" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
<li id="fn:jacy-anthis-disclaimer">
<p>Disclaimer: I don’t condone the author’s conduct. (I have no first-hand information, but it sounds to me like verbal sexual harassment.) Jacy authored many informative articles and a book and runs an organization with a promising strategic vision, so that I can’t help but cite him lest it seem that I plagiarize them. But I’m adding these disclaimers to avoid the impression that I accept such conduct or that it is accepted in my circles. <a class="footnote-backref" href="#fnref:jacy-anthis-disclaimer" title="Jump back to footnote 2 in the text">↩</a></p>
</li>
<li id="fn:splitting-costs">
<p>The processing of the extra donation by operations and accounting at charities may cost some CHF 10–50 or so depending on how often it has to be forwarded and how much of the process is automated. <a class="footnote-backref" href="#fnref:splitting-costs" title="Jump back to footnote 3 in the text">↩</a></p>
</li>
<li id="fn:exploratory-donations">
<p>I’ve also made exploratory donations of CHF 500 or less to signal that I’m seriously considering supporting an organization and it’s worth some staff time to engage with me more so I can learn more about it. I’m ignoring these here. <a class="footnote-backref" href="#fnref:exploratory-donations" title="Jump back to footnote 4 in the text">↩</a></p>
</li>
<li id="fn:subjective-expected-value">
<p>The expected value given infinite intelligence, perfect rationality, complete knowledge, and no limits in time is probably not anywhere close to equal, but I don’t have those. In fact I have very little time. That leads to high variance, and then I can’t tell two options apart anymore. <a class="footnote-backref" href="#fnref:subjective-expected-value" title="Jump back to footnote 5 in the text">↩</a></p>
</li>
<li id="fn:signaling">
<p>If I’m a charity and ten people apply for jobs and ten people ask questions about the organization to decide whether to donate to me, and all of these questions will take a lot of time to answer, then I will want to decide who is really serious about it and who merely wants to fulfill job center requirements or wants to donate CHF 10 if they’re convinced. Absent more information, I treat all requests equally, but if a high-quality cover letter, someone’s reputation, or an up-front donation tells me that they’re serious, I can allocate my time more efficiently. I want to give such costly, honest signals and hope it will enhance our internal coordination. <a class="footnote-backref" href="#fnref:signaling" title="Jump back to footnote 6 in the text">↩</a></p>
</li>
<li id="fn:transparency">
<p>Tom Sittler writes: “From a rule-consequentialist perspective, it may be better to always be fully transparent, and not to make donations decisions based on how they will affect what others think of us.” If this is true, it would be a decisive reasons for me not to split, but I think my approach is completely transparent according to my rule “Act only according to that maxim whereby you can at the same time will that everyone had known all along that you would act that way.” Also the expected value of deciding via unbiased coin flip and splitting equally is equal, so splitting is actually more transparent in that one extra organization can update on it. <a class="footnote-backref" href="#fnref:transparency" title="Jump back to footnote 7 in the text">↩</a></p>
</li>
</ol>
</div>Wild Animal Suffering Research Ideas 20192018-10-28T18:30:00+00:002018-10-28T18:30:00+00:00Dawn Dreschertag:impartial-priorities.org,2018-10-28:/was-research-ideas-2019.html<p>Important questions related to wild animal suffering that I wish I had the time to research.</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#taxonomy-of-resilience-of-wild-animal-welfare-improvements-after-civilizational-collapse">Taxonomy of resilience of wild animal welfare improvements after civilizational collapse</a><ul>
<li><a href="#idea">Idea</a></li>
<li><a href="#relevance">Relevance</a></li>
</ul>
</li>
<li><a href="#invertebrate-deaths-per-crop-yield">Invertebrate deaths per crop yield</a><ul>
<li><a href="#idea_1">Idea</a></li>
<li><a href="#relevance_1">Relevance</a></li>
</ul>
</li>
<li><a href="#maximum-amount-of-exoplanet-surface-area-reachable-over-time">Maximum amount of exoplanet surface area reachable over time</a><ul>
<li><a href="#idea_2">Idea</a></li>
<li><a href="#relevance_2">Relevance</a></li>
</ul>
</li>
<li><a href="#accidental-exodus-will-it-be-simple-not-to-inadvertently-export-invertebrates-into-space">Accidental Exodus: Will it be simple not to inadvertently export invertebrates into space</a><ul>
<li><a href="#idea_3">Idea</a></li>
<li><a href="#relevance_3">Relevance</a></li>
</ul>
</li>
<li><a href="#are-suffering-levels-resilient">Are suffering levels resilient?</a><ul>
<li><a href="#idea_4">Idea</a></li>
<li><a href="#relevance_4">Relevance</a></li>
</ul>
</li>
<li><a href="#academic-movement-building-publishing-sweet-spots">Academic Movement Building: Publishing Sweet Spots</a><ul>
<li><a href="#idea_5">Idea</a></li>
<li><a href="#relevance_5">Relevance</a></li>
</ul>
</li>
</ul>
</div>
<h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<p>As a spin-off off <a href="https://claviger.net/prioritization-research-ideas-2019.html">my research ideas on prioritization</a>, here more object-level research ideas for wild animal suffering alias welfare biology.</p>
<p>If you’re interested in researching one of these questions, please leave a comment or ask me to edit the post so we don’t duplicate effort.</p>
<h2 id="taxonomy-of-resilience-of-wild-animal-welfare-improvements-after-civilizational-collapse"><a class="toclink" href="#taxonomy-of-resilience-of-wild-animal-welfare-improvements-after-civilizational-collapse">Taxonomy of resilience of wild animal welfare improvements after civilizational collapse</a></h2>
<div class="admonition type">
<p class="admonition-title">Questions</p>
<ol>
<li>Which interventions against WAS are most permanent even without sustained human effort, especially in the case of civilizational collapse?</li>
<li>Are there downsides to focusing completely on such interventions, especially with regard to option value?</li>
</ol>
</div>
<h3 id="idea"><a class="toclink" href="#idea">Idea</a></h3>
<p>Interventions (1) whose effects subside when they are not actively carried out anymore, (2) whose effects may be undone by a catastrophe, or (3) that are premised on structures that may be destroyed by a catastrophe are unlikely to be cost-effective compared to interventions that reduce more event-like suffering risks because interventions that are not robustly self-sustaining will have an effect over only a very short period of time in expectation. (Some people put the probability of an <a href="http://effective-altruism.com/ea/1rk/current_estimates_for_likelihood_of_xrisk/">existential catastrophe before 2100 at 19%</a>.)</p>
<p>My intuition is that this (the inferior cost-effectiveness) will remain true even if we find that “<a href="https://claviger.net/prioritization-research-ideas-2019.html#resilience_1">shmingletons</a>” are relatively volatile after all or that the tractability of such risk reduction efforts is rather low. In combination, these may be stronger challenges. The greatest single challenge I’m aware of stems from the <a href="https://foundational-research.org/how-the-simulation-argument-dampens-future-fanaticism">simulation hypothesis</a> because most simulations are likely to be short-lived.</p>
<p>These three categories may not be the only failure modes, and there may also be different paths to success, so it would be helpful to create an exhaustive taxonomy of interventions that classifies them according to their cost-effectiveness based only on this one important factor.</p>
<p>A consideration that could be taken into account that pushes in the opposite direction is option value. If we build highly resilient systems not even we may be able to change them later on.</p>
<h3 id="relevance"><a class="toclink" href="#relevance">Relevance</a></h3>
<p>The resilience of the positive change of an intervention should be <a href="https://claviger.net/cause-selection-2018.html#durability_1">one major criterion</a> for at least shallow analyses of its cost-effectiveness. Such a rubic could speed up the assessment and help us avoid overlooking important factors.</p>
<h2 id="invertebrate-deaths-per-crop-yield"><a class="toclink" href="#invertebrate-deaths-per-crop-yield">Invertebrate deaths per crop yield</a></h2>
<div class="admonition type">
<p class="admonition-title">Questions</p>
<ol>
<li>Are greenhouses only suitable or economically viable for some vegetable species but very low in suffering?</li>
<li>Which crops promise the greatest absolute suffering reduction given current farming methods and optimal, economically viable methods?</li>
</ol>
</div>
<h3 id="idea_1"><a class="toclink" href="#idea_1">Idea</a></h3>
<p>Even though the <a href="https://reducing-suffering.org/net-primary-productivity-land-type/">net primary productivity (<span class="caps">NPP</span>) of cultivated land is not maximal</a> compared to plausible counterfactuals such as forests (greater <span class="caps">NPP</span>) or grassland (similar <span class="caps">NPP</span>), the expected insect suffering caused by crop farming may be many times as great as that in forests because the animals attracted by the plants are continually killing off by farmers before they can influence the <span class="caps">NPP</span> of the land, so that it remains to attract more animals.</p>
<p>This may be different in organic farming systems that are limited in terms of the insecticides that they can use, but they may just use animals that attack the animals that try to feed on the crops, so that the effect on the aggregate suffering may be the same. If the yield is lower but demand is unchanged, the aggregate suffering may even be greater.</p>
<p>I’m more optimistic that this is different in greenhouse systems that physically shield the crop <span class="caps">NPP</span> from any animals. Insofar as the shielding is effective, this approach should make the <span class="caps">NPP</span> virtually disappear (ignoring the waste products that its human consumers will leave behind). <a href="https://www.nationalgeographic.com/magazine/2017/09/holland-agriculture-sustainable-farming/">Such systems can also be highly efficient and economically viable.</a></p>
<p>These system may only be suitable or economically viable for some species of plants, so it’ll be important to determine what the counterfactual is for the production of these particular species, for which species the potential for suffering reduction is greatest, and how this compares to other interventions to reduce insect suffering such as malaria prevention or buying grassland to cover it with gravel.</p>
<h3 id="relevance_1"><a class="toclink" href="#relevance_1">Relevance</a></h3>
<p>This area has potential for social enterprises, e.g., for-profits porting technologies that have been proven to be economically viable in the Netherlands or Japan to other markets. A for-profit could also produce products needed by companies implementing the most impactful greenhouse system if that approach is easier to bootstrap.</p>
<h2 id="maximum-amount-of-exoplanet-surface-area-reachable-over-time"><a class="toclink" href="#maximum-amount-of-exoplanet-surface-area-reachable-over-time">Maximum amount of exoplanet surface area reachable over time</a></h2>
<div class="admonition type">
<p class="admonition-title">Questions</p>
<ol>
<li>What formula best describes the volume of the three-dimensional shape of the expansion of earth-originating life into space over shorter timelines – thousands or tens of thousands of years rather than billions?</li>
<li>What formula best describes the energy output per time of the stars in that volume?</li>
</ol>
</div>
<h3 id="idea_2"><a class="toclink" href="#idea_2">Idea</a></h3>
<p>I model the expansion into space as spherical or at least as third-degree polynomial on the basis that <a href="https://en.wikipedia.org/wiki/Observable_universe#Large-scale_structure">space is homogeneous at the largest scale</a>. (See also <a href="https://en.wikipedia.org/wiki/File:Observable_Universe_with_Measurements_01.png">this image</a>.) But that only starts “at scales between 30 and 200 megaparsecs,” or 100–700 million lightyears, so 50–350 million lightyears in either direction. Assuming we can’t travel quite at light speed, it may be a billion years before the approximation becomes quite justified. (Maybe our local group or supercluster are already homogenous enough inside, but there we’re already at a scale of millions of lightyears.)</p>
<p><a href="https://claviger.net/cause-selection-2018.html#influencing-the-future_2">My observations</a>, even the most optimistic ones, don’t promise that we can influence much more than the next 10,000 years. These observations may be flawed because they can’t observe any counterfactuals, but combined with the probably common intuition that influences don’t last forever, it may be valuable to have a better idea of the shape of earth-originating life in the universe over just the next 10,000s of years.</p>
<p>Additionally, whole-brain emulation may make biological life redundant at some point, so that environments in space will cease to be optimized for it, so that unintentional suffering will be minimized anyway at some point. This may happen even sooner than 10,000 years from now.</p>
<h3 id="relevance_2"><a class="toclink" href="#relevance_2">Relevance</a></h3>
<p>If our expansion into space will be much more circular than spherical or have some other shape with orders of magnitude smaller volume, then efforts to influence the aggregate suffering from space colonization will be radically less important than it currently seems. Those are some of the most effective interventions in the wild animals suffering space, indicating that this is an important consideration.</p>
<p>But this consideration is likely to be irrelevant in cases where space colonization is not run by biological life (either because of whole-brain emulation or because we’ve been superseded), which seems increasingly likely the further technology advances.</p>
<p>A related consideration may remain relevant, namely that the suffering of silicon-based beings will be upper-bounded by the available energy, which will be correlated with the number of solar system reachable over time. The expansion model could cover this case as well by including the energy output of the reachable suns.</p>
<h2 id="accidental-exodus-will-it-be-simple-not-to-inadvertently-export-invertebrates-into-space"><a class="toclink" href="#accidental-exodus-will-it-be-simple-not-to-inadvertently-export-invertebrates-into-space">Accidental Exodus: Will it be simple not to inadvertently export invertebrates into space</a></h2>
<div class="admonition type">
<p class="admonition-title">Questions</p>
<ol>
<li>Is it hard to avoid accidentally carrying invertebrates (such as small insects or their eggs) into space? How hard is it?</li>
<li>What animals can survive and procreate in environments that exist aboard uncrewed spacecrafts?</li>
</ol>
</div>
<h3 id="idea_3"><a class="toclink" href="#idea_3">Idea</a></h3>
<p><a href="https://sentience-politics.org/files/Dello-Iacovo-On-terraforming-wild-animal-suffering-and-the-far-future.pdf">Michael Dello-Iacovo</a> has written about the risks from intentionally and accidentally exporting nonhuman animals to space in the course of space colonization (with a focus on terraforming, which may not be necessary in the long run, e.g., if we move to real or simulated environments in Dyson spheres).</p>
<p>It <a href="https://www.theatlantic.com/science/archive/2018/06/bacteria-can-eat-the-cleaning-products-nasa-uses-to-sterilize-its-spaceships/562016/">seems to be hard</a> but <a href="https://en.wikipedia.org/wiki/Interplanetary_contamination">important</a> to keep bacteria out of spacecrafts, and if there are astronauts on board, it may get even more difficult to keep life out, perhaps even insects or their eggs.</p>
<p>But how hard is it exactly? Once a large exodus from earth starts, even a small risk may lead to lots of insects being exported and then multiplying at high rates.</p>
<h3 id="relevance_3"><a class="toclink" href="#relevance_3">Relevance</a></h3>
<p>If animal farming is superseded by a more efficient technology and we don’t prioritize terraforming, or if we move away from biological life entirely, the remaining risk to wild animals (that I’m aware of) is from being exported into space by accident. This scenario is not as conjunctive as that sentence may imply as it is an additional consideration in all other cases as well.</p>
<p>If these animals are unlikely to get onto spacecrafts or unlikely to survive in space, we can be somewhat less concerned about space colonization.</p>
<h2 id="are-suffering-levels-resilient"><a class="toclink" href="#are-suffering-levels-resilient">Are suffering levels resilient?</a></h2>
<div class="admonition type">
<p class="admonition-title">Question</p>
<ol>
<li>Is it plausible that there are any feedback mechanisms that increase suffering again, be it indirectly, if we reduce it or reduce it the wrong way?</li>
</ol>
</div>
<h3 id="idea_4"><a class="toclink" href="#idea_4">Idea</a></h3>
<p>If we want to reduce insect suffering, we might think that we should euthanize insects (regardless of actually relevant factors such as whether they’re about to lay eggs or are in great pain), but if their population is bottlenecked by their sustenance, we just free up edible biomass to be eaten by another insect some generations later that wouldn’t have existed otherwise. (This relationship may be less than one-to-one.)</p>
<p>Similarly, subsidies may lead to export of the subsidized product, and thus to market distortions and costs disproportionate to the benefits.</p>
<h3 id="relevance_4"><a class="toclink" href="#relevance_4">Relevance</a></h3>
<p>I’m worried that there may be some mechanisms by which efforts to minimizes suffering may be systematically less effective than we think because they (the mechanisms) respond to the existing level of suffering or its rate of increase or decrease. This would make these efforts at suffering minimization greatly less tractable.</p>
<p>For example, per individual, there are probably adaptive mechanisms whereby (1) they adapt to a safer environment by becoming more sensitive within their lifetime or (2) they adapt to a safer environment by becoming more sensitive evolutionarily because more sensitive individuals can perform better within their culture.</p>
<p>But unless these changes span orders of magnitude of individual suffering over ones lifetime, it may be more relevant to consider population-level dynamics, e.g., are there any mechanisms that cause the number of animals to increase more quickly the lower it is?</p>
<h2 id="academic-movement-building-publishing-sweet-spots"><a class="toclink" href="#academic-movement-building-publishing-sweet-spots">Academic Movement Building: Publishing Sweet Spots</a></h2>
<div class="admonition type">
<p class="admonition-title">Questions</p>
<ol>
<li>What attributes make species, topics, and attributes (such as welfare) interesting for researchers?</li>
<li>What makes a species or topics are interesting for welfare biology?</li>
<li>Where are the sweet spots where a hypothetically easy-to-study species that is highly relevant for welfare biology has remained understudied?</li>
</ol>
</div>
<h3 id="idea_5"><a class="toclink" href="#idea_5">Idea</a></h3>
<p>To build more momentum for welfare biology research, it may be important to understand what makes a species and behaviors of a species attractive for academic researchers. (E.g., what makes drosophila so popular?)</p>
<p>Then we can determine what the main criteria are that make a species or topic interesting for welfare biology – e.g., being r-selected, likely to become more frequent due to global warming, hard to keep out of space missions, short lives and violent deaths, and reinforcement learning behavior.</p>
<p>Finally, we can put it all together and look for publishing sweet spots where a hypothetically easy-to-study species that is highly relevant for welfare biology has remained understudied.</p>
<h3 id="relevance_5"><a class="toclink" href="#relevance_5">Relevance</a></h3>
<p>Incentivizing research in welfare biology has been a priority of several charities in the space, and for good reasons, I think, because of my considerations surrounding the <a href="https://claviger.net/cause-selection-2018.html#fundamentality">robustness of research impact</a>.</p>Prioritization Research Ideas 20192018-10-14T18:10:00+00:002018-10-28T10:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2018-10-14:/prioritization-research-ideas-2019.html<p>Important questions related to prioritization that I wish I had the time to research.</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#value-of-prioritization-research">Value of Prioritization Research</a><ul>
<li><a href="#introduction_1">Introduction</a></li>
<li><a href="#relevance">Relevance</a></li>
<li><a href="#assumptions">Assumptions</a></li>
</ul>
</li>
<li><a href="#unified-model">Unified Model</a><ul>
<li><a href="#introduction_2">Introduction</a><ul>
<li><a href="#my-vision">My Vision</a></li>
<li><a href="#some-inchoate-ideas">Some Inchoate Ideas</a></li>
</ul>
</li>
<li><a href="#relevance_1">Relevance</a></li>
<li><a href="#assumptions_1">Assumptions</a></li>
</ul>
</li>
<li><a href="#resilience">Resilience</a><ul>
<li><a href="#introduction_3">Introduction</a></li>
<li><a href="#relevance_2">Relevance</a></li>
<li><a href="#assumptions_2">Assumptions</a></li>
</ul>
</li>
<li><a href="#fundamentality">Fundamentality</a><ul>
<li><a href="#introduction_4">Introduction</a></li>
<li><a href="#relevance_3">Relevance</a></li>
<li><a href="#assumptions_3">Assumptions</a></li>
</ul>
</li>
</ul>
</div>
<h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<p>There are a number of questions that I think could greatly change or refine dominant <span class="caps">EA</span> opinions on prioritization depending on the answers we can find for them. I don’t have time to research all or even a significant fraction of them, so I’m hoping that by compiling them here, others will get interested in them too.</p>
<p>Some of these questions will seem like they intend to cast into doubt opinions of one group or another. That’s just because I think that’s the job of prioritization research. But there’s probably a rough positive correlation between how much I like to probe theories and how likely I think they are.</p>
<p>If you’re interested in researching one of these questions, please leave a comment or ask me to edit the post so we don’t duplicate effort.</p>
<h2 id="value-of-prioritization-research"><a class="toclink" href="#value-of-prioritization-research">Value of Prioritization Research</a></h2>
<div class="admonition type">
<p class="admonition-title">Questions</p>
<ol>
<li>Which paths to impact are how plausible for prioritization research?</li>
<li>Are they sufficiently valuable on the margin to beat more direct interventions?</li>
</ol>
</div>
<h3 id="introduction_1"><a class="toclink" href="#introduction_1">Introduction</a></h3>
<p>I can think of four paths to impact already, but I don’t strong intuitions for how to weigh them nor do I think the list is complete.</p>
<ol>
<li>Research may show that we’ve been sufficiently wrong that continuing to pursue our current strategy would incur a significant opportunity cost.</li>
<li>Research may uncover new, highly cost-effective interventions.</li>
<li>Research may produce a more nuanced picture of the cost-effectiveness landscape – a combination of paths 1 and 2.</li>
<li>Research may cement our current strategy with enough evidence to secure the support of more uncertainty-averse contributors.</li>
</ol>
<p>Path 4 has some appeal to me, though I’m divided on whether that’s rational.<sup id="fnref:uncertainty-aversity"><a class="footnote-ref" href="#fn:uncertainty-aversity">1</a></sup> Path 3, however, is the one I find most interesting, and perhaps it’s also one that’ll require some more explanation, because it touches on the question how I would like to see the knowledge structured that we’ll hopefully generate.</p>
<p>I’ll discuss this question further in the next section.</p>
<p><a href="http://effective-altruism.com/ea/1ri/when_causes_multiply/">Denise Melchin’s</a> and <a href="http://measuringshadowsblog.blogspot.com/2015/08/multiplicative-factors-in-games-and.html">Sam Bankman-Fried’s</a> observations with regard to multiplicative effects between interventions are likely relevant here (see also Carl Shulman’s comment on that article), because the most relevant spaces – such as prioritization itself, <span class="caps">AI</span> safety, welfare biology, and community building – are all still very small and may have many low-hanging fruit left to pluck. Small size and high output elasticity of labor are indicators that we need to be careful not to allocate more resources to the area than it takes to resolve a bottleneck because otherwise it would only exacerbate a bottleneck elsewhere.</p>
<p>Based on the low number of people working on prioritization at the moment – probably lower than <span class="caps">AI</span> safety and community building, but tell me if that’s wrong – it’s possible that prioritization is a greater bottleneck than the work of the areas it leverages.<sup id="fnref:multiplication"><a class="footnote-ref" href="#fn:multiplication">5</a></sup></p>
<h3 id="relevance"><a class="toclink" href="#relevance">Relevance</a></h3>
<p>If we’re considering investing time and money into prioritization research, it would be useful to have some evidence that it’s plausibly the best use of our time. If not, it would still be useful to find out how we can 80/20 it to reap any low-hanging fruit.</p>
<h3 id="assumptions"><a class="toclink" href="#assumptions">Assumptions</a></h3>
<ol>
<li>Ones moral system is aggregative and consequentialist.</li>
<li>It is common enough that the research has value for enough people.</li>
<li>Infinite ethics is somehow solved.</li>
</ol>
<h2 id="unified-model"><a class="toclink" href="#unified-model">Unified Model</a></h2>
<div class="admonition type">
<p class="admonition-title">Questions</p>
<ol>
<li>What structure lends itself to parallel research on many questions while minimizing coordination overhead and maintaining a focus on only the most important issues?</li>
<li>How does this structure integrate all that information into a ranking?</li>
<li>To this end, what metrics are feasible to determine for a wide range of interventions?</li>
<li>What are the best ways to trade off value of information and option value against terminal goal realization, such as minimizing suffering?</li>
<li>What interventions lend themselves to several metrics so that more evidence on their cost-effectiveness can provide information on the priorities of several others at once?</li>
</ol>
</div>
<h3 id="introduction_2"><a class="toclink" href="#introduction_2">Introduction</a></h3>
<p>The typical <span class="caps">EA</span> approach, perhaps pioneered by GiveWell, has served its purpose: Identify the most pressing problems, find the best interventions that address these problems, and then support the organizations that implement these interventions best. But it separates only interventions to investigate further and interventions to dismiss. It is not a clean way of prioritizing among the first group.</p>
<p>What can be more or less cost-effective are these interventions, not the problems. Interventions, however, don’t usually solve only one problem, especially not if everyone gets to decide how they want to categorize undesirable things into problems.</p>
<h4 id="my-vision"><a class="toclink" href="#my-vision">My Vision</a></h4>
<ol>
<li>One problem is the vagueness of what constitutes one problem: whether it could just as well be subdivided (academia being inefficient vs. psychology being inefficient and gender studies being inefficient) or whether it’s instrumental/proximate (meat-eating vs. speciesism). In the first case, academia being inefficient doesn’t cause the subfields to be inefficient but speciesism is one cause of meat-eating (assuming that theory holds).</li>
<li>The second problem is whether an intervention has one main effect and other effects (flow-through effects) are minor or whether the configuration is somehow different. </li>
<li>Finally, we don’t have a common metric that would always allow us to compare the magnitude of the effects an intervention has on a problem.<sup id="fnref:common-metric"><a class="footnote-ref" href="#fn:common-metric">3</a></sup></li>
</ol>
<p>So taken together, we’d like to have a set of problems, an intervention, and a common metric that allows us to compare the effect of the intervention on the problems. Then we could determine the effect distribution of the intervention over the problems and say whether it’s sort of log-normal (very few problems get a lot of effect, others can be ignored), more heavy-tailed (very few problems get a lot of effect but many others get sizable effects too, so they can’t be ignored), uniform (every problem gets addressed about equally much), or some other variation.<sup id="fnref:inus-conditions"><a class="footnote-ref" href="#fn:inus-conditions">4</a></sup></p>
<p>These are all significant challenges. To make progress on them, we’ll need much more research on how to translate various effects into the metrics we care about (or some useful proxy measures) and then aggregate them into a partial ordering. This task can be distributed over many researchers and will require them to answer highly specific questions – much more specific than, say, whether technical or policy <span class="caps">AI</span> safety research is more important.</p>
<p>Meanwhile, other researchers can investigate what interventions show which of the investigated effects and can suggest more effects or different taxonomies of effects for further investigation.</p>
<p>Only then can we derive insights from the complete model. (But as an <span class="caps">MVP</span>, we can just make up taxonomies and performance metrics, aggregate them into a model and share them. That would lend much more structure to prioritization research already. I’ve not seen even that much so far.)</p>
<p>With that basic structure in place, we would incur little coordination overhead and could retain a clear focus on what is most important while we start making finer distinctions between interventions, work out the important differences, and study them separately.</p>
<p>In the end, I hope we’ll no longer be studying broad questions such as whether it’s better to show down negative developments or to improve them, but will invest significant resources into very detailed questions while being confident enough that the value of information will be worth it.</p>
<p>Poor Economics has demonstrated how a space can move from broad questions such as “Does aid help?” to detailed questions such as what cost-effectiveness a certain approach to mass deworming of primary school children in a certain region of Kenya displays towards improvements along some highly comparable metrics. I want to see a similar movement in the prioritization space at large, because at the moment I feel lie we’re still asking a number of questions that may be too broad and vague to answer.</p>
<h4 id="some-inchoate-ideas"><a class="toclink" href="#some-inchoate-ideas">Some Inchoate Ideas</a></h4>
<p>In order to trade off interventions that are highly unlikely to succeed but abysmal if they fail against interventions that change the status quo in incremental, unspectacular ways, or against interventions that have great upsides but destroy option value, or against interventions that are speculative but have great value of information, etc., we need common metrics to compare them.</p>
<p>That does not necessarily have to be one common metric for all of them. If one intervention lends itself to two types of cost-effectiveness metrics, it can serve as a bridge to compare two interventions, each of which only lends itself to one of them and not both the same one.</p>
<p>For example, in one case we may have a risk over time of a sudden explosion of suffering happening while in the other we have a continuous function of total expected suffering over time. Here we can probably use survival analysis to convert the first into another function of expected total suffering over time.</p>
<p>In other cases we may have only importance, tractability, and neglectedness scores; <a href="http://documents.worldbank.org/curated/en/270771529500170694/pdf/WPS8481.pdf">density and centrality</a>; or fundamentality to go on.</p>
<h3 id="relevance_1"><a class="toclink" href="#relevance_1">Relevance</a></h3>
<p>About as relevant as the whole of prioritization research. See above.</p>
<h3 id="assumptions_1"><a class="toclink" href="#assumptions_1">Assumptions</a></h3>
<ol>
<li>Enough people will care about the output of the system.</li>
<li>We’ll be able to disentangle questions sufficiently that people don’t have to work together extremely closely to solve several of them.</li>
</ol>
<h2 id="resilience"><a class="toclink" href="#resilience">Resilience</a></h2>
<div class="admonition type">
<p class="admonition-title">Questions</p>
<ol>
<li>Where do important arguments still rely on overly “absolute” models?<ol>
<li>For simplicity, models may ignore expected decay over time,</li>
<li>rely on “sequence thinking” only,<sup id="fnref:sequence-thinking"><a class="footnote-ref" href="#fn:sequence-thinking">2</a></sup> or</li>
<li>treat a multidimensional optimization problem as if it were one-dimensional.</li>
</ol>
</li>
<li>How long can “shmingletons” (explained below) ideally last – decades or millions of years?</li>
<li>If we use methods like stabilizing feedback and path dependence to create a highly resilient system, how long can we make it last?</li>
<li>How long will extinction last?</li>
<li>The answers to these may be functions with a few dependent variable – which ones should we use and how do we recognize them in real or hypothetical systems?</li>
</ol>
</div>
<h3 id="introduction_3"><a class="toclink" href="#introduction_3">Introduction</a></h3>
<p>This is where it gets a bit more specific. Models of the future that a lot of effective altruists base their decisions on are currently of mostly qualitative nature. In order to reason about systems qualitatively, effects need to point in the same direction or be sufficiently strong or likely that you don’t face non-negligible trade-offs. Where trade-offs are unavoidable, they force you to limit yourself to strategies that are robust enough to be beneficial either way or to coordinate well with many others to try out many strategies that each are highly unlikely to succeed.</p>
<p>A limitation to robust strategies can be costly in terms of impact and option value. Without the evidence that shows that, for example, malaria prevention with bednets is fairly cost-effective, we’d be limited to interventions that we might consider robust (depending on our moral goals), such as building up a better health system (horizontal rather than vertical interventions) or improving infrastructure, which may be radically less cost-effective. Similarly, trying many risky strategies is also inferior to trying only the best among them.</p>
<p>Therefore, more research that could narrow down the set of the most cost-effective interventions can be very valuable, and that will require quantifying the relevant factors in trade-offs more rigorously.</p>
<p>One example of a concept that is still heavily “rounded” is the <a href="https://nickbostrom.com/fut/singleton.html">singleton</a>:</p>
<blockquote>
<p>As I introduced the notion, the term [singleton] refers to a world order in which there is a single decision-making agency at the highest level. Among its powers would be (1) the ability to prevent any threats (internal or external) to its own existence and supremacy, and (2) the ability to exert effective control over major features of its domain (including taxation and territorial allocation).</p>
</blockquote>
<p>I’d be more interested in something like a “shmingleton,” which can “prevent <em>almost</em> any threats … to its own existence and supremacy.” Say, an entity with a vanishingly small put nonzero probability of collapse over time. It seems much more likely and natural to me that we’ll see shmingletons than actual singletons, and given the vastness of the future, even this small probability can make a large difference.</p>
<p>Here we can turn to systems theory and complexity theory to help us model the resilience of systems and get a feeling for the yearly probabilities of collapse of systems, because we can think of a shmingleton as a highly resilient system.</p>
<h3 id="relevance_2"><a class="toclink" href="#relevance_2">Relevance</a></h3>
<p>If we’re at a key point in history at which we can uniquely influence the values that will shape the billions of years of the future, and we lose that ability once they’re set in stone, everything we do only matters insofar as it influences what the values are that will get set in stone.</p>
<p>But if that stone is more like sandstone, the argument becomes less diamondclad. Particularly low tractability, some sort of cluster-thinking penalty, low personal fit, etc. suddenly become much more important considerations.</p>
<h3 id="assumptions_2"><a class="toclink" href="#assumptions_2">Assumptions</a></h3>
<p>I can’t think of any nontrivial assumptions on which the relevance of these questions depends.</p>
<h2 id="fundamentality"><a class="toclink" href="#fundamentality">Fundamentality</a></h2>
<div class="admonition type">
<p class="admonition-title">Questions</p>
<ol>
<li>How can we recognize bottlenecks in the dependency tree that leads to impact such as suffering minimization?</li>
<li>What sorts of dependencies are we dealing with? Do we need a minimum amount of something? Are there several fundamental qualities that are necessary but interchangeable? Are there others of which we need a minimum amount in sum?</li>
<li>Can we use something like Page Rank to determine the relative importance of dependencies to trade off against tractability and neglectedness?</li>
</ol>
</div>
<h3 id="introduction_4"><a class="toclink" href="#introduction_4">Introduction</a></h3>
<p>How do we compare <a href="https://claviger.net/the-bulk-of-the-impact-iceberg.html">very fundamental and thus hopefully very robust interventions</a> (like the research I’m trying to encourage here) to “last brick” types of interventions like bednet distributions?</p>
<p>In the linked article I argue that fundamentality is a big asset thanks to the tremendous leverage it has over interventions it enables or unearths. But even fundamental research builds upon other research as well as communication, money transfer, safety, trust, etc. (Let’s call these things “stocks.”)</p>
<p>Maybe the dependencies between these stocks are of various different shapes: Maybe we need a minimum level of trust, an equilibrium between research that enhances capabilities of and control over something, both of research and marketing (though the considerations above concerning multiplication of impact), or various other types of interdependencies.</p>
<p>Given the dependencies between stocks and the shapes of those dependencies, can we derive the relative importance of stocks? And observing them in our environment, can we derive the marginal effect of increasing them? And from that the marginal cost-effectiveness of increasing them?</p>
<p>The page rank algorithm solves this problem for one type of (literal) link. That may be a start, and further research into this class of algorithms may reveal more appropriate algorithms or ways to adjust the algorithm.</p>
<h3 id="relevance_3"><a class="toclink" href="#relevance_3">Relevance</a></h3>
<p>On the for-profit market, businesses that are too meta are too expensive to run compared to the margins they can charge so that they will fail. Conversely, if businesses are too vertically integrated, effects such as economies of scale will lead to meta-businesses popping up that the businesses can outsource some of their work to.</p>
<p>But outside costly experiments, we don’t currently have any principled way of predicting such dynamics in altruistic contexts.</p>
<h3 id="assumptions_3"><a class="toclink" href="#assumptions_3">Assumptions</a></h3>
<p>When considering this problem, I typically think about it as a graph of</p>
<ol>
<li>discrete, unique nodes,</li>
<li>of which there is a finite number, and</li>
<li>whose interconnections are sufficiently few or simple (e.g., noncircular).</li>
</ol>
<p>I imagine that this problem will need to be heavily simplified with a lot of assumptions about what constitutes a node in the graph in order to merge complicated cases and ignore hopefully irrelevant cases before it’ll become computationally tractable.</p>
<p>When qualitative arguments yield unintuitive results but otherwise appear sound, they pattern-match (for me) the types of arguments a better informed or more intelligent person can produce for almost any position to convince a less informed or intelligent person – Scott Alexander talks about <a href="https://slatestarcodex.com/2019/06/03/repost-epistemic-learned-helplessness/"><em>epistemic learned helplessness</em></a>.<sup id="fnref:scott-alexander-disclaimer"><a class="footnote-ref" href="#fn:scott-alexander-disclaimer">6</a></sup></p>
<p>When an argument still feels that way to people with a 99th percentile <span class="caps">IQ</span> after years of reading about the relevant field, it seems plausible to me that the bottleneck is that the argument needs to be improved. But no one would produce motivated arguments intentionally (I suppose), so turning an argument into a quantitative model can help us discover gaps in our own reasoning and expose assumptions more clearly.</p>
<p>There does not need to be a symmetry in the other direction, i.e., interventions need not have a plurality of effects, but they might unless we define the effects sufficiently inclusively. In health, where we have a taxonomy of diseases, interventions are classified into vertical and horizontal, and the horizontal interventions are the ones with particularly many effects.</p>
<div class="footnote">
<hr>
<ol>
<li id="fn:uncertainty-aversity">
<p>The question whether a better informed, more intelligent community would be convinced by an argument can be answered by becoming this better informed, more intelligent community or though a <a href="https://intelligence.org/files/VingeanReflection.pdf">Vingean reflection</a>–like process – “a theory of formal reasoning that allows an agent to reason with high reliability about similar agents, including agents with considerably more computational resources, without simulating those agents” (Daniel Dewey). <a class="footnote-backref" href="#fnref:uncertainty-aversity" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
<li id="fn:sequence-thinking">
<p>Most of the population seems to overfit their experience to the point that they care surprisingly little about rare events or uncommon considerations. The EA community seems to do that to a much lesser degree, but that does not clarify whether we’ve overcome the bias and are now well calibrated in this regard or whether we just underfit as often as the rest of the population overfits. It may seem that the smaller divergence from the norm is more likely than the larger divergence, but <a href="http://slatestarcodex.com/2014/12/17/the-toxoplasma-of-rage/">strong divergences are important for informative in-group signaling</a>: “A moral action that can be taken just as well by an outgroup member as an ingroup member is crappy signaling and crappy identity politics.” <a class="footnote-backref" href="#fnref:sequence-thinking" title="Jump back to footnote 2 in the text">↩</a></p>
</li>
<li id="fn:common-metric">
<p>For me, the expected difference in suffering over time between futures with and without an intervention has proved widely useful, as a metric and a proxy for what I care about. (<a href="https://claviger.net/robustness-through-suffering-focus.html">I also care about other things.</a>) But I can imagine cases where I’d have trouble applying it, e.g., interventions aimed at improving institutional decision-making. <a class="footnote-backref" href="#fnref:common-metric" title="Jump back to footnote 3 in the text">↩</a></p>
</li>
<li id="fn:inus-conditions">
<p>Effects usually have a plurality of causes – <a href="http://science.jrank.org/pages/8545/Causality-Inus-Conditions.html">John Mackie’s inus conditions</a> are a simplification since the necessary but insufficient conditions can usually be refined further into insufficient conditions that are dependent on not only the presence but a particular degree of another insufficient condition and the absence of a third and so forth before they become sufficient. <a class="footnote-backref" href="#fnref:inus-conditions" title="Jump back to footnote 4 in the text">↩</a></p>
</li>
<li id="fn:multiplication">
<p>The particular application here is that (1) a group of researchers does the prioritization research and (2) a subset of people in AI safety, welfare biology, and other spaces will listen to prioritization research if it finds that their work needs to focus on a particular subproblem to be among the top most effective things for them to do. The effect of the first group on the second can be modeled as multiplication of the impact of the second. Because this is true across the board for all possible areas of altruistic work but is not true of every individual working on the problem (for reasons of personal fit and preferences) and because individuals can switch between spaces, I think it is cleaner to think of it not as prioritization research influencing a number of other spaces but as influencing a number of other EAs. So, barring a great drop in output elasticity in prioritization research, we may need about as many people in prioritization research as we have sufficiently flexible people in other spaces all combined before prioritization research reaches the limits to its growth according to this model. <a class="footnote-backref" href="#fnref:multiplication" title="Jump back to footnote 5 in the text">↩</a></p>
</li>
<li id="fn:scott-alexander-disclaimer">
<p>Disclaimer: I don’t generally endorse the works of the author. Alexander originated a wealth of helpful ideas, so that I can’t help but cite him lest it seem that I plagiarize them. Unfortunately, (1) the community around his blog contains some insalubrious factions, and (2) until roughly 2016, he himself still published articles that presented issues in a skewed fashion reminiscent of the very dynamics he warns of in <a href="https://slatestarcodex.com/2014/12/17/the-toxoplasma-of-rage/">Toxoplasma of Rage</a>. I’m adding these disclaimers to avoid the impression that I accept such intellectual wantonness or that it is accepted in my circles. I don’t know whether he still endorses his old approaches. <a class="footnote-backref" href="#fnref:scott-alexander-disclaimer" title="Jump back to footnote 6 in the text">↩</a></p>
</li>
</ol>
</div>Cooperative Moral Goals2018-09-06T18:30:00+00:002019-05-26T12:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2018-09-06:/robustness-through-suffering-focus.html<p>In this article I summarize my current thinking on how I want to make my actions robustly positive on a normative level by choosing the moral goals to focus on according to cooperativeness heuristics on five levels.</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#moral-uncertainty">Moral Uncertainty</a></li>
<li><a href="#moral-cooperation">Moral Cooperation</a></li>
<li><a href="#compromise-goals">Compromise Goals</a></li>
</ul>
</div>
<h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<p>We’re quick to admit that we haven’t left our tribal stone-age brains behind yet, so that we need to continuously be wary of the biases that come with having a mind that “runs on broken hardware.” But in some domains, it’s not easily settled who is biased and who is applying useful simplifying heuristics.</p>
<h2 id="moral-uncertainty"><a class="toclink" href="#moral-uncertainty">Moral Uncertainty</a></h2>
<p>Effective altruism comprises a number of camps. For the sake of this example I’ll simplify the situation down to two clusters and call them the Visionaries and the Stoics.</p>
<p>The Visionaries are people who yearn for space colonization, abhor existential risks, value making happy beings as much as they value making beings happy, and generally behave sort of like classic (hedonistic) utilitarians. They are usually not very vocal about suffering.</p>
<p>Who is vocal about suffering, however, are the Stoics. They are wary of space colonization, abhor suffering risks, value making beings happy more than making happy beings, and generally behave sort of like preference utilitarians (with an antifrustrationist bent). They are usually not very vocal about maximizing our use of the cosmic commons.</p>
<p>My hypothesis is that this division is driven by our tribal minds and discomfort with uncertainty.</p>
<p>The Visionaries formed or consolidated because of a pre-existing and influential group of people working on existential risks, primarily from <span class="caps">AI</span>. They had built expertise in the area and a community around their work. Eventually they recognized some problems with their work – e.g., that they increase the <a href="https://arbital.com/p/hyperexistential_separation/">probability of astronomical suffering</a> as a side effect of their increasing the probability of astronomical bliss or eudaimonia, or that it’s a common intuition among thoughtful people today to have no or few offspring unless a very high bar for their expected life satisfaction is met.<sup id="fnref:pro-choice"><a class="footnote-ref" href="#fn:pro-choice">2</a></sup></p>
<p>They were also averse to great uncertainty, in particular, they preferred to be able to believe that they were likely making things better rather than worse. But being the rationalists they are, they knew their peers and their own minds wouldn’t let them get away with believing provably false things about reality. The only recourse was to change the element of their decisions that was arguably more antirealist than their epistemology, namely their moral goals. So they gradually self-modified to the de-facto rather hedonistic utilitarian position that allowed them to care to an overriding extend about such things as <a href="https://nickbostrom.com/astronomical/waste.html">astronomical waste</a>.</p>
<p>Likewise, the Stoics formed amidst an existing animal advocacy community. They were heavily invested in reducing the suffering of farmed animals by preventing them from coming into existence in the first place. Eventually they recognized some problems with their work – e.g., that the logic of the larder may go through for most cows farmed for meat, or that they may deprive some sufficiently insensitive individuals of their net positive lives as a side effect of preventing other net negative lives, with all of it hinging on thresholds with huge uncertainty, such as which life is worth living.</p>
<p>They were also averse to great uncertainty, in particular, they preferred to be able to believe that they were likely making things better rather than worse. But being the rationalists they are, they knew their peers and their own minds wouldn’t let them get away with believing provably false things about reality. The only recourse was to change the element of their decisions that was arguably more antirealist than their epistemology, namely their moral goals. So they gradually self-modified to the de-facto rather antifrustrationist preference utilitarian position (or negative, negative-leaning, prioritarian, etc. utilitarian position) that allowed them to care to an overriding extend about suffering.</p>
<p>Surely, this is not true of everyone in those groups. Degrees of pain sensitivity, a sense of security from being loved by one’s parents in early childhood, one’s density of conscious experiences, agreeableness, and various other influences might be hypothesized to contribute to the effect.</p>
<p>Such hypotheses give either set of moral preferences a sense of arbitrariness that undermines my motivation for pursuing it, which, ipso facto, undermines the intensity of the preferences.</p>
<p>This can easily be misunderstood in various ways, so let me be clear that I distinguish at least four types of distinctions here: (1) happiness/suffering, (2) the moral evaluation according to which happiness/suffering is good/bad, (3) the moral preference for maximal/minimal aggregate suffering, and (4) one’s focus on actions that maximize happiness or minimize suffering.</p>
<p>Questioning the relative importance of happiness or suffering doesn’t impinge on what happiness or suffering feel like and needn’t (but might) impinge on 2 or 3. But doing so <em>will</em> change the relative intensity of the preferences and thus what actions end up most choiceworthy.</p>
<h2 id="moral-cooperation"><a class="toclink" href="#moral-cooperation">Moral Cooperation</a></h2>
<p>Maximizing and minimizing moral preferences benefit from the collaboration of many actors on satisfying them. (Unless they minimize something rare and unguessably obscure or one is omnipotent.) The more outré one’s moral preferences, the fewer collaborators one is likely to find. So there is a trade-off where, even with perfectly crisp moral preferences, one is better off compromising on them to some degree to satisfy them to a greater degree.</p>
<p>My conclusion from this consideration has been to maximize cooperation in five ways and maybe even to interpret moral cooperation as a terminal moral goal of mine.</p>
<ol>
<li>
<p>I empathize with a mess of moral preferences whose boundaries and levels of abstraction are fuzzy. So even internally, I need to compromise to maximize my motivation for my actions.</p>
<ol>
<li>
<p>The idea of a <a href="http://www.overcomingbias.com/2009/01/moral-uncertainty-towards-a-solution.html">moral parliament</a> helps me to achieve this. I imagine that all the different moral theories I empathize with are factions in a moral parliament. Then I intuit some rough fractions of the parliament and assign the factions to the fractions. So we have a smaller faction of something deontological looking, a larger faction of something preference utilitarian, etc.<sup id="fnref:levels"><a class="footnote-ref" href="#fn:levels">5</a></sup></p>
<p>Theoretically, you could now just crunch the numbers for every decision, but unfortunately my model is not that precise, not in a simple sense and also not in the sense that I’m unsure whether my deontological looking faction is a smaller actually, say, Kantian one, or whether it’s a much larger two-level utilitarian one that includes the preference utilitarian one among others. I imagine that many people will face such issues.</p>
<p>But the model is still helpful in that a lot of potential actions that I can take elicit near unanimous votes (with abstention). A vote on whether I want to start space colonization and fill the Hubble volume with happy people (or better yet, very simple, very happy, simulated beings) elicits some agreement, some shrugs, but also many loud voices urging me to consider the individuals that are suffering, that will be suffering, or that may be suffering if something goes wrong. It’s not met with unanimous approval. Neither would be a goal such as blowing up the planet.</p>
</li>
</ol>
</li>
<li>
<p>My moral preferences have changed over time through reflection and changing “tribes.”<sup id="fnref:causation-questionable"><a class="footnote-ref" href="#fn:causation-questionable">3</a></sup> This may continue to happen, so I want to already cooperate with my future versions.</p>
</li>
<li>
<p>Other agents alive today are potential collaborators or saboteurs depending on our game theoretic behaviors; the degree to which our preferences complement, augment, or clash with each other; our respective power; and how well we communicate.</p>
<ol>
<li>
<p>If we are going for a more or less power-weighed compromise anyway,<sup id="fnref:power-weighed-compromise"><a class="footnote-ref" href="#fn:power-weighed-compromise">1</a></sup> one may wonder why we shouldn’t let a negotiation decide and be maximally partisan (more so than we would normally be to make token concessions to the other side in return for concessions from them who we need to assume exaggerate their stakes too).</p>
<p>But I think one should avoid escalation when possible as it would only make the process more costly in terms of transaction costs, time, and risk of further escalation. This strikes me as cooperation in a Prisoner’s dilemma–like situation, which is probably a good idea so long as the other side also continues to cooperate. I see no reason to doubt that today.</p>
</li>
</ol>
</li>
<li>
<p>Agents in the future can’t causally affect my actions today, but they can acausally reward or punish me by satisfying or frustrating my other-directed preferences. This is distinct from the fifth case in that I can causally affect them. But can’t do so with certainty – and probably only to a small degree in expectation – so that I’m well-advised to take the preferences into account that they are likely to have as opposed to those that I would like them to have.</p>
</li>
<li>
<p>Fully acausal cooperation.</p>
<ol>
<li>
<p>Conversely, it may be advisable to cooperate with agents in the past as it gives me evidence that agents in the future that are sufficiently correlated with me will cooperate with my moral preferences. This indicates an increased importance of maintaining traditions, at least when our ancestors would not themselves have abandoned them if they had had better information of the sort that we may have today.</p>
</li>
<li>
<p>What may also be called for is cooperating with beings who are not agenty in the sense that they could/would support or thwart our efforts in that it gives us evidence that beings so much more powerful than us that we couldn’t support or thwart their efforts would cooperate with us.</p>
</li>
<li>
<p>Agents throughout the universe or multiverses may also be more likely to cooperate with us in worlds where we cooperate with them. This is where the concept of <a href="http://effective-altruism.com/ea/1gf/multiversewide_cooperation_in_a_nutshell/">superrationality</a> comes in.</p>
</li>
</ol>
</li>
</ol>
<h2 id="compromise-goals"><a class="toclink" href="#compromise-goals">Compromise Goals</a></h2>
<p>The resulting goals are hard to infer precisely, but we can make educated guesses about them. Many gray areas remain for reasons of unknown unknowns, because the compromise goals are sometimes in conflict, or because they are sometimes silent.</p>
<ol>
<li>
<p>Chesterton’s fence becomes more important because not only may old norms still be relevant in ways we haven’t understood, but maintainting them out of respect for our ancestors who cared about them also gives us evidence that our descendants may continue to maintain what we care about. Some caveats:</p>
<ol>
<li>
<p>We wouldn’t want our descendants with better information to maintain norms that we would rescind if we had such better information. This is a difficult call to make.</p>
</li>
<li>
<p>This is in tension with moral progress, if that’s a meaningful concept, so a world with a short or sparsely populated future may call for a more reactionary morality than a world with a long, populous future.</p>
</li>
</ol>
</li>
<li>
<p>For a wide range of other-directed goals it is instrumentally necessary to exist in order to reach them. Also a nontrivial number of agents with such goals may exist or continue to exist in expectation. This suggests that opposing existential risks is fairly convergent because it helps these agents to exist. Some caveats:</p>
<ol>
<li>
<p>I know enough people who would prefer not to have been born (but shy away from suicide because of its low success rate and associated risks). Existential risk reduction may defect against them in the same way that a hypothetical ban on suffering would defect against those that enjoy suffering. My main objection to this line of thought is that agents who are more agenty about other-directed goals will be more likely to welcome their own existence, because for the converse to be the case, their personal suffering would have to outweigh the opportunity cost in moral preference satisfaction that they would pay if they could choose not to have existed. In a purely power-weighed compromise, this would probably settle the issue, because these more agenty agents will lend the greater support (or pose the greater risks) to our goals. But I prefer to live in a world where I’ll not be heavily defected against for having minority preferences (that don’t themselves constitute a defection), and being considerate toward minority preferences myself gives me evidence that others may be considerate toward my minority preferences too. Easier access to safe, institutionalized suicide may be a win-win.</p>
</li>
<li>
<p>It is highly unclear whether continued existence increases or decreases aggregate suffering, depending on what shape it takes, what sorts of minds can suffer, whether we are in a simulation and what the simulator’s intend is, etc. These will need to be weighed against the first compromise goal for each individual scenario whose probability we want to increase or decrease.</p>
</li>
<li>
<p>There may be powerful agents in the future whose other-directed goals may be better served through non-existence.</p>
</li>
</ol>
</li>
<li>
<p>The most widely shared moral foundations (according to the eponymous theory) are <em>care</em> and <em>justice</em>, where <em>care</em> can be understood as an antifrustrationist aspect of ones moral preferences. Especially among the agents (other than my own person moments) that I’m most likely to want to cooperate with – utilitarians of various stripes – the <em>care</em> foundation is strong. This leads me to believe that there are very few agents who don’t care about suffering reduction to some extend and even fewer who would oppose it.<sup id="fnref2:pro-choice"><a class="footnote-ref" href="#fn:pro-choice">2</a></sup> Opposing suffering seems to me like most robust compromise goal at the moment. One caveat:</p>
<ol>
<li>In some cases, this is in tension with the second likely compromise goal, as mention in its second caveat. I think that absent ways of arbitrate such trade-offs, we should steer clear of actions that are partisan in one direction or the other.</li>
</ol>
</li>
</ol>
<p>The result could be to focus on actions that maximize the sum of these goals as well as one can guess it, but because there are many more constraints such as crowdedness, personal aptitudes, etc., for me it rather takes the shape of focusing on actions that maximize one goal without hindering the others.</p>
<div class="footnote">
<hr>
<ol>
<li id="fn:power-weighed-compromise">
<p>Power-weighed compromise is an attractor state because powerful agents could and would want to use their power to change any other compromise to a power-weighed compromise. This would create addional conflict and then return the system to the power-weighed attractor state anyway. But I think we can aim for a compromise that has stronger minority protection properties than a pure power-weighed compromise so long as these concessions cost the more powerful agents less than the conflict would cost them. The marginal value of such concessions is also probably greatly higher for the minority agents than for the powerful agents. <a class="footnote-backref" href="#fnref:power-weighed-compromise" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
<li id="fn:pro-choice">
<p>Note that I think that this also extends to some degree into the direction of population ethics. For example, most people in my circles endorse a “Pro Choice” stance even in the case where (1) in world A, the child is not born and, in world B, has a net positive life, (2) the decrease in well-being of the parents in world B compared to A is offset by the additional well-being of the child in world B, and (3) the resources freed by not having to raise the child in world A are not invested into an at least equal gain of happiness for an existing or marginal person. These assumptions are perfectly plausible for non-EA parents and run counter to the implications of classical utilitarianism (i.e. total utilitarianism). The “Pro Life” position is very similar so long as it allows abstinence. You have to turn to fringy movements like Quiverfull to find a group that decides in accordance with classical utilitarian population ethics. <a class="footnote-backref" href="#fnref:pro-choice" title="Jump back to footnote 2 in the text">↩</a><a class="footnote-backref" href="#fnref2:pro-choice" title="Jump back to footnote 2 in the text">↩</a></p>
</li>
<li id="fn:causation-questionable">
<p>Interestingly, these changes happened at least in some cases before I realized that my environment largely shared them. It would be flattering to think that I arrived at them independently because they are true. It would also be convenient because they are arguably more likely to be true if more people arrive at them independently. But my use of “flattering” and “convenient” may have betrayed that I’m skeptical. The geographic clustering rather suggests that there is maybe something to the broader cultures of the regions that encourages one or the other – perhaps <a href="https://meteuphoric.com/2018/07/01/are-ethical-asymmetries-from-property-rights/">differences in thinking about property</a>, somehow connected to social security paid from taxes? But also social influences within my small peer groups may have subtle effects (as described in the introductory paragraphs). No one went out to convince me or even tell me about suffering-focused ethics. It didn’t even have a name. But my efforts to make sense of the actions and priorities of my peers crystalized into models of morality and the world that, ipso facto, led me to prioritize the same actions. Now these models feel fundamentally, axiomatically, and intuitively correct to me just like one’s own dialect feels intuitively correct to oneself. But we know that it’s just as arbitrary as anyone else’s. (Not fully arbitrary, since it needs to serve as communication tool and error-correcting code. Just roughly similarly arbitrary.) <a class="footnote-backref" href="#fnref:causation-questionable" title="Jump back to footnote 3 in the text">↩</a></p>
</li>
<li id="fn:moral-realism">
<p>Back when I hadn’t thought about metaethics and assumed that moral realism must be true, I predictably dedicated a lot of time to the search for the true morality and to the search for some form of test for the truth value of moral theories. That should probably be the top priority for moral realists. <a class="footnote-backref" href="#fnref:moral-realism" title="Jump back to footnote 4 in the text">↩</a></p>
</li>
<li id="fn:levels">
<p>There are two things that I’ve decided not to merge down to this level, namely my decision procedures (I’m a big fan of the <a href="http://effective-altruism.com/ea/140/integrity_for_consequentialists/">integrity</a> one) and heuristics that follow from considerations of cooperation (more on that in the second section). <a class="footnote-backref" href="#fnref:levels" title="Jump back to footnote 5 in the text">↩</a></p>
</li>
</ol>
</div>Prepared Opportunism2018-09-02T17:00:00+00:002018-09-02T17:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2018-09-02:/prepared-opportunism.html<p>Prepared opportunism may be a neglected strategy in effective altruism.</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#process">Process</a><ul>
<li><a href="#intelligence">Intelligence</a></li>
<li><a href="#evaluation">Evaluation</a></li>
<li><a href="#exploitation">Exploitation</a></li>
</ul>
</li>
<li><a href="#examples">Examples</a><ul>
<li><a href="#reg-fund-management">REG Fund Management</a></li>
<li><a href="#local-groups-in-ceas-community-building-funnel">Local Groups in CEA’s Community Building Funnel</a></li>
</ul>
</li>
</ul>
</div>
<h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<p>Usually, people get thrown into something – say, a cousin getting cancer – and so they start dedicating resources to some almost random cause. But an attempt to distance oneself from that type of behavior – an attempt to do things more intentionally and strategically – can backfire.</p>
<p>I don’t have insight into many <span class="caps">EA</span> organizations, but among the things I can see there is little seizing of unplanned opportunities going on. Actions are usually intentionally selected, planned, and started by the organization at the time the organization deems optimal. That is better than seizing any random opportunity and probably better than just waiting for good opportunities (because they are probably too few), but a mixed strategy may be even better.</p>
<p>The way I have implemented planned opportunism in the past (in a news context), it was a three-step process.</p>
<h2 id="process"><a class="toclink" href="#process">Process</a></h2>
<h3 id="intelligence"><a class="toclink" href="#intelligence">Intelligence</a></h3>
<p>The first step is to brainstorm what classes of opportunities may be interesting. Where I worked (or volunteered) at the time, these were trademark filings, leaked photos from <span class="caps">B2B</span> fairs, <span class="caps">SEC</span> filings, investor reports, presentations, and conference calls of certain public companies, <span class="caps">TV</span> schedules (as long in advance as possible), unused assets and names of assets in browser games, and sometimes revealing comments from insiders on social media (be it just the information that they’re on vacation at a certain time, since their jobs only allowed that in certain circumstances).</p>
<p>A lot of these types and sources of information were highly nonobvious to us, and it took us one or two years to collect these ideas and to automate the retrieval and alerting without too many false positives. Often Google Alerts was not enough (and it didn’t work very well at the time), so I developed some custom software to gather and classify the data.</p>
<p>Depending on what an organization does, the relevant information will be very different, and it will probably take long to come up with the right ideas and gather the right sources.</p>
<h3 id="evaluation"><a class="toclink" href="#evaluation">Evaluation</a></h3>
<p>In my experience in the news context, seizing these opportunities was almost zero cost, and we had too little to report on anyway, so that we didn’t apply a strong filter. But that’s probably unusual. What I’ve heard increasingly often from <span class="caps">EA</span> organizations is that they decline more and more requests for interviews and such because press coverage has been of highly variable value and disvalue. So I think that in general, evaluating whether an opportunity is good enough is probably just as hard as noticing opportunities, and the “Hell, yeah!” heuristic may be a good guide to which ones to accept: decline unless your evaluation is nothing short of “Hell, yeah!”</p>
<p>I read an anecdote somewhere (maybe in <em>The Intelligent Investor</em> or a similar book) where someone was given a bunch of money to invest into some type of business. That was way back when you had to travel around to invest into businesses, so this investor/advisor did that. Eventually he returned and gave the money back because, in his opinion, none of the opportunities had been good enough.</p>
<p>This seemed like an impressive feat to me because this investor not only compared the different current opportunities to each other but also to a wider reference class of past and potential opportunities. That’s what makes this step hard.</p>
<h3 id="exploitation"><a class="toclink" href="#exploitation">Exploitation</a></h3>
<p>This is where the preparedness comes in. When you’re waiting for an unknown opportunity of a certain reference class of opportunities to happen, you want to prepare for it as well as possible. In the news context, I prepared article templates where I’d only have to fill in some titles and dates, and then had an article ready to publish within a minute. (If you have a tool that wakes you up in the middle of the night when an opportunity happens, you’ll be glad not to have to do more thinking than that.) These templates often contained some bit of historical or otherwise topical trivia that few people knew and that was likely to fit with whatever I’d fill in later. And if not, I could still delete it.</p>
<p>Similarly, an activist could prepare by writing op-eds on various references classes of big events that tend to occur around once a year or so and lend themselves to their activist work, and then have them ready to go out (and in mostly copyedited shape too) within an hour. Address sufficiently many of these reference classes, and you can send out an essay every month.</p>
<p>Or one could have a list of contact data of certain politicians ready, and when a discussion on risks from <span class="caps">AI</span> starts in some government body, immediately (say, within the day) send all relevant politician, or all that may be receptive to it, a prepared letter offering advise on the topic.</p>
<p>Or, more speculatively, use databases of mergers and acquisitions of <span class="caps">VC</span>-backed companies to quickly offer donation advice to the founders who may make their exit at the same time.</p>
<h2 id="examples"><a class="toclink" href="#examples">Examples</a></h2>
<h3 id="reg-fund-management"><a class="toclink" href="#reg-fund-management"><span class="caps">REG</span> Fund Management</a></h3>
<ol>
<li>
<p>The <span class="caps">EA</span> Foundation wants to run <span class="caps">EA</span> workshops tailored to wealthy people who may become potential clients for a fund akin to the <a href="https://reg-charity.org/fund/"><span class="caps">REG</span> Fund</a>. (Not a perfect match of the Intelligence stage because it’s a very active process, but there’s no reason to value purity when applying such a strategy since the stages are also just something I made up.)</p>
</li>
<li>
<p>If a funder is ready to invest, they will need to decide whether the person is sufficiently cause-neutral that the cooperation will be worthwhile.</p>
</li>
<li>
<p>If that all works out, they’ll be able to seize that opportunity right away because thanks to experience in managing the <span class="caps">REG</span> Fund, the advisors will be able to recommend large grants right away.</p>
</li>
</ol>
<h3 id="local-groups-in-ceas-community-building-funnel"><a class="toclink" href="#local-groups-in-ceas-community-building-funnel">Local Groups in <span class="caps">CEA</span>’s Community Building Funnel</a></h3>
<ol>
<li>
<p>In the case of the <a href="https://www.centreforeffectivealtruism.org/the-funnel-model/">community building funnel</a>, local <span class="caps">EA</span> groups are there, waiting for interested, value-aligned people to join them. (Again, this is a bit too active to be a perfect fit.)</p>
</li>
<li>
<p>The groups are ideally structured such that unaligned or exploitative people, find them uninteresting.</p>
</li>
<li>
<p>They draw on a body of talks, seminars, etc. that they have planned out in advance to educate interested newcomers. (At least once the local group is relatively established and knows how it wants to structure its events.)</p>
</li>
</ol>Beware Momentum2018-08-03T14:00:00+00:002019-04-02T20:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2018-08-03:/beware-momentum.html<p>I pose the question whether we’re again building too much momentum toward what we think is best and thereby erode our ability to react to new insights and adjust our strategy.</p><div class="admonition type">
<p class="admonition-title">Update, April 2019</p>
<p>Within half a year after I published this, it turned out that the problem – or the first problem – is that long-termist EA organizations could only absorb a tiny fraction of the newly available talent. They can’t scale at the rates at which highly skilled people apply to them.</p>
<p><a href="https://forum.effectivealtruism.org/posts/jmbP9rwXncfa32seH/after-one-year-of-applying-for-ea-jobs-it-is-really-really">Extremely smart people get turned down at high rates</a> because there is almost always someone more talented in the application queue. For example, the Open Philanthropy Project got around a thousand applications (I’m highly uncertain about this number, but a lot) and <a href="https://www.openphilanthropy.org/blog/reflections-our-2018-generalist-research-analyst-recruiting">has so far hired five</a>. “EA applicant,” 16th best medical student out of almost 6,000 to pass a standardized licensing exam, writes, “more than one person seriously encouraged me to apply for the COO position at OpenPhil (a position which went to the person who led the operations for Hillary Clinton's election campaign!).” The demand is such that that’s what the hiring bar is like. He got no offer for 20 positions he applied to.</p>
<p>At the time that I first published this article I didn’t predict what the exact failure mode would be. In hindsight, it seems fairly obvious as it could be interpreted as falling into the same reference class as the problem with movement growth.</p>
</div>
<p>A few years ago, 80,000 Hours et al. noticed that the funding gaps they saw in the spaces they prioritized seemed to be about to shrink below the levels of the talent gaps they saw. This meant they had to perform a costly reversal in the outreach strategy, from promoting earning to give as an underused opportunity to warning people against it whose comparative advantage didn’t seem maximally unambiguous.</p>
<p>Similarly, the growth of what used to be called the <span class="caps">EA</span> “movement” used to be prioritized until people started warning of the dangers of getting it wrong. But at that point it had already been gotten wrong in that today there are a lot of people who identify as EAs but don’t have the time to keep up with the bleeding-edge research (and in other ways too). In my circles, we’re even unsure whether we’re just seeing more informed people and less informed people who’ll eventually follow the first group, or whether we’re seeing a permanent fragmentation of the community.<sup id="fnref:fragmentation"><a class="footnote-ref" href="#fn:fragmentation">1</a></sup></p>
<p>Now I suspect that the same failure mode is repeating in a third domain.</p>
<p>I’m not aware of any plausible theory according to which at least <span class="caps">AI</span> safety focused on alleviating risks of extreme suffering (s-risks) is not the most pressing problem to focus on. But my filter bubble now focuses on education and recruiting for the cause almost exclusively, which also seems worrisome to me, especially when compared to prioritization research. Some considerations:</p>
<ol>
<li>Trivially, the considerations that led us to reconsider our approaches to <span class="caps">ETG</span> and growth were unknown before they were discovered. More unknown unknowns may become known that will lead us to either deprioritize <span class="caps">AI</span> safety a little or change something about our approach to it. But now we’re building such momentum that we’re eroding our ability to make such course adjustments.<sup id="fnref:neglectedness"><a class="footnote-ref" href="#fn:neglectedness">2</a></sup></li>
<li>Relatedly, I worry that the enormous momentum that gets built at the moment will lead to a continued increase in available, skilled applicants to <span class="caps">AI</span> safety organizations. For many possible underlying skill distributions, the qualification of the best applicant increases with the number of applicants. So the current development discourages hiring and growth just as deflation discourages investment. This will lead us to overestimate the current talent gap and may also delay research in general.</li>
<li>Very few people are cause neutral. Those who are have a huge comparative advantage for prioritization research, but if they get convinced that <span class="caps">AI</span> safety research is more important, they may specialize in it to the point where a switch back later would be very costly.</li>
</ol>
<p>These are only considerations, and it is well possible that the current momentum is actually not yet enough, that I’m mistaken about its degree, or a number of other countervailing considerations. So I have three main suggestions:</p>
<ol>
<li>Put a stronger focus again on prioritization research to be prepared if <span class="caps">AI</span> safety is not the most important thing after all, there are low limits to its growth, or we need to adjust our growth strategy.</li>
<li>Spread questions, not answers: A big part of the trouble with <span class="caps">ETG</span>, offsetting, and various other “answers” is that they are useful only on the margin in a certain situation. Spreading them is costly, and once it’s done, the situation will ideally change, requiring spreading of the opposite answer.<sup id="fnref:givewell-ace"><a class="footnote-ref" href="#fn:givewell-ace">3</a></sup> To sidestep these costs and risks we can spread questions, e.g., instead of spreading offsetting, we can focus outreach on making people curious to find out how they can best reduce animals suffering such that the researchers who put out the tentative answers can maintain maximal fidelity through long blog posts or papers and have low costs advertising their conclusions, whether they are, at that time, offsetting, veganism, or working for an animal rights organization.</li>
<li>Marry the generally accepted margin thinking with something like systems theory to gain a better understanding of the above failure mode and how to avoid it. Over longer time scales, we’ll probably also run into issues with exacerbating oscillations if we don’t get our reaction delays right.</li>
</ol>
<p>I’ve consulted some friends before writing this because s-risk-focused <span class="caps">AI</span> safety does seem like the most important thing to me and does still seem neglected to me, and it seemed potentially dangerous to me to risk braking down the momentum in this direction. So please don’t update on these considerations too strongly before someone has put out a good quantitative model trading these factors off against the surely many countervailing ones.</p>
<div class="footnote">
<hr>
<ol>
<li id="fn:fragmentation">
<p>This also contributes to the recent concerns about representativeness: I’d be happy to see the variety of opinions represented fairly that well-informed EAs hold, but it seems to me like a waste of everyone’s time to include opinions that majorities hold only because they haven’t had time to catch up on the intellectual process of the past years. <a class="footnote-backref" href="#fnref:fragmentation" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
<li id="fn:neglectedness">
<p>One might argue that AI safety is so enormously neglected compared to AI research that there’s no such risk for decades at least, but EA and ETG are still neglected when you choose a wide enough reference class: very few people are EAs and I want to start a social enterprise partially because I expect it to be easier to get larger investments than larger grants. Plus, I’m particularly concerned about the momentum among people with the rare advantage of cause neutrality. <a class="footnote-backref" href="#fnref:neglectedness" title="Jump back to footnote 2 in the text">↩</a></p>
</li>
<li id="fn:givewell-ace">
<p>GiveWell and ACE are doing it very deliberately and are careful to measure whether their donors really follow their recommendations or whether they become invested in the particular charities, e.g., whether they switch their donations to a different one when the first announces that their funding gap has been filled. This is a tricky process that has to be worth the costs and risks. <a class="footnote-backref" href="#fnref:givewell-ace" title="Jump back to footnote 3 in the text">↩</a></p>
</li>
</ol>
</div>Current Thinking on Priorities 20182018-03-10T14:00:00+00:002018-11-07T19:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2018-03-10:/cause-selection-2018.html<p>This article documents my current thoughts on how to make the most out of my experiment with earning to give. It draws together a number of texts by other authors that have influenced my thinking and adds some more ideas of my own for a bundle of heuristics that I currently use to make donation decisions. I hope the article will make my thinking easier to critique, will help people make prioritization decisions, will inspire further research into the phenomena that puzzle me, and will allow people to link the right books or papers to me.</p><style>
.entry-content img {
display: block;
width: 80%;
margin-left: auto;
margin-right: auto;
}
.entry-content table td,
.entry-content table th {
padding: .5em 1em;
vertical-align: top;
}
.entry-content table th {
font-weight: bold;
}
.entry-content table th:first-child {
text-align: right;
vertical-align: top;
}
.entry-content table th:first-child div {
transform-origin: top right;
transform: rotate(-90deg) translate(-.4em, -1.2em);
}
.entry-content dd {
margin-left: 1em;
}
</style>
<div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#comparing-causes">Comparing Causes</a><ul>
<li><a href="#upward-trajectories">Upward Trajectories</a></li>
<li><a href="#urgency">Urgency</a></li>
</ul>
</li>
<li><a href="#interventions">Interventions</a><ul>
<li><a href="#feedback-loops">Feedback Loops</a><ul>
<li><a href="#good-proxies">Good Proxies</a></li>
</ul>
</li>
<li><a href="#durability">Durability</a><ul>
<li><a href="#singleton">Singleton</a></li>
<li><a href="#extinction">Extinction</a></li>
<li><a href="#space-colonization">Space Colonization</a></li>
</ul>
</li>
<li><a href="#strategic-robustness">Strategic Robustness</a><ul>
<li><a href="#fundamentality">Fundamentality</a></li>
<li><a href="#value-of-information">Value of Information</a></li>
<li><a href="#instrumental-convergence">Instrumental Convergence</a></li>
<li><a href="#option-value">Option Value</a></li>
<li><a href="#backfiring">Backfiring</a></li>
</ul>
</li>
<li><a href="#counterfactual-of-funding">Counterfactual of Funding</a></li>
<li><a href="#supportability">Supportability</a></li>
<li><a href="#cooperation">Cooperation</a><ul>
<li><a href="#comparative-advantage">Comparative Advantage</a></li>
</ul>
</li>
</ul>
</li>
<li><a href="#influencing-the-future">Influencing the Future</a><ul>
<li><a href="#simple-delay-model">Simple Delay Model</a></li>
<li><a href="#cyclical-model">Cyclical Model</a></li>
<li><a href="#progress-vacuum-model">Progress Vacuum Model</a></li>
<li><a href="#threshold-model">Threshold Model</a></li>
<li><a href="#improvements-vs-delays">Improvements vs. Delays</a></li>
<li><a href="#observations">Observations</a></li>
<li><a href="#reaching-the-next-solar-system">Reaching the Next Solar System</a></li>
</ul>
</li>
<li><a href="#ideas-for-a-model">Ideas for a Model</a></li>
<li><a href="#priorities">Priorities</a><ul>
<li><a href="#influencing-the-singleton">Influencing the Singleton</a></li>
<li><a href="#wild-animal-suffering-research">Wild Animal Suffering Research</a></li>
<li><a href="#preserving-effective-altruism">Preserving Effective Altruism</a></li>
<li><a href="#other-ea-movement-building">Other EA Movement Building</a></li>
<li><a href="#cultured-meat">Cultured Meat</a></li>
<li><a href="#moral-circle-expansion">Moral Circle Expansion</a></li>
<li><a href="#humane-space-colonization">Humane Space Colonization</a></li>
</ul>
</li>
<li><a href="#appendix-changes-since-2015">Appendix: Changes Since 2015</a><ul>
<li><a href="#introduction_1">Introduction</a></li>
<li><a href="#lower-scale">Lower Scale</a><ul>
<li><a href="#feedback-cycles">Feedback Cycles</a></li>
<li><a href="#final-bricks">Final Bricks</a></li>
<li><a href="#short-term-risk-premium">Short-Term Risk Premium</a></li>
</ul>
</li>
<li><a href="#lower-neglectedness">Lower Neglectedness</a><ul>
<li><a href="#the-rational-streetlight-effect">The Rational Streetlight Effect</a></li>
<li><a href="#cost-of-information">Cost of Information</a></li>
<li><a href="#suitability-for-more-types-of-donors">Suitability for More Types of Donors</a></li>
</ul>
</li>
</ul>
</li>
<li><a href="#acknowledgements">Acknowledgements</a></li>
<li><a href="#footnotes">Footnotes</a></li>
</ul>
</div>
<h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<p>Content note: This article is written for an audience of people familiar with cause prioritization and other important <a href="https://concepts.effectivealtruism.org/">effective altruism concepts</a> and who have considered arguments for the importance of long-term impact. To others it may seem confusing unless they get their bearings first through reading such books as <em>Doing Good Better</em> and <em>Superintelligence</em>.</p>
<p>I’m aware of many arguments for the great importance of long-term strategic planning, but when “long-term” refers to millenia and more, it becomes unintuitive that we should be able to have an influence on it today with any but miniscule probability. I’m hoping to avoid having to base my work on <a href="https://en.wikipedia.org/wiki/Pascal%27s_mugging">Pascal’s Mugging</a>–type of scenarios where enormous upsides are used to compensate their miniscule probability for a still-high expected value. So I need to assess when, in what cases, or for how long we can influence the long term significantly and with macroscopic probabilities.</p>
<p>Then I need heuristics to be able to compare, very tentatively, interventions that all can’t rely on feedback cycles or only on feedback cycles around unreliable proxy measures.</p>
<p>Finally, I would like to be able to weigh assured short-term impact (in the sense of “a few centuries at best”) against this long-term impact and develop better heuristics for understanding today which strategies are more likely or unlikely to have long-term impact.</p>
<div class="admonition type">
<p class="admonition-title">Reservations</p>
<ol>
<li>The heuristics I’m using are not yet part of some larger model that would make it apparent when there are gaps in the reasoning. So there probably are huge gaps in the reasoning.</li>
<li>In particular, I’ve come up with or heard about the heuristics during a time when I got increasingly excited about Wild Animal Suffering Research, so I may have inadvertently selected them for relevance to the intervention or even for being favorable to it.<sup id="fnref:heuristics-collection"><a class="footnote-ref" href="#fn:heuristics-collection">1</a></sup></li>
<li>I’m completely ignoring considerations related to infinity, the simulation hypothesis, and faster-than-light communication or travel for now. Especially the simulation hypothesis may be an argument why even slight risk aversion or a slight quantization of probability can lead to a greater focus on the short term.</li>
</ol>
</div>
<h2 id="comparing-causes"><a class="toclink" href="#comparing-causes">Comparing Causes</a></h2>
<p><a href="https://80000hours.org/articles/problem-framework/">The importance, neglectedness, tractability (<span class="caps">INT</span>) framework</a> (or scale, crowdedness, solvability) provides good guidance for comparing between problems (also referred to as causes). It can be augmented with heuristics that make it easier to assess the three factors.</p>
<h3 id="upward-trajectories"><a class="toclink" href="#upward-trajectories">Upward Trajectories</a></h3>
<p>One addition is a heuristic or consideration for assessing scale (and plausibly tractability) that occurred to me and that seems to be very important provided that we are able to make our influence felt over sufficiently long timescales.</p>
<p>I care primarily about reducing suffering (in the broad sense of <a href="https://foundational-research.org/the-case-for-suffering-focused-ethics/">suffering-focused ethics</a> not specifically negative utilitarianism),<sup id="fnref:antirealism"><a class="footnote-ref" href="#fn:antirealism">2</a></sup> so I want my <a href="https://docs.google.com/viewer?a=v&pid=sites&srcid=ZGVmYXVsdGRvbWFpbnxuYmVja3N0ZWFkfGd4OjExNDBjZTcwNjMxMzRmZGE">trajectories</a> to point downward. Suffering of humans from poverty-related diseases <a href="https://en.wikipedia.org/wiki/The_Better_Angels_of_Our_Nature">has long been</a> <a href="https://www.gatesfoundation.org/Who-We-Are/Resources-and-Media/Annual-Letters-List/Annual-Letter-2014">going down</a>. If I were to get active in that area and actually succeed quite incredibly, the result may look like the chart below.</p>
<p><img alt="Speeding up an existing development" src="/images/cause-selection-2018/speeding-up-existing-development.png"></p>
<p>Here the sum of the two darker areas is the counterfactual suffering over time and the lighter area is the difference that I’ve made compared to the counterfactual without me.</p>
<p>But consider the case where I pick out something that I think has an increasing trajectory:</p>
<p><img alt="Dampening a negative development in the short term" src="/images/cause-selection-2018/dampening-negative-development-short-term.png"></p>
<p>When we zoom out,<sup id="fnref:clustering"><a class="footnote-ref" href="#fn:clustering">3</a></sup> this starts to look really big:</p>
<p><img alt="Dampening a negative development in the long term" src="/images/cause-selection-2018/dampening-negative-development-long-term.png"></p>
<p>In this simplified model, the impact of interventions that can and conceivably will solve the issue they’re addressing can be curtailed by intersecting with y = 0 if our influence is sufficiently durable.</p>
<p>This is a heuristic of scale and tractability. It is straightforward that the scale of a problem that is nigh infinite over all time is greater than that of a problem that is finite.</p>
<p>In practice, we will probably see an asymptotic approach toward the x axis as the cost-effectiveness of more work in the area drops further and further. This is where tractability comes in. As you approach zero disutility within some reference class, there is usually some decreasing marginal cost-effectiveness as the remaining sources of disutility are increasingly tenacious ones. If our influence is still felt at that point, it’ll rapidly become less important, something that does not happen (and maybe even the opposite<sup id="fnref:increased-marginal-utility"><a class="footnote-ref" href="#fn:increased-marginal-utility">4</a></sup>) in the case of interventions that try to dampen the exacerbation of a problem. So if I have the chance to affect some trajectory by 1° in a suffering-reducing direction, then I’d rather affect an upward trajectory than a downward one, one that afterwards will at most become constant.<sup id="fnref:slight-downward-trajectory"><a class="footnote-ref" href="#fn:slight-downward-trajectory">5</a></sup></p>
<p>Some EAs have complained that they set out to make things better and now they’re just trying to prevent them from getting much worse. But perhaps that’s just the more impactful strategy.</p>
<h3 id="urgency"><a class="toclink" href="#urgency">Urgency</a></h3>
<p>I’ve often seen urgency considered, but not as part of the <span class="caps">INT</span> framework. I think it is a heuristic of tractability. When there is an event that is important to prevent, and the event is of such a nature that preventing it is all-important but there’s nothing left to do if and once it happens, then all the leverage we have over the value of the whole future is condensed into the time leading up to the potential event.</p>
<p>Survival analysis is relevant to this problem, because there are two cases, one where the event happens and then nothing (e.g., some suffering-intense singleton emerges and continues throughout the future – <a href="https://foundational-research.org/s-risks-talk-eag-boston-2017/">a suffering risk or s-risk</a>) and one where the event doesn’t happen but may, as a result, still happen just a little later.</p>
<p>There is some <a href="https://en.wikipedia.org/wiki/Survival_function">survival function</a> that describes our chances of survival beyond some point in time. It may be some form of exponential decay function. These have a mean, the expected survival duration, which may serve as a better point for comparisons of urgency than the earliest catastrophic event.</p>
<p>If we wait with preventing catastrophes for so long that at the time when we do start to work on them the duration of the future that we can expect to influence is longer than the duration of the future we can expect to survive for, then it seems to me that we waited for too long. Conversely, if the expected end of our influence is earlier than our demise, it might be better to first focus on either more urgent causes or causes that don’t have such a strong event character. But first we’ll need to learn more about the likely shape of the decay function of our influence.</p>
<div class="admonition type">
<p class="admonition-title">Further Research</p>
<p>It would also be interesting to investigate whether we can convert the survival function for an s-risk-type of catastrophe<sup id="fnref:s-risks"><a class="footnote-ref" href="#fn:s-risks">6</a></sup> into an expected suffering distribution, which we could then compare against expected suffering distributions from suffering risks that don’t have event character. The process may be akin to multiplying the failure rate with the suffering in the case of the catastrophe, but since the failure rate is not a probability distribution, I don’t think it’s quite that easy.</p>
</div>
<h2 id="interventions"><a class="toclink" href="#interventions">Interventions</a></h2>
<p>Some of my considerations for comparing interventions are less well established than the scale, tractability, neglectedness framework for problems. Michael Dickens proposed a tentative <a href="http://effective-altruism.com/ea/yp/evaluation_frameworks_or_when_importance/">framework</a> for assessing the effectiveness of interventions, but it is somewhat dependent on the presence of feedback loops. Apart from that, however, it makes the important contribution of considering the strength of the evidence in addition to what the evidence says about the expected marginal impact.</p>
<p>I will propose a few heuristics other than historical impact for evaluating interventions. In each case it would be valuable to not only count the heuristic as satisfied or not but also multiply in once confidence with that verdict. Michael Dickens, for example, generated likely <a href="http://effective-altruism.com/ea/ws/on_priors/">background distributions</a> of intervention cost-effectiveness to adjust his confidence in any particular estimate accordingly.</p>
<h3 id="feedback-loops"><a class="toclink" href="#feedback-loops">Feedback Loops</a></h3>
<p>Confusingly, I’m listing feedback loops as one of the heuristics. But I don’t mean it in the sense that we should necessarily draw on historical evidence when evaluating the cost-effectiveness of an intervention (we should do so whenever possible of course) but that the presence of feedback loops that are short enough will be very valuable for the people running the intervention. So if there’s a chance that they can draw on them, it’s a big plus for the intervention enabling iterative improvements and reduced risks of updating around aimlessly on mere noise.</p>
<h4 id="good-proxies"><a class="toclink" href="#good-proxies">Good Proxies</a></h4>
<p>We can almost never measure the outcomes we care about most directly but almost always have to make do with proxies, so the reliability of these proxies is an important consideration.</p>
<h3 id="durability"><a class="toclink" href="#durability">Durability</a></h3>
<p>There are two failure modes here: the desired change may be reversed at some point in the future or it may have eventually happened anyway. For example, <a href="http://effective-altruism.com/ea/xp/lessons_from_the_history_of_animal_rights/">vegan and vegetarian societies have not been stable in the past</a> or valuable research may get done in academia only a little later than at an <span class="caps">EA</span>-funded institute.<sup id="fnref:cooperatively-waiting-for-research"><a class="footnote-ref" href="#fn:cooperatively-waiting-for-research">7</a></sup> Another failure mode for many interventions is that some form of singleton may take control of all of the future so that we can influence the future only by proxy through influencing what singleton emerges.</p>
<p>I can see three possible futures at the moment: (1) the emergence of a singleton, a wildcard for now, (2) complete annihilation of sentient or just intelligent life, or (3) space colonization. I’ll mention in the following why I’m conflating some other scenarios into these three. (This is heavily informed by Lukas Gloor’s “<a href="http://effective-altruism.com/ea/1k4/draft_cause_prioritization_for_downsidefocused/">Cause Prioritization for Downside-Focused Value Systems</a>.”)</p>
<h4 id="singleton"><a class="toclink" href="#singleton">Singleton</a></h4>
<p>I’ve heard opinions that implied that a permanent singleton is possible and others that leaned toward thinking that value preservation under self-improvement is so hard that value drift is inevitable. Perhaps this path may enable us to create a wonderful populated utopia or attempts at it might lead to “<a href="https://foundational-research.org/cause-prioritization-downside-focused-value-systems/">near misses [that] end particularly badly</a>.” Improving our strategy for preparing for this future seems very valuable to me.</p>
<p>A seeming singleton that turns out unstable after a while may be as deleterious for our ability to influence the future via other means than the singleton itself as an actual (permanent) singleton. Artificial general intelligence (<span class="caps">AGI</span>) may turn any sufficiently technologically advanced state that is not a singleton into a distractor state, in which case many of my thoughts in this text are moot. So this case is crucial to consider.</p>
<h4 id="extinction"><a class="toclink" href="#extinction">Extinction</a></h4>
<p>In the cases of human extinction or extinction of all life, it would be interesting to estimate the expected time it’ll take for another at least similarly intelligent species to emerge. <a href="http://effective-altruism.com/ea/1lf/paper_surviving_global_risks_through_the/">Turchin and Denkenberger</a> have attempted this, yielding a result of something to the order of 100 million years. Such a long delay may significantly reduce the maximal disvalue (and also value, for less suffering-focussed value systems) of space colonization because resources will be more thinly spread (time-consuming to reach) throughout the greatly expanded universe and some other factors.<sup id="fnref:long-delays"><a class="footnote-ref" href="#fn:long-delays">8</a></sup> However, space colonization may still happen relatively quickly if there are more species within our reachable universe who are also just some millenia away from starting to colonize space.<sup id="fnref:models-of-galactic-spreading"><a class="footnote-ref" href="#fn:models-of-galactic-spreading">9</a></sup></p>
<p>But <a href="http://effective-altruism.com/ea/1k4/draft_cause_prioritization_for_downsidefocused/">Lukas Gloor thinks</a> “that large-scale catastrophes related to biorisk or nuclear war are quite likely (~80–90%) to merely delay space colonization in expectation,” with <span class="caps">AGI</span> misalignment and nanotechnology posing risks of a different category.<sup id="fnref:recovery-after-catastrophe"><a class="footnote-ref" href="#fn:recovery-after-catastrophe">10</a></sup> So a wide range of seeming extinction risks may allow our civilization to recover eventually, at which point the future will still hold these three options.</p>
<h4 id="space-colonization"><a class="toclink" href="#space-colonization">Space Colonization</a></h4>
<p>Even if this repeats many times, it is likely that eventually one of these civilizations will either go completely extinct, form a singleton, or colonize space. The third option would eliminate many classes of extinction risks, the ones that are only as severe as they are because we’re physically clustered on only one planet. Extinction risks that only depend on communication will remain for much longer.</p>
<p>Risks from artificial intelligence will probably only depend on communication<sup id="fnref:ai-and-communication"><a class="footnote-ref" href="#fn:ai-and-communication">11</a></sup> so that it will have to turn out that artificial intelligence can be controlled permanently for space colonization to happen. This surely makes this scenario less likely than the previous two. (Or more precisely, though such precise language may imply a level of confidence I don’t have, less than one third of the probability mass of the future may fall on this scenario.) Perhaps, though, we have a greater chance to influence the outcome or influencing it may require different skill sets than influencing artificial intelligence.</p>
<p>But while humans may look out for one another, we have a worse track when it comes to those who are slightly outside our culture (or involved in it only in a unidirectional way where they can’t make demands of their own in “trades” with them). <a href="https://sentience-politics.org/files/Dello-Iacovo-On-terraforming-wild-animal-suffering-and-the-far-future.pdf">Michael Della-Iacovo</a> considers that insects may be spread to other planets (or human habitats in space) as a food source or even by accident. Eventually, farmed animals may follow, and they may return to being wild animals even when animal farming becomes obsolete due to such technologies as cultured meat. People may even spread wild animals for aesthetic reasons. Finally, the expected capacity for disutility of bacteria may be small, but large numbers of them play vital roles in some proposed strategies for terraforming planets.</p>
<p>So the graphs in “Upward Trajectories” are inaccurate in the important way that suffering is likely to expand spherically as some cubic polynomial of the radius from our solar system – not linearly.<sup id="fnref:cubic-spreading"><a class="footnote-ref" href="#fn:cubic-spreading">12</a></sup></p>
<p>If we want to influence a future of space colonization positively, then it is all-important to make sure that our influence survives to the start of the space colonization era. (I will qualify this further in “Reaching the Next Solar System” below.) Then, longer survival of the influence becomes rapidly less important: Communities will likely cluster because of the lightspeed limit on communication latency,<sup id="fnref:faster-than-light"><a class="footnote-ref" href="#fn:faster-than-light">13</a></sup> and if one of these communities loses our influence or would’ve reaped its benefits even without us, it’ll carry this disvalue or opportunity cost outward only into an increasingly narrow cone of uncolonized space.</p>
<h3 id="strategic-robustness"><a class="toclink" href="#strategic-robustness">Strategic Robustness</a></h3>
<p>I specify that I mean <em>strategic</em> robustness because I’ll mention something that I’ll refer to as <em>ethical</em> robustness later, but this strategic robustness is just what others mean when they just say “<a href="http://www.stafforini.com/blog/bostrom/">robustness</a>.” Strategic robustness unfortunately overlaps with what I called durability, but I try to disentangle it by using <em>robust</em> to refer to strategies that are beneficial across many different possible futures while I use <em>durable</em> to refer to strategies that have a chance to survive until we colonize space, if it should come to that. Insofar as the second is included in the first, it is a small (but important) part of the concept, so I hope not to double-count too much.</p>
<h4 id="fundamentality"><a class="toclink" href="#fundamentality">Fundamentality</a></h4>
<p>In <a href="https://claviger.net/the-bulk-of-the-impact-iceberg.html">The Bulk of the Impact Iceberg</a>, I argue that any interventions that observably produce the impact we care about rest on a foundation of endless amounts of preparatory work that is often forgotten by the time the impact gets generated by the final brick intervention. Some of this preparatory work may not get done automatically. This suggests that highly effective final brick interventions are few compared to equally or more effective preparatory interventions.</p>
<p>A particularly interesting question is here whether there are heuristics for determining where the most important bottlenecks are in these trees of preparatory work because that is where we should invest most heavily. Studying advertisement may be helpful for noticing and testing such heuristics. (More on that in the abovelinked article.)</p>
<p>I particularly highlight (and use as one such heuristic for now) research as a highly effective interventions, because, if it is done right and studies something that is sufficiently plausible, it is valuable whether it is successful or seemingly fails.</p>
<h4 id="value-of-information"><a class="toclink" href="#value-of-information">Value of Information</a></h4>
<p>Feedback loops generate valuable information to improve an intervention incrementally. In this context of robustness, however, I’m referring to information that is generated that doesn’t only benefit the very intervention that generates it but is more widely beneficial. People working in the area of <a href="https://claviger.net/cause-area-human-rights-in-north-korea.html">human rights in North Korea</a>, for example, may not be working on the most large-scale problem there is, and tractability and neglectedness may not make up for that, but they may gain skills in influencing politics and gain insights into how to avoid similar nigh-singletons as North Korea in the future.</p>
<h4 id="instrumental-convergence"><a class="toclink" href="#instrumental-convergence">Instrumental Convergence</a></h4>
<p>Anything that we have already determined to be instrumentally convergent – i.e. something that a wide range of agents are likely to do no matter their ultimate goals – is convergent for reasons of its robustness. So all else equal, an intervention toward a convergent goal is a robust intervention. Gathering knowledge, intelligence, and power are examples of such robust capacity building.</p>
<h4 id="option-value"><a class="toclink" href="#option-value">Option Value</a></h4>
<p>Things that destroy option value are bad (all else equal):</p>
<ol>
<li>If you work on some intervention for a decade and then find that it is not the best investment of your time, it’s best if the work still helped you build transferable skills and contacts, which you’ll continue to profit from.</li>
<li>If you never defect against other value systems even when it would be useful to you, you don’t lose the option to continue cooperate with them or to intensify your cooperation.</li>
<li>If you remain <a href="http://effective-altruism.com/ea/18i/hardtoreverse_decisions_destroy_option_value/">neutral or low-profile</a>, you retain the ability to cooperate with a wide range of agents and can <a href="http://globalprioritiesproject.org/wp-content/uploads/2015/05/MovementGrowth.pdf">avoid controversy and opposition</a>.</li>
<li>If you avoid making powerful enemies you can minimize risks to yourself or your organization. The individual or the organization are often points of particular fragility in an intervention so long as relatively few people are working on it or they are fairly concentrated. (The failure of an organization, be it for completely circumstantial reasons, may also discourage others from founding another similar organization.)</li>
</ol>
<p>Relatedly, something that I like to call “control” is useful because it implies higher option value: If you can change course quickly and at a low cost you have more control than if you could do the same but at a somewhat higher cost or more slowly.</p>
<h4 id="backfiring"><a class="toclink" href="#backfiring">Backfiring</a></h4>
<p>Finally, robustness is nil if the risk of very bad counterfactually significant outcomes is too high.</p>
<h3 id="counterfactual-of-funding"><a class="toclink" href="#counterfactual-of-funding">Counterfactual of Funding</a></h3>
<p>Valuable work that is funded by money or time that would’ve otherwise benefited almost as valuable work is not terribly valuable. Resource constraints are rarely so clear cut, so we may never know whether it was great, good, eh,<sup id="fnref:eh"><a class="footnote-ref" href="#fn:eh">14</a></sup> or bad for <a href="https://80000hours.org/2016/02/the-value-of-coordination/">Rob Wiblin to switch to 80,000 Hours</a>, but if you can achieve the same thing whether you’re funded by EAs or by a <span class="caps">VC</span>, then it’ll likely make a big difference.</p>
<h3 id="supportability"><a class="toclink" href="#supportability">Supportability</a></h3>
<p>Even a problem whose solution is tractable may have an intervention addressing it that is hard to support, e.g., because you’d have to have a very particular skill set or the intervention is so far only a hypothetical one.</p>
<h3 id="cooperation"><a class="toclink" href="#cooperation">Cooperation</a></h3>
<p>No two people have completely identical values, and often values are significantly different. Battling this out in a battle of force or wits has rather flat expected utility dependent on the actors relative strengths <a href="https://foundational-research.org/gains-from-trade-through-compromise/">while cooperation lets them find Pareto-efficient points</a>:</p>
<ol>
<li>Depending on the shape of the frontier, this can be an extremely important consideration even for a risk neutral agent with a single value system:<ul>
<li>To use the example from the article linked above, a special deep ecologist may care equally about 8e6 species totalling <a href="http://reducing-suffering.org/how-many-wild-animals-are-there/">2e18 individuals</a> or the same number of species totalling radically more individuals. If most of these individuals suffer badly enough throughout their lives, a suffering reducer may greatly prefer the relatively smaller number. A trade would be very valuable for the suffering reducer – if they’re roughly equally powerful, then many times more than a battle in expectation.</li>
</ul>
</li>
<li>But it is also very decisive for me because I empathize with many value systems, so I want to do the things that are good or neutral for all of them – what I refer to as ethical robustness.</li>
<li>Even if the frontier is such that the gains from trade are minor, a zero-sum game may look more like a battle than a race, so that time is irreversibly lost, time that could be used to avert suffering risks or preserve important suffering-reducing knowledge across existential catastrophes.</li>
<li>More speculatively, insofar as our decisions can be correlated with those of sufficiently similar agents who yet have different moral goals, we can determine the results of the different instantiations of the decision algorithm that we share with them and <a href="https://foundational-research.org/multiverse-wide-cooperation-via-correlated-decision-making/">thus get them to cooperate with our moral system too</a>.</li>
</ol>
<h4 id="comparative-advantage"><a class="toclink" href="#comparative-advantage">Comparative Advantage</a></h4>
<p>Agents with strong commitments to cooperation will still face hurdles to achieving that cooperation because they may not be able to communicate efficiently. They can apply heuristics such as <a href="https://blog.givewell.org/2015/11/25/good-ventures-and-giving-now-vs-later/">splitting their investments</a> (see the “Coordination Issues” section in the linked article) into shared projects equally or focussing on the issues that they know few others <em>can</em> focus on, so capitalizing on their comparative advantage.</p>
<p>Altruists at government institutions or many foundations may face greater requirements to make strong cases for their investment decisions that they can point to later when something goes wrong or a project fails to avoid being blamed personally. Private donors or the Open Philanthropy Project can thus cooperate with the former group by supporting the types of projects that the former group can’t support.<sup id="fnref:audience"><a class="footnote-ref" href="#fn:audience">15</a></sup></p>
<h2 id="influencing-the-future"><a class="toclink" href="#influencing-the-future">Influencing the Future</a></h2>
<p>The first space colonization mission is about to be launched but the team of meteorologists is split over whether the weather is suitable for the rocket launch. Eventually, the highest-ranking meteorologist decides that if their colleagues can’t agree then the weather situation is too uncertain to permit the launch.</p>
<p>A week later, the rocket launches, and it sets a precedent for others. Innovation accelerates, prices drop, and eventually humans expand beyond the solar system far out into the Milky Way launching new missions in all directions by the minute. And all the while they take with them bacteria, insects, and many larger animals, may <a href="https://foundational-research.org/the-importance-of-wild-animal-suffering/">r-strategists</a> among them. Unbeknownst to everyone, though, the launch would’ve been successful even the first time.</p>
<h3 id="simple-delay-model"><a class="toclink" href="#simple-delay-model">Simple Delay Model</a></h3>
<p>What influence does the one-week delay have one millennium later?<sup id="fnref:millennia"><a class="footnote-ref" href="#fn:millennia">16</a></sup> Intuitively, I would think that is has close to no impact, but a very simple model is not enough to confirm that impression.</p>
<p><img alt="Simple delay model" src="/images/cause-selection-2018/simple-delay-model.png"></p>
<p>The chart assumes that I’m the meteorologist and that I’ve turned a world whose suffering would’ve looked like the blue line, <em>t³</em>, into the red-line world, <em>(t-1)³</em>. The yellow line is the difference in suffering between the worlds: <em>3t² - 3t + 1</em>. That’s a ton, and rather than diminishing over time like my intuition intuited, it increases rapidly.</p>
<p>What decreases rapidly, however, is the size of the difference relative to the suffering that increases even more rapidly – the green dotted line: <em>(3t² - 3t + 1) / (t - 1)³</em>. So is it just this morally unimportant relative difference that caused my intuition to view our influence as something like an exponential decay function when really it will increase throughout the next millennia? Dan Ariely certainly thinks that we erroneously reason about money in relative rather than absolute terms, fretting about cents and dollars when buying sponges but lightly throwing around thousands of dollars when renovating a house or speculating on cryptocurrencies.</p>
<h3 id="cyclical-model"><a class="toclink" href="#cyclical-model">Cyclical Model</a></h3>
<p>Or maybe the intuition is informed by cyclical developments.</p>
<p><img alt="Cyclical model" src="/images/cause-selection-2018/cyclical-model.png"></p>
<p>Some overwhelming outside force may limit our potential absolutely at some point in time and force us back, and from that point onward, the suffering trajectory would only be conditional on the overwhelming force rather than our influence from the past.</p>
<p>Seasons, winter in particular, are an obvious example of cyclical forces like that, but in the context of space colonization, I can’t think of reason for it just yet – especially once several solar systems have been colonized with decades of communication delay between them and probably no exchange in commodities given how much longer they would have to travel.</p>
<h3 id="progress-vacuum-model"><a class="toclink" href="#progress-vacuum-model">Progress Vacuum Model</a></h3>
<p>But what about correlated forces rather than mysterious overwhelming outside forces?</p>
<p><img alt="Progress vacuum model" src="/images/cause-selection-2018/progress-vacuum-model.png"></p>
<p>Perhaps that week of delay saw one week of typical progress in charting space, searching for exoplanets, and advancing technology, so that the delay created one more week of “progress vacuum,” which eventually backfired by speeding up progress beyond the slope of the trajectory of the counterfactual world.</p>
<p>But the graph above is probably an exaggeration. When you reduce the supply by slowing some production process down, the demand reacts to that in complex ways due to various threshold effects, and the resulting cumulative elasticity factor is just that, a factor, a linear polynomial. If our influence is a quadratic polynomial, as surmised in the first diagram, then multiplying it with a small linear polynomial is not going to have a significant negative impact in the long term.</p>
<h3 id="threshold-model"><a class="toclink" href="#threshold-model">Threshold Model</a></h3>
<p>But might we face more mighty thresholds?</p>
<p><img alt="Threshold model" src="/images/cause-selection-2018/threshold-model.png"></p>
<p>Now that’s one mighty threshold right there! Once we’re spread throughout space, it becomes harder to think of thresholds like that because it has to have an expected impact that is at least commensurate to a quadratic polynomial of time or radius. But so long as the radius is constant and we’re still down to earth, a conference may be all that’s needed: a research group may present some seminal paper at a conference – delay them by one week, and they’ll hurry up and have a typo more in the presentation, but they will still present their results to the world at the same moment.</p>
<p>In conclusion, I think, based on just these rough considerations, that the cyclical model will continue to lose relevance but that the correlated speed-up (or supply/demand elasticity) model and the threshold model – the first probably just a smoothed out version of the second – will continue to be highly relevant for as long as we’re still in one solar system or even on one planet.</p>
<h3 id="improvements-vs-delays"><a class="toclink" href="#improvements-vs-delays">Improvements vs. Delays</a></h3>
<p>Intentionally delaying things that others care about is a bit uncooperative. The intervention that I’m most excited about at the moment, making space colonization more humane for nonhumans, would instead aim to reduce the suffering footprint space colonization without interfering with its rate, so for example 0.9<em>t</em>³ instead of (<em>t</em>-1)³. The difference to the counterfactual is now a cubic polynomial itself!</p>
<p>For short time scales, the delay approach is still ahead:</p>
<p><img alt="Improvement vs. delay (short term)" src="/images/cause-selection-2018/improvement-vs-delay-short-term.png"></p>
<p>But then the cubic polynomial of the improvement approach of course quickly catches up:</p>
<p><img alt="Improvement vs. delay (long term)" src="/images/cause-selection-2018/improvement-vs-delay-long-term.png"></p>
<p>Now how can my intuition that impact decays still be true when we’re dealing with a polynomial that is yet another degree higher?</p>
<ol>
<li>Thresholds would have to look different than papers presented at conferences to limit out impact in this scenario.</li>
<li>A progress vacuum is also not as straightforward. But if we put in a lot of work to advance a technology for making space habitable that minimizes suffering and so make it the go-to technology by the time the exodus starts (see “Humane Space Colonization” below), we’re establishing it through some degree of path dependence.<ol>
<li>If it turns out that it wasn’t the most efficient technology, the path dependence may not be enough to lock it in permanently – just as it is imaginable that Colemak might eventually replace Qwerty.</li>
<li>Or fundamental assumptions of the technology we locked in that way may cease to apply – more comparable to how Neuralink might eventually replace Qwerty.</li>
<li>But in either case, our impact may only degrade to delay levels, not to zero.</li>
</ol>
</li>
</ol>
<p>But if our impact still decays and if the period until the cubic polynomial overtakes the quadratic one is suitable in length, then we might still be better off delaying. Perhaps there’s potential for a moral trade here: We’ll help you with your technological progress but in return you have to commit to long-term funding of our work to make it more humane.</p>
<div class="admonition type">
<p class="admonition-title">Further Research</p>
<ol>
<li>What are the strongest risks to our impact with the improvement approach?</li>
<li>Can we build some sort of technological obsolescence model off the idea that assumptions follow a tree structure where technologies closer to the root make fewer assumptions – they are harder to invent but there are fewer ways to make them obsolete – whereas the opposite is true for inventions closer to the leaves, and that in the shape of an exponential relationship, thus introducing a factor that can perhaps, in some way I haven’t quite fleshed out, still explain my decay intuition?</li>
<li>When r-strategists reproduce they might "multiply" exponentially within the Malthusian bounds. If we can hope to affect little of the future, this may be highly relevant. Even if not, might we be able exert greater force on the polynomial than a delay has by influencing the factor of the exponent of such exponential but bounded growth?</li>
</ol>
</div>
<h3 id="observations"><a class="toclink" href="#observations">Observations</a></h3>
<p>Which of these models might best fit any observations that we can already make? To get at least some intuition for what we’re dealing with, I want to draw comparisons to some better-known phenomena.</p>
<table>
<tr>
<th></th>
<th>Factual</th>
<th>Counterfactual</th>
</tr>
<tr>
<th>
<div>
People<sup id="fnref-etg-influence"><a class="footnote-ref" href="#fn-etg-influence">17</a></sup>
</div>
</th>
<td>
Plausibly more ephemeral (than concepts<sup id="fnref-concepts-and-ideas"><a class="footnote-ref" href="#fn-concepts-and-ideas">18</a></sup>), because (1) they still die.
Plausibly more durable, because (1) their values may drift, talents and passions change, etc. at a higher rate than they can promote them to others.
</td>
<td>
Plausibly more ephemeral, because (1) the usual concerns with replaceability.
</td>
</tr>
<tr>
<th>
<div>
Companies<sup id="fnref-uspto"><a class="footnote-ref" href="#fn-uspto">19</a></sup>
</div>
</th>
<td>
Plausibly more ephemeral, because: (1) they fail, probably in most cases, for a host of other reasons than that their core idea becomes obsolete and (2) because they are highly concrete and thus fragile.<sup id="fnref-companies"><a class="footnote-ref" href="#fn-companies">20</a></sup>
Plausibly more durable, because: (1) they can adapt (change their nature) to survive – the idea is not their essence as it is for concepts.
</td>
<td>
Plausibly more durable, because: (1) they are unlikely to would have been founded anyway, under the same name, if they hadn’t been founded when they were.<sup id="fnref-counterfactual-homonymy"><a class="footnote-ref" href="#fn-counterfactual-homonymy">21</a></sup>
</td>
</tr>
<tr>
<th>
<div>
Languages
</div>
</th>
<td>
Plausibly more durable, because: (1) they have higher path dependence than most<sup id="fnref-most"><a class="footnote-ref" href="#fn-most">22</a></sup> concepts.
</td>
<td>
Plausibly more durable, because: (1) they are unlikely to would have developed the same way anyway if they hadn’t developed when they did.<sup id="fnref-independent-language-development"><a class="footnote-ref" href="#fn-independent-language-development">23</a></sup>
</td>
</tr>
<tr>
<th>
<div>
Cities
</div>
</th>
<td>
Plausibly more durable, because: (1) they have higher path dependence than most concepts and (2) they are often continuously nourished by something geographic, such as a river, which is probably unusually permanent.
</td>
<td>
Plausible more ephemeral, because: (1) at least some places are so well-suited for cities<sup id="fnref-city-places"><a class="footnote-ref" href="#fn-city-places">24</a></sup> that if one hadn’t been founded there when it was, it would’ve been founded there little later.
</td>
</tr>
</table>
<p>A human with stable values can work for highly effective short-term-focused charities, advocate for them, or earn-to-give for them for 50 years, which may serve as a lower bound for the durability of the influence we can plausibly have.</p>
<p>There are <a href="http://www.bbc.com/news/business-16611040">few companies</a> that survived more than millenium (but probably wouldn’t have if it didn’t inhabit a small, stable geographic niche), and hoping to create one that scales and yet survives even a century is probably unreasonable, since <a href="http://www.gongol.com/research/economics/companyage/">large companies are not very old</a>. A large company may be better positioned to get old than a small one, but there’s probably not enough space for enough large companies so that they can have the same outliers that there are among small companies.</p>
<p>Languages and cities still exert influences several millennia later, but because of the high probability that the oldest cities would’ve been founded anyway in short intervals because of their geography, languages are probably the stronger example. But they only have “extrapolatory power” for very path dependency–causing concepts.</p>
<p>Language may have emerged <a href="https://en.wikipedia.org/wiki/Proto-Human_language#Date_and_location">some 50–100 thousand years ago</a>, but the proto-languages that can still be somewhat reconstructed were spoken a mere <a href="https://linguistics.stackexchange.com/questions/13192/oldest-proto-languages">10,000</a> or possibly <a href="https://en.wikipedia.org/wiki/Proto-Afroasiatic_language">18,000</a> years ago.</p>
<div class="admonition type">
<p class="admonition-title">Further Research</p>
<p>I haven’t found evidence for whether linguistic reconstruction errs more on the side of reconstructing noise – that a proto-language could’ve had many different forms with no systematic impact on today’s languages, so that attempts at its reconstruction yield a wide range of different results – or on the side of silence – that the reconstruction is not possible even though today’s languages would’ve been different had the proto-language been different. It may be interesting to weigh the factors that indicate either direction.</p>
</div>
<p>So we can perhaps hope for our influence to last between half a century and a couple thousand years.</p>
<p>If we want our influence to reach other solar systems – and we travel there in a waking state – we have to travel these, let’s say, 10 light years at the 100th to 1000th part of the speed of light. That’s some 30–300 times the 356,040 km/h that <a href="http://www.guinnessworldrecords.com/news/2016/7/nasa%E2%80%99s-juno-probe-recognised-by-guinness-world-records-as-fastest-ever-spacecraft">Helios 2</a> reached, probably the fastest human-made object in space to date.<sup id="fnref:time-dilation"><a class="footnote-ref" href="#fn:time-dilation">25</a></sup> That doesn’t seem unattainable.</p>
<p>Furthermore, in a sleeping or suspended (in the case of emulations) state the distance doesn’t matter. Such missions will probably be expensive, so they’ll need to be funded by many stakeholders that will have particular interests and goals, such as harnessing the energy of another sun. They’re more likely to fund a mission with suspended passengers who can’t change the goals of the mission when the stakeholders can no longer intervene.</p>
<p>Unfortunately, that doesn’t necessarily mean that the new ships will be launched soon and none of our influence will be lost during the travel. Stakeholders will aim to make a good trade-off between an earlier launch and higher speeds, because waiting for technological innovation in, say, propulsion systems may be worth it and result in an earlier arrival. So the launch may be delayed, and if earlier enterprises get their predictions wrong and innovation happens faster than they thought, then their ships may even be overtaken by later, more value-drifted ships.</p>
<h3 id="reaching-the-next-solar-system"><a class="toclink" href="#reaching-the-next-solar-system">Reaching the Next Solar System</a></h3>
<p>The jump from colonizing our solar system (where communication is feasible) and colonizing others is huge: The <a href="https://www.universetoday.com/15585/diameter-of-the-solar-system/">diameter of our solar system</a> is in the area of 0.001 light years – around 9 light hours – while the <a href="https://www.spaceanswers.com/deep-space/whats-the-nearest-solar-system-to-our-own/">closest solar system</a>, around the star Epsilon Eridani, is 10.5 light years away and another one, around the star Gliese 876, full 15.3 light years. (Infeasibly far for light-speed communication with our solar system.) So the different levels of feasibility of communication mean that a range of existential risks will diminish much earlier than any reduction of risk from value drift.</p>
<h2 id="ideas-for-a-model"><a class="toclink" href="#ideas-for-a-model">Ideas for a Model</a></h2>
<p>Maybe I’ll find the time to create a quantitative model to trade off short-term-focused and long-term-focused activities in view of the scenario where space colonization either happens before any civilizational collapse occurs or where our influence survives the collapse or collapses.</p>
<p><a href="http://getguesstimate.com/">Guesstimate</a> has the added benefit that the limit to 5,000 samples introduces a quantization that, incidentally, protects against Pascal’s Mugging type of calculations unless you reload very often. That will make the conclusions we can draw from it more intuitive or intuitively correct.</p>
<p>The model should take at least these factors into account:</p>
<ol>
<li>Our colonization of space can probably be modeled as a spherical expansion at some fraction of the speed of light.<ol>
<li>It might take into account that faster-than-light travel might be discovered.</li>
<li>It should assume small fractions or otherwise take relativistic effects into account.</li>
</ol>
</li>
<li>It should take the expansion of space into account, which will put a limit on the maximum region of space that can be colonized – less than the Hubble volume.</li>
<li>It should consider the number and perhaps the differences in density of solar systems<ol>
<li>Since a large fraction of our time will be spent in transit, it may also be important to investigate whether the transit would likely be spent in a suspended or sleep state<sup id="fnref:ems"><a class="footnote-ref" href="#fn:ems">26</a></sup> that should preserve all properties of the civilization or in an active state where the civilization continues to evolve.</li>
</ol>
</li>
<li>If we assume that there’s no faster-than-light communication, then civilizations will probably cluster or otherwise have little effect on each other. Solar systems may be a sensible choice for the expanse of such clusters because they also have an energy source.<ol>
<li>Whole brain emulation seems to me (mostly from considerations in <em>Superintelligence</em>) more quickly achievable than large-scale space colonization, so between already-colonized regions of space, communication and travel will become similar concepts.<ol>
<li>Emulation may also become possible through training models of yourself with Neuralink.</li>
</ol>
</li>
</ol>
</li>
<li>It should include some sort of rate of decay of influence, perhaps modeled as a risk per time, resulting into a decay function such as <a href="http://mathworld.wolfram.com/ExponentialDecay.html">exponential decay</a>. The following “risks” should contribute to it, but separately as I will argue below:<ol>
<li>Decay due to destruction of the civilizational cluster.<ol>
<li>Crucial so long as we’re huddled together on earth, afterwards probably limited to planets more than to whole civilizational clusters.</li>
</ol>
</li>
<li>Decay due to random “value” drift, but not limited to values<ol>
<li>Likely to be relevant for whole civilizational clusters, so probably close to impossible to overcome for us today.</li>
</ol>
</li>
<li>Decay (of our influence) due to independent, counterfactual discovery of our contribution.<ol>
<li>Dito.</li>
</ol>
</li>
</ol>
</li>
</ol>
<h2 id="priorities"><a class="toclink" href="#priorities">Priorities</a></h2>
<p>I think it makes more sense for me to put effort into prioritizing between different broad “classes of interventions” than individual charities, because if the charities are at all competent, they’re likely more competent than me. And within each cause area, I should just be able to ask the charities which one I should support, or else they’d fail my cooperativeness criterion. (Except, I may not notice many defections.)</p>
<p>Just this “class of intervention” is something they’re probably somewhat locked in to, either by by-laws, comparative advantage through early specialization, or some feature of the team’s ethical system.</p>
<p>So below I try to make sometimes a bit artificial distinctions and then try to apply my heuristics in the form of pros and cons of the cause area and the class of interventions. When an intervention has more cross-cutting benefits, then this scheme breaks down quickly – moral circle expansion, for example, can be beneficial in how it influences futures without singleton and in how it influences what singleton emerges, but the latter benefits are counted only in the “influencing the singleton” section. I think the usefulness of this section doesn’t go much beyond a mere brainstorming.</p>
<p>And note that it doesn’t make sense to count the pros and cons below because their weights are vastly different or my confidences are vastly different or they’re highly correlated or they’re just questions or some combination of these.</p>
<h3 id="influencing-the-singleton"><a class="toclink" href="#influencing-the-singleton">Influencing the Singleton</a></h3>
<p>My heuristics don’t help much to evaluate this one until I have more clarity on how to convert a hazard function (failure rate) into an expected suffering distribution. I continue to see immense variance in this whole future scenario because of the wide range of different worlds we might get locked in to.</p>
<table>
<tr>
<th></th>
<th>Pro</th>
<th>Con</th>
</tr>
<tr>
<th><div>Problem</div></th>
<td>
<dl>
<dt>Scale or upward trajectories</dt>
<dd>See <span class="caps">WASR</span> below. The right singleton may prevent or limit it. Conversely, the wrong singleton may create enormous suffering in other forms.</dd>
<dt>Neglectedness</dt>
<dd>Still substantial.</dd>
</td>
<td>
<dl>
<dt>Tractability</dt>
<dd>Very hard to get right, though Lukas Gloor guesses that the chances are macroscopic.</dd>
</dl>
</td>
</tr>
<tr>
<th><div>Intervention</div></th>
<td>
<dt>Durability</dt>
<dd>Maximal, ipso facto.</dd>
<dt>Urgency</dt>
<dd>Plausibly most urgent. In particular, strong <span class="caps">AGI</span> may come before space colonization.</dd>
<dt>Robustness</dt>
<dd>Even if it doesn’t, the absence of a singleton (through <span class="caps">AGI</span>) in the face of advanced technology may even generally be a distractor state. In that case, space colonization is unlikely to proceed fast enough to prevent an <span class="caps">AGI</span> from quickly controlling all (up to that point) colonized space. (More on that in the section “Ideas for a Model.”)</dd>
</td>
<td>
<dt>Supportability</dt>
<dd>I know too little about how suffering focussed various organizations are and feel like the space needs mostly people with very specific and rare skill sets.<sup id="fnref-fri"><a class="footnote-ref" href="#fn-fri">27</a></sup></dd>
<dt>Feedback loops</dt>
<dd>Unlikely or distant at the moment and there don’t seem to be reliable proxies.</dd>
<dt>Robustness</dt>
<dd>If Robin Hanson is to be believed, then futures are possible in which <span class="caps">AGI</span> can be controlled (and no singleton emerges) even if the alignment problem remains unsolved.</dd>
</td>
</tr>
</table>
<h3 id="wild-animal-suffering-research"><a class="toclink" href="#wild-animal-suffering-research">Wild Animal Suffering Research</a></h3>
<table>
<tr>
<th></th>
<th>Pro</th>
<th>Con</th>
</tr>
<tr>
<th><div>Problem</div></th>
<td>
<dl>
<dt>Scale or upward trajectory of problem</dt>
<dd><span class="caps">WAS</span> is likely to explode upon the colonization of space, and perhaps there are interventions to dampen this exacerbation.</dd>
<dt>Tractability</dt>
<dd>Welfare biology may well be at a similar turning point as medicine in the early 19th century or so.</dd>
<dt>Neglectedness</dt>
<dd>So far, I see only two or three small organizations working on the problem.</dd>
</dl>
</td>
<td>
<dl>
<dt>Upward trajectory questionable</dt>
<dd>Space colonization may happen late enough that a singleton (that will likely not care about spreading biological life in the first place) can meanwhile obviate my reasons to suspect great scale.</dd>
</dl>
</td>
</tr>
<tr>
<th><div>Intervention</div></th>
<td>
<dl>
<dt>Foundational</dt>
<dd><span class="caps">WAS</span> Research does research<sup id="fnref-was-research"><a class="footnote-ref" href="#fn-was-research">28</a></sup> and Animal Ethics encourages it, so they fall within the category of the most tractable type of preparatory work I’ve identified.</dd>
<dt>Counterfactually durable</dt>
<dd>It addresses something that is unlikely to be addressed anyway any time soon. In fact, even their and Animal Ethics’ efforts to get more academics on board are progressing slowly.</dd>
<dt>Factually durable</dt>
<dd>Knowledge, as opposed to values, may survive some existential catastrophes for long enough to be recovered by the new civilization once it’s no longer busy with survival only.</dd>
</dl>
</td>
<td>
<dl>
<dt>Not terribly durable</dt>
<dd>The last pro may be a weak one because even if the knowledge survives catastrophes and societal value drifts, it’ll still take a highly developed civilization to make it likely that it’ll be put to use again.<sup id="fnref-durability-of-knowledge"><a class="footnote-ref" href="#fn-durability-of-knowledge">29</a></sup></dd>
<dt>Invites defection</dt>
<dd>It is tempting to defect against Quiverfull-type maximization of life value systems, as evident in my uncharitable and flawed comparison. Conscientious cooperators can probably avoid it.</dd>
<dt>Robustness varies</dt>
<dd>Though research promises some robustness, it could be greater if it weren’t focussed on something as specific as ecology. Practice in statistics, for example, would be more transferable for the researchers than practice in ecology, and the results, negative or positive, could be informative for a wider range of altruists.</dd>
</dl>
</td>
</tr>
</table>
<h3 id="preserving-effective-altruism"><a class="toclink" href="#preserving-effective-altruism">Preserving Effective Altruism</a></h3>
<p>Movement building so far – probably highly effective but also fairly well funded – has aimed to grow the movement or grow its capacity – one aiming for greater numbers the other aiming for greater influence, better coordination, fewer mistakes, etc. But if effective altruism is destroyed, this would be a bad outcome according to even more value systems than in the case of existential catastrophes, so it would pay to invest heavily into reducing the probability of a <a href="https://www.google.com/url?q=https://www.facebook.com/groups/OMfCT/permalink/1719738125007765/&sa=D&ust=1519548900092000&usg=AFQjCNGwWYIrnlQ5vn2rVfXnjm1nj5gsew">permanent collapse of the movement</a>.</p>
<p>Avenues to achieving this may include actual protection of the people – safety nets have been <a href="http://effective-altruism.com/ea/1k6/the_almighty_hive_will/">discussed</a> and <a href="https://www.facebook.com/groups/peerfunding/">tried</a> – and preservation of knowledge and values. Texts may be sealed into “time capsules,”<sup id="fnref:to-the-moon"><a class="footnote-ref" href="#fn:to-the-moon">30</a></sup> but they need to be the right texts. Instructions that require a high level of technology may be useless for a long time after a catastrophe, and Lukas Gloor also notes that today’s morally persuasive texts were written for a particular audience – us. Read by a different audience in a very different world, they may have different effects from what we hoped for or probably none at all.<sup id="fnref:fanaticism"><a class="footnote-ref" href="#fn:fanaticism">31</a></sup> So such texts may need to be written specifically to be as timeless as possible.<sup id="fnref:text-test"><a class="footnote-ref" href="#fn:text-test">32</a></sup></p>
<table>
<tr>
<th></th>
<th>Pro</th>
<th>Con</th>
</tr>
<tr>
<th><div>Problem</div></th>
<td>
<dl>
<dt>Tractability</dt>
<dd>Tractability of <span class="caps">EA</span> may be higher in a worse world – the world after an existential catastrophe – because there is more to improve and so also more easily improveable things.</dd>
</dl>
</td>
<td>
<dl>
<dt>Tractability</dt>
<dd>Conversely, the tractability of <span class="caps">EA</span> may also be lower in a worse world because many highly effective interventions rely on nonparochial work which depends on technology and infrastructure for communication, travel, and transport.</dd>
</dl>
</td>
</tr>
<tr>
<th><div>Intervention</div></th>
<td>
<dl>
<dt>Durability</dt>
<dd>An intervention – probably research for now – would directly address the question of the durability of <span class="caps">EA</span>. Cases for the effectiveness of facilitating <span class="caps">EA</span> abound, so I won’t reproduce them here. (The raison d’être of 80,000 Hours, Giving What We Can, Centre for Effective Altruism, and others.)</dd>
<dt>Robustness</dt>
<dd>The generality of <span class="caps">EA</span> lends interventions in this space exquisite robustness in terms of fundamentality, instrumental convergence, and option value.</dd>
<dt>Cooperation</dt>
<dd>EAs are probably one of the groups best positioned to understand and adhere to cooperation.</dd>
<dt>Common sense</dt>
<dd>Movement building is considered a high priority, a high prior for estimating the value of movement preservation.</dd>
</dl>
</td>
<td>
<dl>
<dt>Robustness</dt>
<dd>Not really a con, just not a pro, is that the research is probably only as good as the instrumental goal it tries to achieve so that it may be double-counting to list the robustness of research as a separate pro.</dd>
<dt>Feedback loops</dt>
<dd>Small-scale catastrophes may allow for some feedback mechanism, but they’re hopefully few and far in between.</dd>
<dt>Supportability</dt>
<dd>I don’t know of anyone who’s currently working on this.</dd>
</dl>
</td>
</tr>
</table>
<h3 id="other-ea-movement-building"><a class="toclink" href="#other-ea-movement-building">Other <span class="caps">EA</span> Movement Building</a></h3>
<table>
<tr>
<th></th>
<th>Pro</th>
<th>Con</th>
</tr>
<tr>
<th><div>Problem</div></th>
<td>
<dl>
<dt>Tractability</dt>
<dd>Comparatively excellent insofar as we can tell from the decent feedback loops.</dd>
</dl>
</td>
<td></td>
</tr>
<tr>
<th><div>Intervention</div></th>
<td>
<dl>
<dt>Robustness</dt>
<dd>Excellent (see above). Instrumental convergence and fundamentality are ipso facto, and I’m also happy with how well organizations like <span class="caps">CEA</span>, <span class="caps">EAF</span>, and <span class="caps">RC</span> have budgeted option value.</dd>
<dt>Feedback loops, control, option value</dt>
<dd>Decent. (But this overlaps with the first point.)</dd>
<dt>Cooperation</dt>
<dd>Seems to be going better than in at least some other spaces.</dd>
</dl>
</td>
<td>Durability: I don’t know how likely <span class="caps">EA</span> is to re-emerge if it gets lost, that is, how often it would be invented anyway, but otherwise it seems very fragile with regard to both existential catastrophes and mundane value drift.</td>
</tr>
</table>
<h3 id="cultured-meat"><a class="toclink" href="#cultured-meat">Cultured Meat</a></h3>
<table>
<tr>
<th></th>
<th>Pro</th>
<th>Con</th>
</tr>
<tr>
<th><div>Problem</div></th>
<td></td>
<td>
<dl>
<dt>Scale</dt>
<dd>Maybe only relevant so long as humans continue to exist in carbon-based form.</dd>
</dl>
</td>
</tr>
<tr>
<th><div>Intervention</div></th>
<td>
<dl>
<dt>Fairly durable</dt>
<dd>Compared to values spreading approaches to antispeciesism, cultured meat seems more promising on this front, but the cons in the same category may be more interesting.</dd>
<dt>Fundamentality</dt>
<dd>Research, again, promises some fundamentality by its nature.</dd>
<dt>Counterfactual of funding</dt>
<dd>The research can be funded through for-profit startups.</dd>
<dt>Cooperation</dt>
<dd>With most value systems at least. Classic utilitarians who bite the Repugnant Conclusion bullet may mourn the decimation of cows farmed for beef, who are said to have net positive lives.</dd>
</dl>
</td>
<td>
<dl>
<dt>Robustness</dt>
<dd>Research in biochemistry is again rather specific research. (See wild animal suffering research above.)</dd>
<dt>Exposed to many existential risks</dt>
<dd>The very advanced technology necessary to culture meat is likely to remain first inaccessible and then prohibitively expensive for a long time after a civilization-destroying event. During those centuries or millennia, the knowledge of how to produce it may get lost. Or, especially when the research is funded through for-profits, the public perception of cultured meat may suffer like that of genetic engineering even without catastrophic events.</dd>
<dt>Counterfactual durability</dt>
<dd>I’ve heard opinions that cultured meat is the sort of thing that would have to be invented necessarily due to the inefficiency of traditional animal farming.</dd>
</dl>
</td>
</tr>
</table>
<h3 id="moral-circle-expansion"><a class="toclink" href="#moral-circle-expansion">Moral Circle Expansion</a></h3>
<p>I used to call this section “antispeciesist values spreading” but <a href="http://effective-altruism.com/ea/1l0/why_i_prioritize_moral_circle_expansion_over/">Jacy Reese draws the line more widely</a>,<sup id="fnref:jacy-anthis-disclaimer"><a class="footnote-ref" href="#fn:jacy-anthis-disclaimer">33</a></sup> so I’ll go with his reference class and name for it. I have trouble applying my heuristics here because of how meta the intervention is. Jacy’s article should be more enlightening.</p>
<table>
<tr>
<th></th>
<th>Pro</th>
<th>Con</th>
</tr>
<tr>
<th><div>Problem</div></th>
<td>
<dl>
<dt>Scale</dt> <dd>All-encompassing.</dd>
</dl>
</td>
<td>
<dl>
<dt>Scale</dt> <dd>Possibly cut short by singleton.</dd>
</dl>
</td>
</tr>
<tr>
<th><div>Intervention</div></th>
<td>
<dl>
<dt>Robustness</dt>
<dd>Research (as conducted by the Sentience Institute) should be fairly fundamental and widely relevant, the intervention is unlikely to backfire, and it’s so meta that similar considerations apply as for movement building and preservation above.</dd>
</dl>
</td>
<td>
<dl>
<dt>Durability</dt>
<dd>There are probably modes of <span class="caps">MCE</span> that are optimized for durability (maybe humane space colonization could fall into this category) but more generally <span class="caps">MCE</span> should be susceptible to all the same risks that <span class="caps">EA</span> movement building is exposed to.</dd>
</dl>
</td>
</tr>
</table>
<h3 id="humane-space-colonization"><a class="toclink" href="#humane-space-colonization">Humane Space Colonization</a></h3>
<p>This is what I’m most excited about at the moment because it promises to be a very strong contender for most effective long-term intervention – but much more research is needed.</p>
<p>The idea is that SpaceX and others may start shooting people into space in a few decades and may start to put them on Mars too. When that time comes, they’ll be looking for technologies that’ll allow people to survive in these environments permanently. They’ll probably have a range of options, and will choose the one that is most feasible, cheap, or expedient. There may even be a path dependence effect whereby the proven safety of this one technology and the advanced state of its development make it hard for other technologies to attract any attention or funding.</p>
<p>This may not be the technology that would’ve minimized animal suffering, though. So in order to increase the chances that the technology that gets used in the end and that perhaps sees some lock-in is the one that we think is most likely to minimize animal suffering, we need to invest into differential technological development such that at the time that the technology is needed by SpaceX and company, the one that is most feasible, cheap, and expedient coincides with the one that minimizes animal suffering.</p>
<p>A social enterprise that aims to achieve this could be bootstrapped on the basis of vertical agriculture, greenhouse agriculture, and zero-waste housing technologies and then use its know-how and general capacity to research low-suffering technologies for making space habitable.</p>
<table>
<tr>
<th></th>
<th>Pro</th>
<th>Con</th>
</tr>
<tr>
<th><div>Problem</div></th>
<td>
<dl>
<dt>Scale or upward trajectories</dt>
<dd>See Wild Animal Suffering Research.</dd>
<dt>Neglectedness</dt>
<dd>No one is working on it specifically as far as I know. Two or three organizations are working on welfare biology in general.</dd>
</dl>
</td>
<td>
<dl>
<dt>Tractability</dt>
<dd>Unknown.</dd>
</dl>
</td>
</tr>
<tr>
<th><div>Intervention</div></th>
<td>
<dl>
<dt>Counterfactual of funding</dt>
<dd>If it can be run as social enterprise, the funding will have exceptionally low counterfactual value.</dd>
<dt>Durability</dt>
<dd>Highly likely to influence space colonization scenarios in the long term.</dd>
<dt>Robustness</dt>
<dd>Acceptable because greenhouse agriculture may also be a way to cut down on insect suffering in the short term already.</dd>
<dt>Urgency</dt>
<dd>High given how much research is probably necessary, but see con.</dd>
<dt>Cooperative</dt>
<dd>Most people are not, in my experience, astronomical waste–type of consistent classic utilitarians, so most people would probably be neutral toward the project and many may even welcome it without being explicit, agenty suffering reducers.</dd>
<dt>Counterfactually durable</dt>
<dd>See Wild Animal Suffering Research.</dd>
<dt>Feedback loops</dt>
<dd>Maybe, thanks to deserts and the like on earth.</dd>
</dl>
</td>
<td>
<dt>Robustness</dt>
<dd>Possibly but unlikely relevant for singleton scenarios. Also less robust than the most robust research <span class="caps">WASR</span> could do because the corporate research will often have other, overriding priorities.</dd>
<dt>Urgency</dt>
<dd>Plausibly lower than singleton-affecting work if <a href="https://sentience-politics.org/files/Dello-Iacovo-On-terraforming-wild-animal-suffering-and-the-far-future.pdf">Michael Della-Iacovo</a> is right.</dd>
</td>
</tr>
</table>
<h2 id="appendix-changes-since-2015"><a class="toclink" href="#appendix-changes-since-2015">Appendix: Changes Since 2015</a></h2>
<p>This section is only interesting for those who would like to know why I went through the developments that I went through since 2015 – e.g., because you may find yourself in a similar situation as I did back then. Everyone else can feel free to skip this section.</p>
<h3 id="introduction_1"><a class="toclink" href="#introduction_1">Introduction</a></h3>
<p>In 2014, I started out being fan of GiveWell’s and Animal Charity Evaluators’ top charities. (See my <a href="https://claviger.net/my-cause-selection-denis-drescher.html">previous post of this sort</a>.) But these two organizations have the strength and limitation that you don’t need to trust them (or any subjective judgement calls of theirs) because they aim to draw only on hard, objective evidence that can be published in a review. (When they do need to draw on judgment calls, they make them transparent, so that you can fill in your own instead and see how it changes the result.)</p>
<p>The best evidence comes from trying things and seeing what works – from feedback loops. But in some cases feedback loops can be very long; the feedback can be the absence of a catastrophe, which is often difficult to know without knowledge of the counterfactual without the intervention; or the intervention can be so fundamental that it is easily forgotten by the time the feedback comes in. Highly effective interventions can fall into one or several of these categories.<sup id="fnref:ace-long-term"><a class="footnote-ref" href="#fn:ace-long-term">34</a></sup> A search for relatively safe and reliable ways of doing good would have to exclude many of them.</p>
<p>But what I care about terminally is to <em>maximize my expected impact</em> in reducing suffering not to <em>minimize the variance</em> of my estimates of it.</p>
<p>The organizations that provide sufficient evidence (either by implementing a well-proven intervention, like <span class="caps">AMF</span>, or running trials on their own interventions like GiveDirectly) are few and limited to areas that are fairly easy to study.</p>
<p>Michael Dickens argues that the background distribution of cost-effectiveness over interventions is likely to be <a href="http://effective-altruism.com/ea/ws/on_priors/">log-normal or Pareto</a>. The heavier tail of Pareto distributions (or maybe <a href="https://arxiv.org/abs/cond-mat/0412004">power-law distributions in general</a>) makes it seem like the more likely choice to me.</p>
<p>If you have a heuristic that can find some of the most cost-effective interventions (such as the scale, tractability, neglectedness one), but you first draw a sample from the search space, you’ll already lose some of the interventions that the heuristic could’ve recognized.</p>
<p>But, what is worse, the sampling is not random but probably correlated with lower scale and lower negletedness (though perhaps not or in some cases even negatively with tractability).</p>
<h3 id="lower-scale"><a class="toclink" href="#lower-scale">Lower Scale</a></h3>
<h4 id="feedback-cycles"><a class="toclink" href="#feedback-cycles">Feedback Cycles</a></h4>
<p>In order to accrue evidence of effectiveness, interventions need to have short feedback cycles. Studies run for a few years or decades, even natural quasi-experiments rely on cycles in the form that events in a plausible reference class must’ve happened already, and recently enough that there’s data on them and that the study retains some external validity for today.</p>
<p>But the future is long, so the interventions that have the greatest impact are ones that realize this impact throughout as long stretches of the future as possible. In order for such a maximally effective intervention to have strong evidence of its effectiveness, it needs to have short feedback cycles <em>and</em> great positive impact far beyond those cycles. That is something that radically fewer interventions are likely to achieve than just having great positive impact in the long term.</p>
<p>Perhaps short-term effectiveness is highly correlated with long-term effectiveness, but I’m not aware of evidence for that. I’d expect the world to change too much for the training benefits from short feedback loops to be valuable in the long term. On the contrary, high confidence into principles that used to warrant it but don’t anymore is anecdotally hard to overcome. If you know of any good reasons to suspect a correlation, please comment.</p>
<p>Assuming a low correlation, what we’ll get from selecting for feedback loops is likely to be interventions that are highly effective in the long term at most at the rate of the background distribution, so we’ve excluded both clearly ineffective and potentially highly effective interventions. There are many more of the first, if we’re right about the background distribution, so we’ve excluded more of them, but if all the interventions span orders of magnitude in cost-effectiveness differences, then some sort of impact-adjusted rate may make up for that.</p>
<h4 id="final-bricks"><a class="toclink" href="#final-bricks">Final Bricks</a></h4>
<p>What I mentioned above in reference to <a href="https://claviger.net/the-bulk-of-the-impact-iceberg.html">The Bulk of the Impact Iceberg</a> is another important factor that leads to an underinvestment into some types of preparatory work.</p>
<h4 id="short-term-risk-premium"><a class="toclink" href="#short-term-risk-premium">Short-Term Risk Premium</a></h4>
<p>Another consideration that I’m highly unsure about is that additional constraints in for-profit investing usually come at a premium, so adding more constraints (short, recent feedback cycles, well-studied by academia, no conflicts of interest, etc.) may also come at a price in impact.</p>
<h3 id="lower-neglectedness"><a class="toclink" href="#lower-neglectedness">Lower Neglectedness</a></h3>
<h4 id="the-rational-streetlight-effect"><a class="toclink" href="#the-rational-streetlight-effect">The Rational Streetlight Effect</a></h4>
<p>In my post on <a href="https://claviger.net/expected-utility-auctions.html">Expected Utility Auctions</a> I argue that if you have a large set of highly effective interventions of which, say, 1% have strong evidence behind them and get singled out by evaluators for that reason, altruists will all consider this particular 1% but all the remaining 99% will only receive a random allocation of the left-over attention. With so much attention converging on this 1%, it’ll become greatly less neglected than the remaining 99%.</p>
<p>This may be offset by larger funding gaps – i.e. more slowly dropping marginal utility for money – but I don’t know of evidence that this is the case, and where it is the case, it may be due to the very lack of attention I want to address, e.g., because it’s harder to scale for a charity implementing such an intervention because of the lack of attention from potential hires and thus greater difficulty in scaling up. Moreover, if the marginal cost-effectiveness of one intervention drops more slowly than that of another intervention but the second intervention is much more cost-effective on average, it’ll take a while to drop down to the level of the first intervention.</p>
<p>Moreover, individual charities implementing interventions need to have funding gaps sufficient large to warrant the work that goes into evaluating them even this is possible. That also excludes some small organizations, though evaluators are now working on pipelines to enable such organizations to bootstrap their way up.</p>
<h4 id="cost-of-information"><a class="toclink" href="#cost-of-information">Cost of Information</a></h4>
<p>The higher-quality evidence comes at a monetary price, too, so interventions where the millions or so for the creation of such evidence have already been paid should tend to be fairly well-funded already.</p>
<h4 id="suitability-for-more-types-of-donors"><a class="toclink" href="#suitability-for-more-types-of-donors">Suitability for More Types of Donors</a></h4>
<p>Such evidence also enables grantors that are accountable to many stakeholders to make safe investments, not only GiveWell. Any donor who is not exposed to accountability pressures should use their comparative advantage to support the interventions that governments and some foundations would not be able to support.</p>
<h2 id="acknowledgements"><a class="toclink" href="#acknowledgements">Acknowledgements</a></h2>
<p>This piece benefitted from comments and support by Anami Nguyen, Lukas Gloor, Martin Janser, Naoki Peter, and Michal Pokorný. Thank you.</p>
<h2 id="footnotes"><a class="toclink" href="#footnotes">Footnotes</a></h2>
<div class="footnote">
<hr>
<ol>
<li id="fn:heuristics-collection">
<p>But I’ve tried to compensate for that by compiling heuristics of others and searching past articles of mine for heuristics I thought relevant at the time. <a class="footnote-backref" href="#fnref:heuristics-collection" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
<li id="fn:antirealism">
<p>This is based on my now <a href="https://plato.stanford.edu/entries/moral-anti-realism/">antirealist</a> perspective, which is some sort of strategically constrained emotivism. <a href="https://foundational-research.org/gains-from-trade-through-compromise/">Gains from Trade through Compromise</a>, Moral Tribes, and The Righteous Mind are good introductions to what I’m referring to by “strategically constrained” (the article) and antirealism (the books) respectively. My moral perspective is the result of a <a href="http://www.overcomingbias.com/2009/01/moral-uncertainty-towards-a-solution.html">moral parliament</a> (note that the article appears to be written from a realist perspective) that consists mostly of a type of two-level <a href="https://foundational-research.org/hedonistic-vs-preference-utilitarianism/">preference utilitarianism</a> with some negative and classic utilitarian minorities. The result looks akin to a <a href="http://effective-altruism.com/ea/140/integrity_for_consequentialists/">deontological decision procedure</a> and something like lexical threshold prioritarianism. My intuition is also that suffering <a href="http://www.nickbostrom.com/papers/experience.pdf">scales</a> <a href="https://foundational-research.org/how-the-simulation-argument-dampens-future-fanaticism#Many_copies_of_a_brain_dont_matter_much_more_than_one_copy">linearly</a> with the number of beings suffering to the same degree. <a class="footnote-backref" href="#fnref:antirealism" title="Jump back to footnote 2 in the text">↩</a></p>
</li>
<li id="fn:clustering">
<p>When talking about the long term, I either mean millennia – periods long enough that we can suffer existential catastrophes or start colonizing space – or all of the future. Humans, as the currently most powerful species, are still compressed in one spot in the universe, so that I think that what we do now will matter more (might even matter greatly more if our influences can last long enough) for the rest of the future than what any similarly sized cluster of people will do once they have colonized enough of the surrounding universe that communication between the clusters becomes difficult. <a class="footnote-backref" href="#fnref:clustering" title="Jump back to footnote 3 in the text">↩</a></p>
</li>
<li id="fn:increased-marginal-utility">
<p>Just as problems will typically become harder to solve to the same degree the closer they are to being solved absolutely, so they will typically become easier to solve to the same degree the farther they are away from being solved absolutely. <a class="footnote-backref" href="#fnref:increased-marginal-utility" title="Jump back to footnote 4 in the text">↩</a></p>
</li>
<li id="fn:slight-downward-trajectory">
<p>Or plausibly one with a downward trajectory that is still so far away from being solved that I don’t expect any significant drops in marginal utility throughout a period that I can hope to affect. (More on that in the "Model" section below.) But that judgment will be even more error prone than that of a trajectory alone. <a class="footnote-backref" href="#fnref:slight-downward-trajectory" title="Jump back to footnote 5 in the text">↩</a></p>
</li>
<li id="fn:s-risks">
<p>Anything from incremental but extreme increases in suffering to sudden explosions with afterwards constant extreme suffering are considered s-risks. <a class="footnote-backref" href="#fnref:s-risks" title="Jump back to footnote 6 in the text">↩</a></p>
</li>
<li id="fn:cooperatively-waiting-for-research">
<p>If the EA-funded institute waits for (from our perspective) interesting research to get done elsewhere, it can invest the time into interesting research that would not otherwise get done. But if the EA-funded institute does the interesting research first the non-EA researchers will probably research something uninteresting (again, from our perspective) in that time, so less interesting research will get done. This may at first seem uncooperative, and it’ll probably pay to stay alert to cases where it is, but it is the observed preference of the non-EA researchers to research the interesting topic in question, so they pay no price for our gain. <a class="footnote-backref" href="#fnref:cooperatively-waiting-for-research" title="Jump back to footnote 7 in the text">↩</a></p>
</li>
<li id="fn:long-delays">
<p><a href="https://foundational-research.org/cause-prioritization-downside-focused-value-systems/">Lukas Gloor</a>: “Any effects that near-extinction catastrophes have on delaying space colonization are largely negligible in the long run when compared to affecting the quality of a future with space colonization – at least unless the delay becomes very long indeed (e.g. millions of years or longer).” <a class="footnote-backref" href="#fnref:long-delays" title="Jump back to footnote 8 in the text">↩</a></p>
</li>
<li id="fn:models-of-galactic-spreading">
<p>I’m still very much looking for models of intergalactic spreading that take into account lightspeed limits and cosmic expansion; trade-offs between expansion and more efficient usage of close space and resources (e.g., through a switch to silicon-based life); and trade (including moral trade) with other space-colonizing species. <a class="footnote-backref" href="#fnref:models-of-galactic-spreading" title="Jump back to footnote 9 in the text">↩</a></p>
</li>
<li id="fn:recovery-after-catastrophe">
<p>From “<a href="http://effective-altruism.com/ea/1k4/draft_cause_prioritization_for_downsidefocused/">Cause Prioritization for Downside-Focused Value Systems</a>”: “This leaves us with the question of how likely a global catastrophe is to merely delay space colonization rather than preventing it. I have not thought about this in much detail, but after having talked to some people (especially at FHI) who have investigated it, I updated that rebuilding after a catastrophe seems quite likely. And while a civilizational collapse would set a precedent and reason to worry the second time around when civilization reaches technological maturity again, it would take an unlikely constellation of collapse factors to get stuck in a loop of recurrent collapse, rather than at some point escaping the setbacks and reaching a stable plateau (Bostrom, 2009), e.g. through space colonization. I would therefore say that large-scale catastrophes related to biorisk or nuclear war are quite likely (~80–90%) to merely delay space colonization in expectation.[17] (With more uncertainty being not on the likelihood of recovery, but on whether some outlier-type catastrophes might directly lead to extinction.)” See also footnote 17 in the original. <a class="footnote-backref" href="#fnref:recovery-after-catastrophe" title="Jump back to footnote 10 in the text">↩</a></p>
</li>
<li id="fn:ai-and-communication">
<p>Unless perhaps they need great amounts of resources, such as energy, or some other additional limiting factor. <a class="footnote-backref" href="#fnref:ai-and-communication" title="Jump back to footnote 11 in the text">↩</a></p>
</li>
<li id="fn:cubic-spreading">
<p>Space has three spatial dimensions and is, at the largest scale, filled with matter evenly. When our civilization expands into space, we’ll always try to reach the resources that are closest and not yet claimed by others, so that there’s no reason why we would expand only in one direction (at a sufficiently large scale). At the largest scale, we’d get a volume of our civilization and thus suffering of ¾·π·r³, but first there are many irregularities that will lead to a much more complex (but still cubic) polynomial – irregularities such as cultural differences between mostly isolated clusters of our civilization, perhaps evolutionary differences, irregularities in the density of solar systems and galaxies, etc. <a class="footnote-backref" href="#fnref:cubic-spreading" title="Jump back to footnote 12 in the text">↩</a></p>
</li>
<li id="fn:faster-than-light">
<p>Faster-than-light communication or even travel would change my models significantly. <a class="footnote-backref" href="#fnref:faster-than-light" title="Jump back to footnote 13 in the text">↩</a></p>
</li>
<li id="fn:eh">
<p><code>¯\_(ツ)_/¯</code> <a class="footnote-backref" href="#fnref:eh" title="Jump back to footnote 14 in the text">↩</a></p>
</li>
<li id="fn:audience">
<p>I would guess that this article is mostly read by an audience of private donors and perhaps someone at the Open Philanthropy Project or another similar foundation, so I can gear it a little bit toward that audience. <a class="footnote-backref" href="#fnref:audience" title="Jump back to footnote 15 in the text">↩</a></p>
</li>
<li id="fn:millennia">
<p>A millennium is long enough that it seems unintuitive to me that a little delay at the beginning should have an effect at the end of it but it is short enough that the accessibility of marginal resources (solar systems and such) is not systematically lower at the end due to such effects as the expansion of space or increase in entropy. <a class="footnote-backref" href="#fnref:millennia" title="Jump back to footnote 16 in the text">↩</a></p>
</li>
<li id="fn:etg-influence">
<p>For example, ETGing for highly effective short-term-focused charities for 50 years of one’s life. <a class="footnote-backref" href="#fnref:etg-influence" title="Jump back to footnote 17 in the text">↩</a></p>
</li>
<li id="fn:concepts-and-ideas">
<p>I’ll refer to these as concepts or ideas (whichever is clearer in the context) or influences when I only mean the difference between their factual and counterfactual duration of existence. <a class="footnote-backref" href="#fnref:concepts-and-ideas" title="Jump back to footnote 18 in the text">↩</a></p>
</li>
<li id="fn:uspto">
<p>Databases of the application for and expiry of trademarks (such as the USPTO) and of companies themselves may provide enough data to try to fit a decay function to it. <a class="footnote-backref" href="#fnref:uspto" title="Jump back to footnote 19 in the text">↩</a></p>
</li>
<li id="fn:companies">
<p>In comparison to languages for example, a company either exists or doesn’t, but Proto-Indo-European, albeit incomprehensible to us, still exerts some influence on languages today. <a class="footnote-backref" href="#fnref:companies" title="Jump back to footnote 20 in the text">↩</a></p>
</li>
<li id="fn:counterfactual-homonymy">
<p>I hope that sentence is comprehensible even if I got the grammar wrong. <a class="footnote-backref" href="#fnref:counterfactual-homonymy" title="Jump back to footnote 21 in the text">↩</a></p>
</li>
<li id="fn:most">
<p>At least in the sense of “more than average.” <a class="footnote-backref" href="#fnref:most" title="Jump back to footnote 22 in the text">↩</a></p>
</li>
<li id="fn:independent-language-development">
<p>Is this true? Or do similar people put in similar regions tend to independently evolve similar languages due to all the similarities in their requirements, circumstances, etc.? Eyes have evolved independently a couple of times after all. <a class="footnote-backref" href="#fnref:independent-language-development" title="Jump back to footnote 23 in the text">↩</a></p>
</li>
<li id="fn:city-places">
<p>I don’t know if this is a minority or the majority of cities. <a class="footnote-backref" href="#fnref:city-places" title="Jump back to footnote 24 in the text">↩</a></p>
</li>
<li id="fn:time-dilation">
<p>The time dilation is about 1.6 seconds at that speed over 1,000 years, so no help there. <a class="footnote-backref" href="#fnref:time-dilation" title="Jump back to footnote 25 in the text">↩</a></p>
</li>
<li id="fn:ems">
<p>By this I mean to cover the possibilities of physical travel in a sleeping state and travel of <a href="https://en.wikipedia.org/wiki/The_Age_of_Em">ems</a> (as Robin Hanson calls them) stored on hard drives of sorts. <a class="footnote-backref" href="#fnref:ems" title="Jump back to footnote 26 in the text">↩</a></p>
</li>
<li id="fn:fri">
<p>My favorite is FRI, but I’m still waiting for it to emerge as a suffering-focused AI alignment powerhouse. (My heuristic here is whether an organization seems to be closer to the category of looking for something to research and when it finds it pounce on it to make the best of it or the category of having a wealth of topics to research and spending considerable time not on searching for but on prioritizing between topics.) It’ll probably take a while longer to build up such capacity. MIRI seems to be popular at FRI, but it is also (finally) well-funded this year. Are there any other suffering-focused organizations to consider? <a class="footnote-backref" href="#fnref:fri" title="Jump back to footnote 27 in the text">↩</a></p>
</li>
<li id="fn:was-research">
<p>The name may’ve given it away already. <a class="footnote-backref" href="#fnref:was-research" title="Jump back to footnote 28 in the text">↩</a></p>
</li>
<li id="fn:durability-of-knowledge">
<p>Knowledge of how to reduce suffering at no gain to the suffering reducers themselves is probably not more durable than persuasive texts on ethics (Lukas Gloor gave Animal Liberation as an example). If we (people like me) had to still fight harder for our survival, we may not have had the resources (time, leisure, iodine, etc.) to become more altruistic in the first place. Just as our civilization gives us those resources, it also makes altruism cheaper through economies of scale, gains from trade, transport, communication, etc. If this civilization is destroyed and some people are just struggling to rebuild it, then (1) knowledge whose implementation is feasible today will be infeasible for a long time, throughout which it may get lost, and (2) those who can’t themselves fight or bargain to be included, such as nonhuman animals, will (again) be left behind for the longest time. Neither approach, scientific knowledge or moral arguments, are likely to be enough. However, a technology that will continue to be feasible, will profit its operator as much or more than an alternative, and benefits the animals would be the most likely technology to survive. <a class="footnote-backref" href="#fnref:durability-of-knowledge" title="Jump back to footnote 29 in the text">↩</a></p>
</li>
<li id="fn:to-the-moon">
<p>See the paper <a href="http://effective-altruism.com/ea/1lf/paper_surviving_global_risks_through_the/">Surviving global risks through the preservation of humanity's data on the Moon</a> by Turchin and Denkenberger. <a class="footnote-backref" href="#fnref:to-the-moon" title="Jump back to footnote 30 in the text">↩</a></p>
</li>
<li id="fn:fanaticism">
<p>He’s particularly worried about very bad outcomes from fanaticism for example about EA ideas when they are read at a time when people are less aware of biases or more prone to superstitious thinking. But if the probability of fanaticism is small compared to a failure of the texts to have any effect (let’s call it a null result) but the outcome may be bad enough to offset that, then, without act-omission distinction, fanaticism doesn’t seem clearly worse than simple failure by a factor of more than 1–2 or so, while the probability of a null result seems orders of magnitude more likely to me. <a class="footnote-backref" href="#fnref:fanaticism" title="Jump back to footnote 31 in the text">↩</a></p>
</li>
<li id="fn:text-test">
<p>A test of that might be to ask people from many different backgrounds and cultures to summarize such texts, preferably people who don’t know enough of the culture of the author to adjust for the difference. If the results are similar, they’re more likely to be suitable texts. <a class="footnote-backref" href="#fnref:text-test" title="Jump back to footnote 32 in the text">↩</a></p>
</li>
<li id="fn:jacy-anthis-disclaimer">
<p>Disclaimer: I don’t condone the author’s conduct. (I have no first-hand information, but it sounds to me like verbal sexual harassment.) Jacy authored many informative articles and a book and runs an organization with a promising strategic vision, so that I can’t help but cite him lest it seem that I plagiarize them. But I’m adding these disclaimers to avoid the impression that I accept such conduct or that it is accepted in my circles. <a class="footnote-backref" href="#fnref:jacy-anthis-disclaimer" title="Jump back to footnote 33 in the text">↩</a></p>
</li>
<li id="fn:ace-long-term">
<p>The criteria of Animal Charity Evaluators should allow them to consider the long term, but that is not what I’ve observed in practice. <a class="footnote-backref" href="#fnref:ace-long-term" title="Jump back to footnote 34 in the text">↩</a></p>
</li>
</ol>
</div>Cause Area: Human Rights in North Korea2017-11-20T00:00:00+00:002020-10-09T00:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2017-11-20:/cause-area-human-rights-in-north-korea.html<p>The suffering that the North Korean regime inflicts on its citizens is a lesser source of suffering than malaria worldwide (but not compared to individual highly malarial countries of similar population as North Korea) or industrial agriculture in <span class="caps">US</span> states of similar population. However, it may be on par or even exceed that inflicted on the <span class="caps">US</span> American prison population, a cause prioritized by the Open Philanthropy Project. There are risky but promising interventions, which could be scaled up if more funding were available. The cause area seems well suited for hits-based giving by major donors looking for funding gaps. The government change in South Korea of May 9, 2017, may further increase the marginal utility of funding.</p><div id="pec-encrypted-content" style="display:none">csUwxl36L8/bAVPg1cVnTA==;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;^</div>
<div id="pec-decrypted-content">
<h4><i></i></h4>
</div>
<form id="pec-decrypt-form">
<p>
Old articles can be embarrassing. If you would like to access them anyway, you may
<a href="https://bit.ly/3jPN8tT" target="_blank">request access here</a>. Please indicate
who you are in case I don’t know you or don’t recognize you from your email address.
</p>
<input type="password" id="pec-content-password" placeholder="Password" />
<button type="submit" id="pec-decrypt-content">Decrypt</button>
</form>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/core.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/enc-base64.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/cipher-core.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/pad-nopadding.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/md5.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/aes.js"></script>
<script type="text/javascript">
(function () {
var strip_padding = function (padded_content, padding_char) {
/*
* Strips the padding character from decrypted content.
*/
for (var i = padded_content.length; i > 0; i--) {
if (padded_content[i - 1] !== padding_char) {
return padded_content.slice(0, i);
}
}
};
var decrypt_content = function (password, iv_b64, ciphertext_b64, padding_char) {
/*
* Decrypts the content from the ciphertext bundle.
*/
var key = CryptoJS.MD5(password),
iv = CryptoJS.enc.Base64.parse(iv_b64),
ciphertext = CryptoJS.enc.Base64.parse(ciphertext_b64),
bundle = {
key: key,
iv: iv,
ciphertext: ciphertext
};
var plaintext = CryptoJS.AES.decrypt(bundle, key, { iv: iv, padding: CryptoJS.pad.NoPadding });
try {
return strip_padding(plaintext.toString(CryptoJS.enc.Utf8), padding_char);
}
catch (err) {
// encoding failed; wrong password
console.log(err);
return false;
}
};
var init_decryptor = function () {
var decrypt_btn = document.getElementById('pec-decrypt-content'),
password_input = document.getElementById('pec-content-password'),
encrypted_content = document.getElementById('pec-encrypted-content'),
decrypted_content = document.getElementById('pec-decrypted-content'),
decrypt_form = document.getElementById('pec-decrypt-form');
decrypt_btn.addEventListener('click', function () {
// grab the ciphertext bundle
var parts = encrypted_content.innerHTML.split(';');
// decrypt it
var content = decrypt_content(
password_input.value,
parts[0],
parts[1],
parts[2]
);
if (content) {
// success; display the decrypted content
decrypted_content.innerHTML = content;
decrypt_form.parentNode.removeChild(decrypt_form);
encrypted_content.parentNode.removeChild(encrypted_content);
// any post processing on the decrypted content should be done here
}
else {
// ¯\_(ツ)_/¯
password_input.value = '';
}
});
};
document.addEventListener('DOMContentLoaded', init_decryptor);
})();
</script>Friends of the Civilization2017-09-06T18:00:00+00:002017-09-06T18:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2017-09-06:/friends-of-the-civilization.html<p>We are the Friends of the Civilization, and we call him Elua. Elua shelters us.</p><div id="pec-encrypted-content" style="display:none">flv72mfr0gdmbc7hG6+FoA==;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;^</div>
<div id="pec-decrypted-content">
<h4><i></i></h4>
</div>
<form id="pec-decrypt-form">
<p>
Old articles can be embarrassing. If you would like to access them anyway, you may
<a href="https://bit.ly/3jPN8tT" target="_blank">request access here</a>. Please indicate
who you are in case I don’t know you or don’t recognize you from your email address.
</p>
<input type="password" id="pec-content-password" placeholder="Password" />
<button type="submit" id="pec-decrypt-content">Decrypt</button>
</form>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/core.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/enc-base64.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/cipher-core.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/pad-nopadding.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/md5.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/aes.js"></script>
<script type="text/javascript">
(function () {
var strip_padding = function (padded_content, padding_char) {
/*
* Strips the padding character from decrypted content.
*/
for (var i = padded_content.length; i > 0; i--) {
if (padded_content[i - 1] !== padding_char) {
return padded_content.slice(0, i);
}
}
};
var decrypt_content = function (password, iv_b64, ciphertext_b64, padding_char) {
/*
* Decrypts the content from the ciphertext bundle.
*/
var key = CryptoJS.MD5(password),
iv = CryptoJS.enc.Base64.parse(iv_b64),
ciphertext = CryptoJS.enc.Base64.parse(ciphertext_b64),
bundle = {
key: key,
iv: iv,
ciphertext: ciphertext
};
var plaintext = CryptoJS.AES.decrypt(bundle, key, { iv: iv, padding: CryptoJS.pad.NoPadding });
try {
return strip_padding(plaintext.toString(CryptoJS.enc.Utf8), padding_char);
}
catch (err) {
// encoding failed; wrong password
console.log(err);
return false;
}
};
var init_decryptor = function () {
var decrypt_btn = document.getElementById('pec-decrypt-content'),
password_input = document.getElementById('pec-content-password'),
encrypted_content = document.getElementById('pec-encrypted-content'),
decrypted_content = document.getElementById('pec-decrypted-content'),
decrypt_form = document.getElementById('pec-decrypt-form');
decrypt_btn.addEventListener('click', function () {
// grab the ciphertext bundle
var parts = encrypted_content.innerHTML.split(';');
// decrypt it
var content = decrypt_content(
password_input.value,
parts[0],
parts[1],
parts[2]
);
if (content) {
// success; display the decrypted content
decrypted_content.innerHTML = content;
decrypt_form.parentNode.removeChild(decrypt_form);
encrypted_content.parentNode.removeChild(encrypted_content);
// any post processing on the decrypted content should be done here
}
else {
// ¯\_(ツ)_/¯
password_input.value = '';
}
});
};
document.addEventListener('DOMContentLoaded', init_decryptor);
})();
</script>The Bulk of the Impact Iceberg2017-09-04T18:00:00+00:002017-09-04T18:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2017-09-04:/the-bulk-of-the-impact-iceberg.html<p>A year or two ago, I first noticed that the way I thought about impact, who causes it, and what replaceability meant did not quite make sense. These concerns lead first to my article “<a href="https://claviger.net/attribution-moloch.html">The Attribution Moloch</a>” and now to this one, an addendum of sorts. Here, I will introduce several considerations that should lead us to value preparatory work – in particular research – higher or even higher than we already do.</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#the-iceberg">The Iceberg</a></li>
<li><a href="#research">Research</a></li>
<li><a href="#additive-delays">Additive Delays</a></li>
<li><a href="#the-shapley-value">The Shapley Value</a></li>
</ul>
</div>
<h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<p><img src="/images/the-bulk-of-the-impact-iceberg/gwwc-donation-totals-2014.png" alt="Giving What We Can donation totals as of September 2014" style="width: 49%;"/>
<img src="/images/the-bulk-of-the-impact-iceberg/gwwc-donation-totals-2017.png" alt="Giving What We Can donation totals as of September 2017" style="width: 49%;"/></p>
<p>I think this change between September 2014 and today is a good thing. We seem to be investing more strongly into preparatory work as opposed to only placing the final bricks in almost complete impact edifices. Since these are, I think, totals over all time, the current allocation (say, this year) must be even further slanted toward organizations like <span class="caps">CEA</span>, 80,000 Hours, and hopefully <a href="https://was-research.org/">Wild Animal Suffering Research</a> than is apparent from the total.</p>
<p>The reason I’m happy to see it is that it shows that we may be solving one or more of the challenges of: personal risk aversion, coordination, and understanding who generates the bulk of the impact. I’ll cover the first two only briefly, because it seems to me that they’re known, and mostly focus on the last.</p>
<ol>
<li>
<p>Decreasing marginal utility of money for generating happiness is a good reason to be somewhat risk averse personally, but while we can pretty clearly notice this decrease throughout the area of $20,000 to $80,000 or so, it only becomes noticeable only <a href="http://reducing-suffering.org/when-should-altruists-be-financially-risk-averse/">in much higher areas for altruistic investments</a> – perhaps three orders of magnitude higher. So when we transfer our intuitions about risk aversity from personal finance to donations, we’ll likely lose out on expected value at no gain.</p>
<ol>
<li>A different perspective on this is that when we strive for an agent-relative goal such as personal wealth maximization, we also personally bear all the risk. But when we strive for an agent-neutral goal such as maximizing happiness or minimizing suffering, the successes and failures are shared to the degree that the actors coordinate their attempts: If there are 1000 plausible ideas of which a random one will succeed, and if 100 EAs coordinate to all try out different ones, 10 trials per <span class="caps">EA</span>, then the success is not the success of the one who happens to get lucky but the success of all of them collectively. It was a perfectly riskless endeavor for them.</li>
</ol>
</li>
<li>
<p>GiveWell currently recommends seven charities, of which three are really well known. But Open Phil has made grants to almost 180 charities; and add to that a bunch of <span class="caps">EA</span> charities they haven’t made grants to. These are probably charities who can plausibly outdo GiveDirectly in direct impact. But when people talk about deciding whether to invest into low risk–low reward or high risk–high reward interventions, it sounds as if they’re deciding between some roughly equal options or even with a slant toward low risk and low reward. The number of charities – though not necessarily the total funding gap just yet – indicates more of a 1:20 split. In effect, the probably less than 50% of donations that go to high risk–high reward interventions seem to vanish compared to the donations that focus on GiveWell top charities. This argument alone is not sufficient to show that this is bad, but it shows that if it is bad (and I think so for other reasons), then we are (or were) not so much falling for the streetlight effect but for a coordination problem.</p>
</li>
<li>
<p>Finally, I think that a lot of people – though unfortunately probably few of the people I can reach with this article – are likely to undervalue preparatory work compared to final-brick type of work. This is what I will explain further in the following.</p>
</li>
</ol>
<h2 id="the-iceberg"><a class="toclink" href="#the-iceberg">The Iceberg</a></h2>
<p>When a friend of mine founded a charity, she did year-long research, got pro-bono counseling from specialist friends of hers, got seven people together to be legally able to found the organization, and got one of them to provide the office space for the meeting. Then a lawyer, who charged a particularly friendly rate, made the final strokes to the by-laws, so we could all sign them. (I was one of those seven people.)</p>
<p>This might be an analogy for the work of the Against Malaria Foundation. Since <span class="caps">AMF</span> steps in and subsequently people don’t die, it might seem like <span class="caps">AMF</span> is to be credited with saving those people’s lives. Especially when it’s actually the case that they would’ve died without <span class="caps">AMF</span> and didn’t die with <span class="caps">AMF</span>, it would seem that way. But while <span class="caps">AMF</span> certainly made the difference between them dying or not dying, everyone else, without whom these people would’ve died too, also needs to be credited: distribution partners, net producers, donors, organizations like GiveWell who funnel donations to <span class="caps">AMF</span>, researchers who found bednets to be a good idea, everyone who prevented the third world war, everyone’s parents, et al.</p>
<p>In the above analogy, the distribution partners or the net producers may be like the lawyer; <span class="caps">AMF</span>, perhaps, is like the cofounder who provided the room and one of the signatures; my friend is like GiveWell and the Disease Control Priorities (<span class="caps">DCP2</span>) study together; and the friends she interviewed may’ve been like the 20-odd teams that conducted RCTs on malaria prevention with bednets.</p>
<p>The particular impact of the founding of her charity or conducing a net distribution is thus a joint achievement of all the above people and many more. So there are cases, perhaps most of them, where a lot of effort had to first be invested before one charity could plug some final hole or place some final brick and make all the impact come to fruition. The bulk of the iceberg probably lies below the GiveWell-recommended tip.</p>
<h2 id="research"><a class="toclink" href="#research">Research</a></h2>
<p><img alt="Dependency tree of contributions" src="/images/the-bulk-of-the-impact-iceberg/impact-tree.png"></p>
<p>Much of the bulk of the iceberg is research, which has the interesting property that often negative results – if they are the result of a high-quality, sufficiently powered study – can be useful. If the 100 EAs from the introduction (under 1.a.) are researchers, they know that one of the plausible ideas got to be right, and 99 of them have already been shown not to be useful, then the final <span class="caps">EA</span> researcher can already eliminate 99% of the work with very little effort by relying on what the others have already done. The bulk of that impact iceberg was thanks to the other researchers. Insofar as research is a component of the iceberg, it’s a particularly strong investment.</p>
<h2 id="additive-delays"><a class="toclink" href="#additive-delays">Additive Delays</a></h2>
<p>There’s another reason for why the bulk of the iceberg is likely big. Let’s conceive of impact as a reduction of the integral over time of something we care to eliminate – say, suffering. An intervention like bednets will slightly reduce the slope of the curve (perhaps for a limited time) thereby reducing the aggregate suffering over all time. The earlier this reduction happens, the better.</p>
<p><img src="/images/the-bulk-of-the-impact-iceberg/impact-speed-up.png" alt="Impact speed-up" style="display: block; margin: 0 auto;" width="628" height="388" /></p>
<p>Each step closer to the impact is dependent upon one or more previous steps, just like GiveWell’s contributions were dependent on studies, electronic money transfer, the invention of LLINs, Sci-Hub, etc. A delay of one of them – say, LLINs haven’t been invented – means that all dependent steps closer to the impact and all steps that, in turn, depend on them, are delayed as well. And among all the dependencies on one “level,” the one that has the longest delay is the one that determines the delay. The same happens at all further levels in the dependency tree, and all these delays add up.</p>
<p>In practice, I would not expect clean levels like that to emerge, but insofar as there are inventions or discoveries that are highly necessary for some impact to emerge, one strategy that would follow from this perspective would be to try to speed up the development of the necessary condition that one expects will take longest to come about.</p>
<p>A counterargument might be that in some environments, something quick and simple can serve as a proof of concept to attract investment or revenue that can then enable the development of the more difficult innovation.</p>
<h2 id="the-shapley-value"><a class="toclink" href="#the-shapley-value">The Shapley Value</a></h2>
<p>There’s really no one true way of assigning partitions of this positive impact to the different contributors, but economists have come up with some <a href="https://en.wikipedia.org/wiki/Shapley_value#Properties">sane-sounding axioms</a> (informally summarized below) to describe one conception of justice and a method that uniquely assigns partitions accordingly. That is the <a href="https://en.wikipedia.org/wiki/Shapley_value">Shapley value</a>. (Another method of allocation with different goals and results is <a href="https://en.wikipedia.org/wiki/Core_(game_theory)">the core</a>.)</p>
<ol>
<li>
<p>Efficiency: it allocates the whole impact.</p>
</li>
<li>
<p>Symmetry: equal contributors receive equal shares.</p>
</li>
<li>
<p>Linearity: if you consider two cooperative efforts as one, each contributor’s share in the joint case will be the sum of the contributor’s shares in the split case.</p>
</li>
<li>
<p>Zero/dummy/null player: someone who contributes nothing, get none of the impact assigned.</p>
</li>
</ol>
<p>Note that this is wholly unnecessary for prioritization. We can choose our actions so to maximize the total impact without applying any algorithm that splits it up again and assigns it to individual contributors. I have, in the past, argued against even attempting such an allocation because all algorithms I had been aware of had unacceptable implications. I probably still do, because the Shapley value is probably hard to calculate in most situations effective altruists will find themselves in. (Not so, however, in <a href="https://support.google.com/adwords/answer/6394265?hl=en">advertisement</a>.)</p>
<p>When we’re talking about replaceability in <span class="caps">EA</span>, we’re usually taking some action and some likely alternative involving inaction and comparing them, e.g., comparing the world with a world without <span class="caps">AMF</span> but otherwise identical. (<a href="https://80000hours.org/2016/02/the-value-of-coordination/">80,000 Hours has touched on it here.</a>) When we’re talking about comparative advantage, however, we’re considering a number of counterfactuals and a number of different actions and reactions to find the maximally impactful variant. I don’t know for sure, but it seems to be like making full use of one’s comparative advantage should be equivalent to maximizing one’s Shapley value.</p>
<p>A consideration of replaceability (defined in this limited way) alone, however, does not have the same properties. Say, we want to determine the contribution of C in a cooperation of A, B, and C – denoted as the set {A, B, C}. Replaceability would only consider the case where {A, B} cooperate and the case where {A, B, C} all cooperate and compare them. That may not be sufficient. Rather, to determine the Shapley value of C, we need to consider all eight subsets of {A, B, C} and calculate the mean of C’s contributions in all cases that involve C.</p>
<p>Another difficulty with actually calculating a Shapley value is that in practice cooperation usually happens between groups of very large cardinality (just think of all the contributors in the graph above) where each contributor contributes some at least infinitesimal part. There is a reformulation of the Shapley value that captures this, but I don’t know how helpful it’ll be even just for illustration purposes.</p>
<p>The main take-away for me is to make sure that I sufficiently value also nonobvious preparatory work that has made some impact possible and to value preparatory work that is currently being done that may make the difference between impact materializing earlier, only much later, or never at all.</p>Direct Suffering Caused by Various Animal Foods2017-06-04T00:00:00+00:002020-11-09T00:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2017-06-04:/direct-suffering-caused-by-various-animal-foods.html<p>I converted Brian Tomasik’s <a href="http://reducing-suffering.org/how-much-direct-suffering-is-caused-by-various-animal-foods/">How Much Direct Suffering Is Caused by Various Animal Foods?</a> to <a href="https://www.getguesstimate.com/models/8960">Guesstimate</a>. We now have ranges, distributions, and the sensitivity analysis to draw on to refine the estimates. I also added two columns to determine the suffering of the average per capita consumption, which seems to me like the more intuitive figure; refined the estimates with additional research; and added organic eggs for comparison.</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#brians-reservations">Brian’s Reservations</a></li>
<li><a href="#faunalytics-research">Faunalytics Research</a></li>
<li><a href="#assorted-considerations">Assorted Considerations</a><ul>
<li><a href="#leghorn-lifespans">Leghorn Lifespans</a></li>
<li><a href="#cow-lifespans">Cow Lifespans</a></li>
<li><a href="#sentience-multipliers">Sentience Multipliers</a></li>
<li><a href="#suffering-multipliers">Suffering Multipliers</a></li>
</ul>
</li>
<li><a href="#footnotes">Footnotes</a></li>
</ul>
</div>
<h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<p><a href="https://www.getguesstimate.com/models/8960"><img alt="Direct suffering Guesstimate." src="/images/direct-suffering-caused-by-various-animal-foods/direct-suffering-guesstimate.png"></a></p>
<p>Brian Tomasik’s classic “<a href="http://reducing-suffering.org/how-much-direct-suffering-is-caused-by-various-animal-foods/">How Much Direct Suffering Is Caused by Various Animal Foods?</a>” investigates the most easily observable suffering footprint of various products derived from animal products.</p>
<p>The page, in it’s latest incarnation, contains a JavaScript calculator for putting in one’s own values. But I’ve gotten used to using ranges and having a sensitivity analysis to draw on to decide which cells to research further. So <a href="https://www.getguesstimate.com/models/8960">I transferred Brian’s model to Guesstimate</a>, mostly drawing on his research but also adding some of my own when I couldn’t decide what the variance should look like.</p>
<p>Finally, I’ve also added some more data points to it:</p>
<ol>
<li>
<p><strong>I researched the average consumption (in the <span class="caps">US</span>).</strong>
People tend to think in portions rather than in kg, and they tend to think in terms of cutting out animal products rather than reducing their consumption by an absolute delta. The new suffering times average consumption column greatly reduces the gap between beef and milk because people tend to consume so much milk, and it makes chicken meat worse than eggs because people tend to consume so much chicken meat.</p>
</li>
<li>
<p><strong>I’ve added a row for organic eggs.</strong>
I’ve often been asked how organic eggs stack up against battery cage eggs (or even more ambiguously, how aviary system eggs stack up). I never knew, because while the conditions are arguably better,<sup id="fnref:fowel"><a class="footnote-ref" href="#fn:fowel">1</a></sup> the chickens used are usually not Leghorns, so they produce fewer eggs per year and at more steeply declining rates so that they’re killed earlier. Therefore, more chickens are needed to satisfy the same demand.<sup id="fnref:elasticity"><a class="footnote-ref" href="#fn:elasticity">2</a></sup><sup id="fnref:thresholds"><a class="footnote-ref" href="#fn:thresholds">3</a></sup></p>
<p>I still don’t know – my estimate is much to close and unreliable for that – but it looks like organic eggs fare a bit better. The difference is less than even just one order of magnitude at the moment, so expect it to change when you touch some of the more sensitive inputs.</p>
</li>
</ol>
<h2 id="brians-reservations"><a class="toclink" href="#brians-reservations">Brian’s Reservations</a></h2>
<p>In 2014, Brian added some reservations to his article:</p>
<blockquote>
<p>I’m somewhat less gung-ho about these numbers than when I first wrote this piece because in practice, the side effects of meat consumption on wild animals and Earth’s long-run future probably matter much more than the (horrific) direct impacts on farm animals themselves. Of course, evaluating the net impact of these indirect side effects is much trickier.</p>
<p>Whatever the sign is of the indirect effects, indirect effects should be more similar across animals than the suffering figures are across animals, since cows and chickens don’t differ as much in their environmental impacts as in their direct suffering per kilogram. Hence, these neglected factors should tend to drive the (absolute value of the) ratios of per-kilogram impact estimates across species closer toward 1.</p>
</blockquote>
<p>Read <a href="https://foundational-research.org/the-importance-of-wild-animal-suffering/">The Importance of Wild-Animal Suffering</a> for more information.</p>
<h2 id="faunalytics-research"><a class="toclink" href="#faunalytics-research">Faunalytics Research</a></h2>
<p><a href="https://faunalytics.org/animal-product-impact-scales/">Faunalytics</a> has done a very similar investigation. They probably did more detailed research and used different and finer categories, which seem to match more closely categories of animal product–based dishes than the animal products that go into them. They present point estimates rather than ranges and opted not to weigh the intensity of the suffering:</p>
<blockquote>
<p>As one final note on the methodology, some readers may have encountered other impact estimates where subjective multipliers are applied to the amount of suffering experienced. We did not do so, treating each day of life as one day regardless of the quality of that life. Although we believe that differences in quality of life and suffering are probable, biases due to anthropomorphization or lack of sufficient data are also likely and, in our view, more problematic. It is worth noting that many of those animals for whom quality of life is likely lowest (e.g., layer hens, farmed fish) are already high on the impact list for other reasons. Any reader who would prefer to recalculate the estimates with additional subjective days of suffering can do so using the data and code files available on the Open Science Framework.</p>
</blockquote>
<h2 id="assorted-considerations"><a class="toclink" href="#assorted-considerations">Assorted Considerations</a></h2>
<p>Most of the considerations are not mine but <a href="http://reducing-suffering.org/how-much-direct-suffering-is-caused-by-various-animal-foods/#Appendix_Where_the_numbers_came_from">Brian’s</a> and <span class="caps">ACE</span>’s (<a href="https://docs.google.com/document/d/1mBS1eFoB2jNb1dqie8quBxGMPxdxNcd-I4AHqEnvnyk/edit">Dairy/Eggs <span class="caps">AYLA</span> and <span class="caps">AEPY</span>
</a>, <a href="https://docs.google.com/document/d/1ndtMnIwu-758kE2zNs2qJHMLB1buzLhm93O0JiY-U4w/edit">Meat Land Animal Equivalents Per Person Per Year</a>, <a href="https://docs.google.com/spreadsheets/d/1YSkZDTWacpkmnZMdRsIMLuCOdIILNVmGTIQAomZDxD4/edit#gid=0">Leafleting Impact Calculator</a><sup id="fnref:leafleting"><a class="footnote-ref" href="#fn:leafleting">6</a></sup>), so check out their work for more guidance. (I’ve also usually linked and copied their summaries into the Guesstimate cells.)</p>
<h3 id="leghorn-lifespans"><a class="toclink" href="#leghorn-lifespans">Leghorn Lifespans</a></h3>
<p>I’ve been reading various different figures on the average lifespans of Leghorns. These chickens can of course live many years, but they produce most eggs in their youth, their first year, and then fewer and fewer with every further year. Therefore, their average age at death is determined by a calculation that probably involves the monetary value of their meat, rent, and the price of new pullets.<sup id="fnref:comparison"><a class="footnote-ref" href="#fn:comparison">4</a></sup> The age that makes for the most profitable killing may be different between countries.</p>
<ol>
<li><a href="https://web.archive.org/web/20121103064431/http://awfc.ca/english/works/pub/disposehens.htm">Brian’s Canadian source</a> (before 2007 and perhaps before 2000 judging by the dates of the sources) indicates that they live for around 500 days and lay 288 eggs p.h. p.a. (per hen per year).</li>
<li>Norwood and Lusk (2011) write, “Cage eggs are assumed to be white eggs from a White Leghorn type breed. This hen will produce 509 eggs throughout its 2.21 years of life.” So around 800 days at 230 eggs p.h. p.a.</li>
<li><span class="caps">ACE</span>’s analysis of the <span class="caps">USDA</span> statistics from 2012 and 2014 indicates that they live around 412 to 515 days.</li>
<li>My different approach to almost the same data (I only have access to the 2012 statistics) puts the average age at 602 to 634 days (at 271–274 eggs p.h. p.a.).</li>
</ol>
<p>I discount Brian’s source only because it’s Canadian and over a decade old, so the economic conditions were surely different enough to explain the difference. Norwood and Lusk have their own university farms where they may allow their chickens to get older than they usually do; their figure is the highest of all. I rely most on the <span class="caps">USDA</span> statistics and there on my reanalysis.<sup id="fnref:reanalysis"><a class="footnote-ref" href="#fn:reanalysis">5</a></sup></p>
<h3 id="cow-lifespans"><a class="toclink" href="#cow-lifespans">Cow Lifespans</a></h3>
<p>Some data points:</p>
<ol>
<li><a href="https://www.quora.com/Whats-the-average-life-span-of-a-cow">Bovine enthusiast</a>: 5–6 years</li>
<li><a href="http://livestocktrail.illinois.edu/dairynet/paperdisplay.cfm?contentid=354">University of Illinois</a>: “The typical cow remains in the milking herd less than 4 years even though peak milk production related to maturity ordinarily does not decline until 8 or 9 years of age.” Cited by <a href="https://en.wikipedia.org/wiki/Dairy_cattle#Management">Wikipedia</a> as evidence for a lifespan of four years, but the author probably intended 4 + (1 to 2) years because the cow as to grow up first before she can produce milk. (Does someone want to correct it?)</li>
<li>Brian: <a href="https://albertamilk.com/ask-dairy-farmer/how-long-does-the-average-dairy-cow-live/">Alberta Milk (n.d.)</a>: “The typical dairy cow lives an average of five years, with the first two years focused on providing a strong foundation for the healthy development of the cow. From age two, the mature cow will become a productive member of the milking herd (meaning, she will produce milk).”</li>
<li><a href="http://extension.psu.edu/animals/dairy/nutrition/heifers/monitoring-heifer-growth/customized-dairy-heifer-growth-chart">Penn State</a> indicates that conception happens at around 15 months</li>
<li><a href="https://www.pubs.ext.vt.edu/404/404-285/404-285.html">Virginia Tech</a> implies that calves are born for the first time at 24 months.</li>
</ol>
<p>Most confusion seems to stem from the time the cow has to grow up before she can be artificially impregnated by the farmer, and then the time that her child grows up in utero before lactation starts. But there are still differences of three to four years even so. The result is not very sensitive to this input, so I haven’t researched it further, but if someone know what the most reliable data are, then I’d love to adapt the Guesstimate!</p>
<h3 id="sentience-multipliers"><a class="toclink" href="#sentience-multipliers">Sentience Multipliers</a></h3>
<p>For the sentience multipliers – numbers from the interval [0.0, 1.0] – I opted for something that captured my conflicting intuitions about sentience and looked properly arbitrary at the same time so not to signal sophistication where there is none. The formulae ended up looking like <code>=max(0, 1-lognormal(2.5, 1)/100)</code>. A log-normal distribution mirrored at a vertical axis so that it bulges up right before 1, and crudely tweaked so not to go below 0.</p>
<p>My intuitions are (1) that sentience may come in degrees, may reach different degrees for different individuals of the same species, may fluctuate for the same individual; (2) that experts disagree over the sentience of different species so that they have degrees of sentience with different probabilities; or (3) that people can legitimately have different opinions on the degree of sentience of an individual just as they can have differing opinions on how democratic each of the <span class="caps">US</span>, Switzerland, China, and North Korea are.</p>
<p>In each of these cases, my distributions mean different things, but the general shape seems to match all of them similarly well or badly.</p>
<ol>
<li>Chickens farmed for meat are killed very young, after around six weeks, so they receive a slightly lower sentience distribution than chickens farmed for eggs, because I think I was less sentient as a child too.</li>
<li>Fish like to cause little scientific kerfuffles (just search <a href="https://animalcharityevaluators.org/researchlibrary/"><span class="caps">ACE</span>’s Research Library</a> for “fish”) over whether they’re conscious or feel pain. Someone who has read this far is probably acquainted with expected value and will not use the possibility that fish might not suffer as an excuse to buy and eat them. That said, I was surprised by how sensitive the result was to this input. It still takes a probability of 0.0001 (or odds of 1:10,000) to press the expected suffering per kg for salmon to the level of milk, and there are probably nowhere nearly enough fish experts in the world to be sure at such an extreme level even if they all agreed, but as it is, the input is correlated with the result at r² = 0.58 (and looks correlated too in the scatter plot).</li>
</ol>
<h3 id="suffering-multipliers"><a class="toclink" href="#suffering-multipliers">Suffering Multipliers</a></h3>
<p>The suffering multipliers were somewhat informed by Dr. Sara Shields’ and Dr. Bailey Norwood’s estimates cited in Veganomics (and, for the latter, in <em>Compassion, by the Pound</em>).<sup id="fnref:nick-cooney-disclaimer"><a class="footnote-ref" href="#fn:nick-cooney-disclaimer">7</a></sup></p>
<ol>
<li>For the most part they agree nicely, but there’s the controversy over whether chickens farmed for meat have lives worth living. On this issue, they have strongly different opinions. My intuitions are informed by theirs, but seeing that a flu has a disability weight > 1 for me, you may make sure that you agree here. In particular, if you’re a classic utilitarian (of the <a href="https://www.youtube.com/watch?v=1CcgrdFF6ik">rat god variety</a> ;-)), this may make a sign-changing difference for you.</li>
<li>People are also notoriously unsure whether or to what degree Dr. Norwood factored in such things as sentience, transport, slaughter, slaughter-to-lifespan ratio, etc. in his estimates – that would be double-counting in our case. I don’t know anything about the context of Dr. Shields’ estimates, so I’ve tried not to rely on them too much.</li>
<li>The model only covers suffering, so the probably positive lives of cows farmed for meat is not captured. If offsetting happiness is a thing for you – as it may be for me in this case – then the result for beef is probably uninformative for you.</li>
<li>Note also that I’ve created a bit of arbitrary fuzziness around the value of 1 for the cow farmed for meat because different individuals are bound to have different life experiences.</li>
</ol>
<h2 id="footnotes"><a class="toclink" href="#footnotes">Footnotes</a></h2>
<div class="footnote">
<hr>
<ol>
<li id="fn:fowel">
<p>For example, according to FOWEL – and yes, I’m aware of the <a href="http://directactioneverywhere.com/theliberationist/2016/3/10/9oyu9qkx8jemid36kmihdh4ce0hiy4">big</a> <a href="http://www.openphilanthropy.org/blog/initial-grants-support-corporate-cage-free-reforms">debate</a> around this recently, but I’m from Europe where no one believes that there’s something worse than battery cages. So I’m shifting the burden of proof to DxE here. ;-) (In reality, I don’t think I have any influence on the prioritization decisions of Open Phil, Mercy for Animals, The Humane League, et al., so I tuned out of the debate to focus on things that I can affect.)</p>
<p>Norwood and Lusk (2011) on the topic: “In the FOWEL model, the cage system receives a score of zero [out of ten], a barn system (barn with hens uncaged on the floor) receives a score of 5.9, and an organic system scores 7.8.” They continue: “Much of the differences in opinion can be explained by different perceptions of the importance and magnitude of the mortality rates. Based on the available evidence, our estimate of the mortality rate in cage systems is about 3 percent. By contrast, we estimate mortality rates of 7 percent in cage free systems and 9 percent in free range systems …. Organic systems have higher mortality rates of about 13 percent because of feed restrictions. Organic producers cannot supplement animal feed with ‘unnatural’ synthetic (man made) amino acids. Another obstacle is the fact that a farmer cannot treat a sick animal with antibiotics and then sell the animal for organic food. This causes some farmers to deny antibiotics to sick animals. As a result, hens suffer. A number of animal scientists in the US believe organic production is cruel to hens for this reason.” <a class="footnote-backref" href="#fnref:fowel" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
<li id="fn:elasticity">
<p>I’m completely ignoring supply and demand elasticity here, because a shift to organic production would lead to a slight price increase due to the production and a big price increase due to warm-fuzzy premiums for retailers, thus probably reducing consumption. I’m ignoring these, because my audience are individuals potentially much more motivated by other factors than price. (Also, the calculations would get a heckuvalot more complex.) <a class="footnote-backref" href="#fnref:elasticity" title="Jump back to footnote 2 in the text">↩</a></p>
</li>
<li id="fn:thresholds">
<p>This tradeoff makes sense from a classic utilitarian or negative utilitarian perspective, but it doesn’t so much for me. I perceive some sort of threshold between milder discomfort and extreme suffering, and think about milder discomfort more in a classic utilitarian way (where it can be traded off against happiness) while I think about extreme suffering more in a prioritarian way (where it is lexically worse than any happiness could be good). Maybe it’s not lexical but just a really big jump, not sure. So if in one system few chickens suffered extremely (above the threshold) it would be lexically worse than one where many chickens suffered mildly (below the threshold), but neither would necessarily be of positive net value. But my threshold to extreme suffering is low (even a flu has a disability weight > 1 for me), so it probably doesn’t make a difference in this case. <a class="footnote-backref" href="#fnref:thresholds" title="Jump back to footnote 3 in the text">↩</a></p>
</li>
<li id="fn:comparison">
<p>One advantage of having vegan friends is that I don’t have to worry that they might eat me. Another, it seems, is that I won’t have to worry that they might kill me off to conserve government funds when I reach retirement age. <a class="footnote-backref" href="#fnref:comparison" title="Jump back to footnote 4 in the text">↩</a></p>
</li>
<li id="fn:reanalysis">
<p>ACE’s result seemed really surprisingly low to me, so I figured there’s a chance that ACE might’ve gone wrong here. I’m still not sure, though.</p>
<p>I only have access to the <a href="http://usda.mannlib.cornell.edu/usda/nass/ChickEgg//2010s/2013/ChickEgg-02-27-2013.pdf">2012 USDA statistics</a> (which cover 2011) and created <a href="https://docs.google.com/spreadsheets/d/1MkPLiQPGRXMLNGQbflX4mJZu7FEK9hhRRL1NDGNAxzk/edit">a spreadsheet for the calculations</a> (disregard the second sheet for now).</p>
<p>ACE’s approach was to divide the monthly totals of all <em>layer</em> hens “on hand” (stated prominently at the beginning of the report) by the monthly rate of “disposal” of all sorts – selling to slaughter, destruction of corpses, etc.</p>
<p>But on page 46, the report seems to distinguish “layers,” “pullets,” and “other chickens,” implying that whenever they talk about “layers,” they’re not referring to all chickens but only to the ones at sufficient age to lay eggs and also excluding whatever “other chickens” is. But the rate of “disposal” of chickens may (and I’m genuinely unsure here) not make that distinction, so that the total we should be using is not the total only of layers but that of layers, pullets, and other chickens all added up. This total is about 10 million higher in the mean over the two years.</p>
<p>Using the rate of replenishment and the (higher for some reason) rate of “disposal” to estimate the months of age of a chicken when it is killed, I get the two-year averages of 19.78 months and 20.83 months respectively. With months of around 30.4 days, that’s 602–634 days or 1.65–1.74 years. I’ve used these higher numbers in the Guesstimate for now. But note that this input is just about uncorrelated with all outputs, so it doesn’t really matter either way. <a class="footnote-backref" href="#fnref:reanalysis" title="Jump back to footnote 5 in the text">↩</a></p>
</li>
<li id="fn:leafleting">
<p>Sorry for using the dirty word, but it’s about much more than just leafleting! Please don’t dox me or call me names because I’ve said “leafleting.” Aaah! <a class="footnote-backref" href="#fnref:leafleting" title="Jump back to footnote 6 in the text">↩</a></p>
</li>
<li id="fn:nick-cooney-disclaimer">
<p>Disclaimer: I don’t condone the author’s conduct. (I have no first-hand information, but the description of a former employee matches gender-based gaslighting.) Cooney authored several informative books, so that I can’t help but cite him occasionally lest it seem that I plagiarize them. But I’m adding these disclaimers to avoid the impression that I accept such conduct or that it is accepted in my circles. <a class="footnote-backref" href="#fnref:nick-cooney-disclaimer" title="Jump back to footnote 7 in the text">↩</a></p>
</li>
</ol>
</div>Fandom Fundraising III: Recommendations2017-02-18T14:00:00+00:002017-03-04T12:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2017-02-18:/lessons-from-fandom-fundraising-recommendations.html<p>From 2011 to 2015, I’ve been involved in charity fundraising efforts that raised over $300,000 for several charities. I hope others can draw on some of the experiences documented here to repeat this success. This is the third of three articles and contains my recommendations for anyone who might want to replicate our efforts.</p><style>
blockquote {
color: #00A;
}
</style>
<div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#failures">Failures</a><ul>
<li><a href="#prioritization">Prioritization</a></li>
<li><a href="#donor-retention">Donor Retention</a></li>
<li><a href="#project-postmortems">Project Postmortems</a></li>
<li><a href="#collaborations-with-untrusted-parties">Collaborations with Untrusted Parties</a></li>
<li><a href="#donation-matching">Donation Matching</a></li>
<li><a href="#long-term-planning">Long-Term Planning</a></li>
</ul>
</li>
<li><a href="#unknown-signs">Unknown Signs</a><ul>
<li><a href="#miscellaneous-activities">Miscellaneous Activities</a></li>
<li><a href="#avoiding-brand-risks">Avoiding Brand Risks</a></li>
<li><a href="#not-incorporating">Not Incorporating</a></li>
<li><a href="#not-reinvesting-more">Not Reinvesting More</a></li>
<li><a href="#onboarding-of-volunteers">Onboarding of Volunteers</a></li>
</ul>
</li>
<li><a href="#bullets-dodged">Bullets Dodged</a><ul>
<li><a href="#politics">Politics</a></li>
<li><a href="#trademarks">Trademarks</a></li>
<li><a href="#technical-challenges">Technical Challenges</a></li>
<li><a href="#cannibals">Cannibals</a></li>
</ul>
</li>
<li><a href="#recommendations">Recommendations</a><ul>
<li><a href="#timing">Timing</a></li>
<li><a href="#targeting">Targeting</a></li>
<li><a href="#prioritization_1">Prioritization</a></li>
<li><a href="#diplomacy">Diplomacy</a></li>
<li><a href="#volunteers">Volunteers</a><ul>
<li><a href="#the-trouble-with-volunteers">The Trouble with Volunteers</a></li>
<li><a href="#bringing-it-home">Bringing It Home</a></li>
</ul>
</li>
</ul>
</li>
<li><a href="#finally">Finally</a></li>
</ul>
</div>
<p>You may want to start by reading the <a href="/lessons-from-fandom-fundraising-chronology.html">introduction to this series</a> of three blog posts if you haven’t already.</p>
<ol>
<li>
<p><a href="/lessons-from-fandom-fundraising-chronology.html">The first part</a> gives a chronological overview of our activities.</p>
</li>
<li>
<p><a href="/lessons-from-fandom-fundraising-strategies.html">The second part</a> describes how we organized our various events and is mostly interesting for someone who wants to see if there is something they can borrow from our experiences.</p>
</li>
<li>
<p>This third part highlights our learnings and makes recommendations as to how I think projects like these might be replicated in the future.</p>
</li>
</ol>
<p>My friend Martin Wilson contributed major parts of this post. His contributions are highlighted in blue.</p>
<blockquote>
<p>My name is Martin Wilson, and I was one of the principal founding members of Bronies for Good. From 2011 to 2014, I’ve been involved in charity fundraising efforts that raised over $300,000 for several charities. I hope others can draw on some of the experiences documented here to repeat this success.</p>
<p>I was heavily involved in most of our major fundraising campaigns, and worked hand in hand in with Denis Drescher to manage some of our key initiatives. In this brief retrospective, I’ll be both adding to some of the existing discourse that Denis has discussed sharing some of my own insights on successes and missed opportunities.</p>
</blockquote>
<h2 id="failures"><a class="toclink" href="#failures">Failures</a></h2>
<h3 id="prioritization"><a class="toclink" href="#prioritization">Prioritization</a></h3>
<p>Our biggest failure was that we took too long to learn about effective altruism and thus squandered a lot of potential.</p>
<p>I also think that supporting Against Malaria Foundation (<span class="caps">AMF</span>) was a mistake given the opportunity costs of not supporting something like Animal Equality (<span class="caps">AE</span>) earlier.</p>
<p>If <span class="caps">AMF</span> was 10–100 times as effective as the weighted-average project we supported before, and <span class="caps">AE</span> is 10–100 times as effective as <span class="caps">AMF</span>, then it lets our pre-2014 fundraising shrink to a relatively paltry €16–1,600 in <span class="caps">AE</span>-equivalent euros. And I have trouble empathizing with worldviews according to which these multipliers (or denominators) would be much smaller.</p>
<p>Supporting <span class="caps">AE</span>, finally, may similarly turn out to have been a mistake, but I don’t see any alternatives that would be better clearly enough to offset some minor problems we would incur: other top candidates like Animal Charity Evaluators, the Effective Altruism Foundation, the Machine Intelligence Research Institute, the Good Food Institute, etc. are not (yet) tax-exempt in an <span class="caps">EU</span> country, particularly Germany. This probably rules them out for most convention programs. The <span class="caps">EA</span> Foundation may very soon have this tax-exemption, but I think it is harmful to be too unreliable in one’s support of charities and I’m too uncertain about the relative ranking of the <span class="caps">EA</span> Foundation and <span class="caps">AE</span>, so that I don’t think I will want to switch.</p>
<h3 id="donor-retention"><a class="toclink" href="#donor-retention">Donor Retention</a></h3>
<p>Until 2014, we used a platform for fundraising that did not allow us to collect names and email addresses of our donors to send personalized messages to them. It did allow for the posting of “blog posts,” which would be distributed to all participants, but the email dispatch seemed to happen synchronously as part of the request, and so timed out after a minute or so. I suspect that we only reached a fraction of the donors this way. Many probably also opted out of these messages, which we might’ve avoided if we had been able to target them better.</p>
<p>Eventually we set up a Mailchimp list and send a “blog post” around inviting the readers to join our new list. Only a fraction of them signed up. Since, the list has again grown to over 900 subscribers, but we might have several times as many had we recorded them from the start.</p>
<h3 id="project-postmortems"><a class="toclink" href="#project-postmortems">Project Postmortems</a></h3>
<p>My main feelings about our projects were first fear that something might go wrong, and finally relief that most of it had worked out within parameters. So when we completed a project, we were (or I was) just happy to have survived it and quickly moved on. As a result, I now have trouble reconstructing the totals of many older projects because they are buried in millions of words of chat log.</p>
<p>Later, I started to record all important data in Google Sheets and publish them in blog posts. These are the resources I drew on to compile this article series.</p>
<h3 id="collaborations-with-untrusted-parties"><a class="toclink" href="#collaborations-with-untrusted-parties">Collaborations with Untrusted Parties</a></h3>
<p>Collaborations with big names promised exposure to many donors, but early on we made the mistake of collaborating with these big names on their terms. Unfortunately, they were chaotic and unconcerned about legal constraints, which could have endangered the trust people put in us.</p>
<h3 id="donation-matching"><a class="toclink" href="#donation-matching">Donation Matching</a></h3>
<p>I provided <a href="http://www.benkuhn.net/matching-results">donation matching</a> for some of our own fundraisers but did so from my donation budget without explaining this fact or the concept of counterfactual validity to the donors, which I now consider objectionable and uncooperative. I should not have counted this money toward my budget or generally refrained from using matching for our fundraisers.</p>
<h3 id="long-term-planning"><a class="toclink" href="#long-term-planning">Long-Term Planning</a></h3>
<p>We started to wonder about our end game when the fandom was already quickly declining – around 2014. We should have made precautions for this from the start. Unfortunately we didn’t know about effective altruism at that point, so this ties back to the first mistake.</p>
<h2 id="unknown-signs"><a class="toclink" href="#unknown-signs">Unknown Signs</a></h2>
<p>These are decisions where I am unsure whether they were good or bad on balance. Experimentation with the opposite decision may be interesting.</p>
<h3 id="miscellaneous-activities"><a class="toclink" href="#miscellaneous-activities">Miscellaneous Activities</a></h3>
<p>Apart from our fundraisers, we also engaged in some miscellaneous activities such as blood drives and appeals against bullying and discrimination. Some of these may only have diverted time and attention (also the attention of our donors), while others may have put off donors with opinions other than ours on topics fairly irrelevant to the charities we supported. Keeping a leaner identity may have been smarter. On the other hand, the activities did generate some attention an engagement and may have opened doors to groups that shared our opinions.</p>
<h3 id="avoiding-brand-risks"><a class="toclink" href="#avoiding-brand-risks">Avoiding Brand Risks</a></h3>
<p>We were often asked whether we have or want to start local chapters in various locations. Organizations such as Oxfam and Giving What We Can are doing this, but we were afraid to allow it lest less informed people at those chapters express opinions we would not endorse and thus jeopardize our image. This may have limited our impact, but may also have averted risks.</p>
<h3 id="not-incorporating"><a class="toclink" href="#not-incorporating">Not Incorporating</a></h3>
<p>In Germany, where I have been throughout the time, we had Your Siblings as a legal entity that could handle money. In the <span class="caps">US</span>, we decided against forming a 501(c)(3) because of the costs and bureaucracy involved. Instead we usually arranged for our donors to donate directly to the charities. Few of our donors asked for donation receipts for tax reasons, but especially some larger donors did and it seemed to be necessary for convention in some countries, so we took the tax status of our charities in Germany and <span class="caps">US</span> into account.</p>
<h3 id="not-reinvesting-more"><a class="toclink" href="#not-reinvesting-more">Not Reinvesting More</a></h3>
<p>We would’ve been limited in how much we can reinvest because of our legal status; any investment could’ve only come from our own donation budgets or directly from a donor. Still, my personal investments were lower than they could’ve been; this may not be true for others of our group who had lower incomes. I usually jumped at any particularly good investment opportunities, particularly in the past few years, but still they were limited. It is not clear to me how we could’ve reached much greater scale with additional investment that would’ve been within my donation budget. Beyond that, it would’ve been valuable to pay a salary to those of us who were struggling financially so they could dedicate more time to our work.</p>
<h3 id="onboarding-of-volunteers"><a class="toclink" href="#onboarding-of-volunteers">Onboarding of Volunteers</a></h3>
<p>The organization has fallen apart because we didn’t manage to recruit new team members at the same rate that we lost them due to college/university degrees or full-time work. The people we did bring on board were often not committed enough to learn the ropes and disappeared again, leading to high onboarding cost at low return.</p>
<p>Martin has analyzed the problem in greater detail.</p>
<blockquote>
<p>While I agree in principle with the premise that adopting effective altruism sooner would have enabled us to generate a stronger overall impact with the fundraising dollars we managed to raise, a greater potential was missed in building a strong, lasting core of fandom volunteers. According to data from the Brony Study research project, a series of collaborative survey studies on adult fans of <em>My Little Pony: Friendship is Magic</em>, the median age of fans was around 20.79 years old.</p>
<p>Through representative sampling, we can extrapolate that the majority composition of the community lies within the young, post-secondary school range. Individuals of this age are typically enrolled as full- or part-time students in postsecondary institutions, or are engaged in part-time labor force participation independent of academic study. And while young adults are increasingly pressed for time, the possibility of course credit or “impact-centric work” could have helped motivate many of the young adult fans of <em>My Little Pony: Friendship is Magic</em> to plug into organizations in their community. These bright minds could have brought an enormous degree of passion, creativity, and energy to organizations with deep service-centric programs. The need in nonprofit organizations for such volunteers is also great: <a href="http://www.volunteeralive.org/docs/Strategic%20Volunteer%20Engagement.pdf">Over 80% of registered third sector organizations in the <span class="caps">US</span> acknowledge the use of volunteers to meet critical service outcomes.</a> Regrettably, many potential volunteers faded into the demands of their personal lives, or were directed to numerous, disparate efforts and projects.</p>
<p>Our own attempts with volunteer engagement came in the form of our Nurse Redheart’s Roundup campaigns, where we encouraged our constituents to donate to local blood banks and centers in their communities. Donors would subsequently log their participation through an online form and be randomly entered into a sweepstakes to win token prizes. While the three instances of the campaign yielded a relatively modest return of 59.5 liters of blood, the campaign could have greatly increased its potential impact by seeking active collaborations with collection agencies in close proximity to population centers with active meetup groups. The replenishing of blood is a particularly salient issue, since much needed transfusion stocks can <a href="http://www.americasblood.org/stoplight.aspx">dwindle perilously during the down periods between major disasters</a>.</p>
<p>In sum, more of a robust recruitment and engagement process for community volunteers, along with greater outreach to corresponding service organizations, could have generated numerous fruitful collaborations and projects for the community as a whole. Especially if we had plugged into existing paradigms on volunteer engagement, such as Service Learning, which could allow university students in particular to pursue volunteer work as <a href="https://www.learningtogive.org/resources/academic-service-learning-reflection-concept">coursework, research, or even culminating experiences</a>.</p>
</blockquote>
<h2 id="bullets-dodged"><a class="toclink" href="#bullets-dodged">Bullets Dodged</a></h2>
<h3 id="politics"><a class="toclink" href="#politics">Politics</a></h3>
<p>We were always very careful not to make enemies. Star Trek’s Prime Directive about noninterference may sound stupid at first, but it has one crucial advantage: When you get in touch with some new civilization or fandom, then what you’re really doing is probably that you get in touch with one particular faction of it. The opposing factions (of whom you’re not aware) will interpret this as you siding with their opponent. You will be assailed from all sides before you know that there is even a war going on. Or some less dramatic version.</p>
<p>In fandoms there are people who pour months (or years) of work and ten thousands of dollars into some fandom project of theirs only to then have some competitor overtake them by luck or by having poured five times as much money into it. They despise each other in proportion to their investment and earlier or later one will subtly sabotage the other and the war begins.</p>
<p>When I first got in touch with the fandom, I was concerned with trying to find the main multipliers who I would have to send my press releases to. Soon I found the top one, but then it turned out that it was only the voice of one faction of the fandom.</p>
<p>When you’re trying to achieve something, it can be really bad to have opponents. So I tried hard to treat every cooperation partner equally. For news sites this meant that I would put in the extra minutes it took to reformat the press release for all the other news sites even if they only had a hundredth part of the readership of the big one. (They would only copy and paste the text, so it needed to be formatted such that it would look okay right away when it landed on the website.)</p>
<p>For online radio channels, it meant that we would try to get everyone to play the same premiere recording in parallel (this never quite worked) or stream it ourselves and have them restream it from us.</p>
<h3 id="trademarks"><a class="toclink" href="#trademarks">Trademarks</a></h3>
<p>Months or years of work have gone into many fandom projects that were eventually shut down by Hasbro because, obviously, they contained ponies. These ponies, and especially their names, are trademarked.</p>
<p>We’ve always been careful to minimize the risk of being shut down with a Cease <span class="amp">&</span> Desist notice. Depending on the area of one’s activism, this can be an important consideration.</p>
<h3 id="technical-challenges"><a class="toclink" href="#technical-challenges">Technical Challenges</a></h3>
<p>We faced a number of challenges with regard to the platforms we used for our fundraising.</p>
<p>Betterplace is free – it even pays for the transfer fees – but depending on how you use it, you may have trouble getting names and email addresses of your donors. When we used its <span class="caps">API</span>, we were able to implement this part of the funnel ourselves, but that was only warranted because we needed the system anyway for the Your Siblings website. Otherwise the CauseVox fee may be the smaller expense.</p>
<p><span class="caps">AMF</span>’s fundraiser system is hard to use – you have to fill in a lot of data redundantly, and the WorldPay interface is highly confusing. We probably lost some donors who couldn’t figure it out. The advantage is of course, that it is provided directly by <span class="caps">AMF</span>, so that no intermediary deducts any fee from the donations, except for transfer fees.</p>
<p>CauseVox, eventually, was recommended to us by Charity Science. It is highly usable and charges a relatively modest fee when you use the “free” plan. It’s our current solution. (It might’ve been risky to continue to use Your Siblings’ system after I left the organization, so we rather switched.)</p>
<h3 id="cannibals"><a class="toclink" href="#cannibals">Cannibals</a></h3>
<p>Some say that there are millions of bronies, but the number of people who are active enough that one could reach them for any kind of charity campaign may be in the six-digit area.<sup id="fnref:brony-count"><a class="footnote-ref" href="#fn:brony-count">1</a></sup> Only a fraction of them are well-off and altruistic enough to become donors.</p>
<p>There were several charities active in the fandom – and I’m using “charity” in a loose sense that includes individuals organizing yearly charity events. Only one other one of them displayed in their decisions anything resembling a concern for effectiveness. With such a small donor base to draw on, and people usually having informal donation budgets, it is likely that someone who donates heavily to a brony fundraiser for a CalArts scholarship will later not have enough spare change to save some lives in a Bronies for Good fundraiser.</p>
<p>Even worse, when a convention runs an auction for a charity that is, say, a thousand times less effective than another charity that we fundraise for at the same convention, then a lot of people will not donate to us because they want to save the money for the auction. They may be indifferent to the difference in effectiveness between the charities or not even realize that the auction benefits a different charity.</p>
<p>In the cases where we were on good terms with a convention but the convention wanted to dedicate the auction to a much less effective charity, we have repeatedly attempted to explain effectiveness considerations to convention organizers, but the ones that were not convinced from the start also didn’t come around. It turned out to be much easier to annoy them with our attempts than to educate them. Since our collaborations with conventions are extremely valuable, we decided to ignore the squandered funds and focus on staying on good terms with the respective conventions.</p>
<p>Seeing how donation drives usually try to capitalize on the Christmas spirit, they usually take place in December. We could have exploited this dynamic, and run our drives in November, so that we may miss out on the Christmas spirit, but will instead profit from reduced cannibalism. (Well, we’ll be the cannibals.) This may be analogous to the property of being <a href="https://en.wikipedia.org/wiki/The_Evolution_of_Cooperation#Axelrod.27s_tournaments">“provocable” of a tit-for-tat agent</a> in an interated prisoner’s dilemma, so it will be important to be tranparent about our algorithm.</p>
<h2 id="recommendations"><a class="toclink" href="#recommendations">Recommendations</a></h2>
<h3 id="timing"><a class="toclink" href="#timing">Timing</a></h3>
<p>If you want to repeat what we have done in another community, then it will be important to join at the right moment, just when it becomes evident that the community is growing rapidly. My feeling is that in the case of the pony fandom, this means that joining in the first year is optimal (late 2010 to late 2011), joining in the second year still works well (late 2011 to late 2012), joining in the third year will take more effort (late 2012 to late 2013), joining in the fourth year may no longer be worth it (late 2013 to late 2014), and after that you should rather look for another opportunity. But other communities probably expand and decay at very different paces.</p>
<h3 id="targeting"><a class="toclink" href="#targeting">Targeting</a></h3>
<p>The pony community attracted more than average sensitive people and probably people who were also more altruistic than the average. This was a great asset in motivating people.</p>
<p>Problems were, however, that there was no particular selection for rationality in place and that many fans were school or college students so that they didn’t have a lot of money. These are also criteria that you could use to decide between different opportunities.</p>
<p>$300,000 may fall short of the expectations that some organization have into fundraising campaigns that require the degree of effort ours did. Better targeting and possibly diversification by educating and supporting people like us in many communities may be a more efficient strategy.</p>
<h3 id="prioritization_1"><a class="toclink" href="#prioritization_1">Prioritization</a></h3>
<p>Of course, <a href="https://www.effectivealtruism.org/">a solid grasp of prioritization</a> is crucial. There is no use in organizing for months to raise $10,000 if it goes to a charity that is a thousand times less effective than some top charity; you could just transfer $10 to the more effective one and spend the rest of the time cuddling.</p>
<h3 id="diplomacy"><a class="toclink" href="#diplomacy">Diplomacy</a></h3>
<p>When you do spot an opportunity to join a promising community at a good time, then it may be best to become a member of that community yourself. We had the great asset that we were first greatly enamored of the show and then wanted to make something of it. Conversely, it would seem dishonest to try to pass as a fan of something that you don’t really care about. It’s not recommended. If you are clearly not enthusiastic about the locus of the community, then you can try any of these:</p>
<ol>
<li>
<p>You can be completely transparent about what you’re doing and why (think of <span class="caps">REG</span>’s work outside of poker). Like-minded people will realize that the value of your activism is not dependent on your in-group-ness. (This may not have worked with ponies, because the number of sufficiently like-minded people were probably in the hundreds, rather than in the thousands, like our donors.)</p>
</li>
<li>
<p>You can find people who are already organizing charity events in the community or submit a press release that calls for volunteers to a community hub. These people could then do most of the things we have been doing while you supervise, coordinate, and support them with advice, connections, money, legal entities, etc. If you are actually in some similar community, then you can communicate your like-mindedness through your brand. Imagine Better has attempted something like this by choosing a name that communicates their association with the Harry Potter fandom (it’s from a <span class="caps">J.K.</span> Rowling quote) without being limited to it.</p>
</li>
</ol>
<p>If you wish to concentrate your efforts on a small number of communities, then it’ll be important to be very careful about how you position yourself within the communities. As mentioned throughout the text, it is easy to first think that you identified some nexus of communication in a fandom only to then realize that it is only the nexus of communication of one faction of it, and that by associating with it, you have made enemies among the other factions. When this happens accidentally, it’s usually possible to mend fences with the other factions, but that requires that you first become aware of the situation, which can take time. My feeling is that the most likely failure scenario is that the majority of people in the community will see you as a rather non-agenty part of the outgroup, so they will just ignore you.</p>
<p>For similar reasons, it’s safer to maintain an image of independence from other groups in the fandom, in the sense that you collaborate with them, but that you are not a subsidiary or project of theirs.</p>
<p>If, conversely, you have identified a number of mildly promising communities and want to fundraise in all of them, then less caution may be warrented (but this is untested). By seeking a strong cooperation partner as interface to the community and relying on its brand and network, you can save a lot of time, gaining traction more quickly. Meanwhile, the diversification can keep your risks managable.</p>
<p>We have, for example, observed projects associating with radio stations and thus profiting from the established brand of the station, its reach among listeners, its contacts among prominent people they have interviewed, and its contacts to conventions (for recording and streaming of panels). In other communities, other forms of media may suggest themselves.</p>
<h3 id="volunteers"><a class="toclink" href="#volunteers">Volunteers</a></h3>
<h4 id="the-trouble-with-volunteers"><a class="toclink" href="#the-trouble-with-volunteers">The Trouble with Volunteers</a></h4>
<blockquote>
<p>While discussing the need for volunteer engagement is all well and good, a quick and dirty problematization of volunteer coordination reveals some rather substantial obstacles: </p>
<ol>
<li><strong>Time.</strong> Between Family, Work, School, and other commitments, being able to coordinate even a single local volunteer group would have exceeded the available operational bandwidth of a non-incorporated group. The amount of uncompensated hours would have also been an enormous demand to make to a single member.</li>
<li><strong>Organizational structure.</strong> Thinking through how we, as part-time non-incorporated organization, could put together a system that allowed volunteers to connect with partner organization is a daunting challenge on it’s own. Perhaps something akin to a geek-themed, volunteer-focused version of LinkedIn might serve here?</li>
<li><strong>Quality control/training.</strong> While volunteers are a wonderful asset to any organization, not all volunteers are useful contributions. Some organizations, such as San Francisco Suicide Prevention, require volunteers to undergo extensive training on <span class="caps">HIPAA</span> laws and phone etiquette. Ensuring that any volunteers sent their way would complete the training, let alone prove effective at the intensive emotional nature of this work. Thinking through quality control issues with potential partners and liability pitfalls would potentially undermine some collaborations completely. How candidates could plug into specific projects or existing services that an organization offers is also worth exploring as well.</li>
<li><strong>Resistance to community/sub-culture.</strong> While most organizations will jump at the chance to have a steady supply of volunteers, not all organizations are led by open-minded, welcoming individuals. While we have not experienced directly, I have been on the receiving end of an organization refusing to accept a monetary donation form efforts I was involved in due a certain member of the nonprofit’s Board of Directors expressing misgivings about an active association with adult fans of <em>My Little Pony: Friendship is Magic</em>.</li>
<li><strong>Volunteer liability.</strong> The finding, placing, and ongoing management of a volunteer: While this would obviously vary depending on the organization, creating a mutual understanding on liability issues around a volunteer’s participation with a potential partner would be a question that merits careful and systematic review.</li>
<li><strong>“How is this pony-related?”</strong> And last, but certainly not least: How to market to potential volunteers how and why they should get involved with a certain partner, and how it feeds back into the undercurrent of interest that binds the community together and generates effective calls to action.</li>
</ol>
<p>Our small, unincorporated group also attempted to recruit volunteers; however, we often found that while individuals were interested in working with us, many lacked tangible skills, maturity, or personal stability to remain involved in a capacity that would sustain the organization in the long term.</p>
</blockquote>
<h4 id="bringing-it-home"><a class="toclink" href="#bringing-it-home">Bringing It Home</a></h4>
<blockquote>
<p>While this is by no means an exhaustive list or reflection, I hope it provides a kind of reflection point on how our experience in attempting to engage volunteers where they lie. And while it would be a wicked challenge, channeling energy and attention to pursue a more systematic, volunteer-centric strategy to spreading altruism can be instrumental in fostering empathy, generosity, and localized connections. Imagine if the spontaneous creativity, passion, and innovation that permeates online affinity communities could reach beyond its confines into the troubled world around it. Where we, as agents of altruism, channel these elements into real, service-focused work, this work can bring an impact that can help alleviate the deep, systemic inequalities and traumas of consumer capitalism. All while providing fertile ground for the personal self-discovery and passion of young adults the world over. Ideate, create, and transform!</p>
</blockquote>
<h2 id="finally"><a class="toclink" href="#finally">Finally</a></h2>
<p>If I’ll notice that I’ve forgotten something major and add it later, I’ll send out an update on Facebook.</p>
<p>That’s it. Good luck!</p>
<div class="footnote">
<hr>
<ol>
<li id="fn:brony-count">
<p>The strongest period on record at the news site that I also worked for – early 2012, shortly after we started recording – saw 33,000 returning users. If they are not truely unique but different browsers and devices used by the same person, we may be at 10,000 unique readers. The largest news site in the fandom had 10–100 times the reach, probably around 30 times, so we’re at 100,000 to 1 million people, somewhat skewed downwards. <a class="footnote-backref" href="#fnref:brony-count" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
</ol>
</div>Fandom Fundraising II: Strategies2017-02-11T12:00:00+00:002017-02-11T12:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2017-02-11:/lessons-from-fandom-fundraising-strategies.html<p>From 2011 to 2015, I’ve been involved in charity fundraising efforts that raised over $300,000 for several charities. I hope others can draw on some of the experiences documented here to repeat this success. This is the second of three articles and gives a detailed account of how we organized specific campaigns.</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#conventions">Conventions</a><ul>
<li><a href="#running-a-table-or-booth">Running a Table or Booth</a></li>
<li><a href="#auctions">Auctions</a></li>
</ul>
</li>
<li><a href="#charity-albums">Charity Albums</a><ul>
<li><a href="#resources">Resources</a></li>
<li><a href="#process">Process</a></li>
<li><a href="#learnings">Learnings</a></li>
</ul>
</li>
<li><a href="#album-premiere-livestreams">Album Premiere Livestreams</a><ul>
<li><a href="#resources_1">Resources</a></li>
<li><a href="#timeline">Timeline</a></li>
</ul>
</li>
<li><a href="#art-livestreams">Art Livestreams</a><ul>
<li><a href="#resources_2">Resources</a></li>
<li><a href="#process_1">Process</a></li>
<li><a href="#learnings_1">Learnings</a></li>
</ul>
</li>
<li><a href="#panel-livestreams">Panel Livestreams</a></li>
<li><a href="#gaming-livestreams">Gaming Livestreams</a></li>
<li><a href="#comic-books">Comic Books</a><ul>
<li><a href="#resources_3">Resources</a></li>
<li><a href="#process_2">Process</a></li>
<li><a href="#learnings_2">Learnings</a></li>
</ul>
</li>
</ul>
</div>
<p>You may want to start by reading <a href="/lessons-from-fandom-fundraising-chronology.html">the introduction to this series</a> of three blog posts if you haven’t already.</p>
<ol>
<li>
<p><a href="/lessons-from-fandom-fundraising-chronology.html">The first part</a> gives a chronological overview over our activities.</p>
</li>
<li>
<p>This second part describes how we organized our various events and is mostly interesting for someone who wants to see if there is something they can borrow from our experiences.</p>
</li>
<li>
<p><a href="https://claviger.net/lessons-from-fandom-fundraising-recommendations.html">The third part</a> highlights our learnings and makes recommendations as to how I think projects like these might be replicated in the future.</p>
</li>
</ol>
<h2 id="conventions"><a class="toclink" href="#conventions">Conventions</a></h2>
<p>The part of our fundraising that is still going strong is everything we do at conventions. In the following, I will use GalaCon as a model. In its five years, we have raised over €77,000 at GalaCon alone.</p>
<p>The conventions that we’re cooperating with originally reached out to us, but we could’ve probably reached out to them just as well.</p>
<p>We haven’t had the ambiguous case that a convention offered us a comped table, did not accept our recommendation for the charity auction, and yet chose a charity for the auction that could likely be in the same ballpark in terms of effectiveness as our recommendation would’ve been. That case is probably similar enough to the one where they accept our recommendation that I don’t need to cover it separately.</p>
<h3 id="running-a-table-or-booth"><a class="toclink" href="#running-a-table-or-booth">Running a Table or Booth</a></h3>
<ol>
<li>
<p>If the convention offers a comped table, then we can use it to accept donations and give out merchandize as thank yous, and to advertize effective altruism and our charity or charities of choice.</p>
</li>
<li>
<p>Here, we make it a point to call what we do “accepting donations” and “giving out merchandize as thank yous” (not “selling”). This allows people to feel good about having donated, keep them in a generous frame of mind, and avoids any framing (common at conventions) where “this is overpriced” would make sense. Some of our gifts are clearly symbolic relative to the donation threshold, e.g., our buttons/pins that people can wear as trophies, so the framing around thank you gifts is probably more accurate too.</p>
</li>
<li>
<p>Our merchandize has included the following – with the donation thresholds above which we give them out:</p>
<ol>
<li>
<p><em>Bound Together</em>, our comic book (≥ €25).</p>
</li>
<li>
<p>Bit coins, brass coins similar to those (called “bits”) used in the show (≥ €1 or €2).</p>
</li>
<li>
<p>Different buttons, circular 5.5 cm across and custom shaped metal pins (≥ €5).</p>
</li>
<li>
<p>Postcards designed by fandom artists (≥ €10–12).</p>
</li>
<li>
<p>Indie collectors cards, produced by artists in cooperation with us and GalaCon (≥ €12).</p>
</li>
<li>
<p>Drawings by artists (≥ €10–12).</p>
</li>
<li>
<p>Animal Equality and Animal Charity Evaluators booklets (no threshold).</p>
</li>
<li>
<p>A general <span class="caps">EA</span> leaflet (no threshold).</p>
</li>
</ol>
</li>
<li>
<p>The strongest sellers have been:</p>
<ol>
<li>
<p>Bound Together – especially in Germany and especially back when people didn’t have it yet.</p>
</li>
<li>
<p>Buttons – especially metal pins and back when people didn’t have them yet.</p>
</li>
<li>
<p>Bit coins – which are pretty new, so people don’t have enough of them yet.</p>
</li>
</ol>
</li>
<li>
<p>Our stock is always counted beforehand and split in bundles of 25, 50, or something small like that.</p>
<ol>
<li>
<p>We try to count after each day (most cons are two-day events around my parts).</p>
</li>
<li>
<p>This is easy, because you can easily count all the untouched bundles and multiply the count by the bundle size. Then you count the leftover from the opened bundle and add it. Finally, you subtract the sum from the original total you had in stock.</p>
<ol>
<li>If we’re too tired and fail to count after the first day and we have enough stock, we can count the untouched bundles in the morning and leave the opened bundle aside. Then we can reconstruct the tally of the first day after the con.</li>
</ol>
</li>
</ol>
</li>
<li>
<p>Counting the money:</p>
<ol>
<li>
<p>At a very basic level, <a href="http://www.wikihow.com/Count-Money">I follow this counting method</a>.</p>
</li>
<li>
<p>First, I need to find a safe, sheltered, and silent place, ideally somewhere at someone’s home behind locked doors.</p>
</li>
<li>
<p>I at least have trouble recognizing people in different clothing and with different hairstyles unless I’ve known them for a bit. If I need to transport the money to another place, I try to make myself look as different as easily possible from how I looked on stage.</p>
</li>
<li>
<p>When I have arrived at a suitable place, I need a lot of space, rubber bands, and plastic bags.</p>
</li>
<li>
<p>I try to keep as many small subtotals as feasible, because I will make errors, and when I do, I want to have to recount (several times) as small of a fraction of the money as possible.</p>
<ol>
<li>Evidently, making the subtotals maximally small gets you back to not having subtotals at all. I prefer something around ten bills/coins.</li>
</ol>
</li>
<li>
<p>I count every subtotal (and later the number of stacks) several times.</p>
</li>
<li>
<p>I write down every little thing and use a calculator even for very simple operations.</p>
<ol>
<li>I try to be unself-conscious enough to type 20 times 6 into a calculator. After getting 50% of my sleep for two nights in a row and then working a 16-hour day at the table, I might even forget to hyphenate a compound modifier whose first part does not end in <em>-ly</em> and which modifies a noun phrase that follows it!</li>
</ol>
</li>
<li>
<p>I want to make the bank teller happy, so I align all bills in the same direction. If I didn’t do it, though, it would have the advantage that I could marvel at the superhuman speed (two or three bills per second maybe?) at which the tellers count, realign, and test the bills all at once.</p>
</li>
<li>
<p>Finally, I roll up the bills, still sorted, and put the coins into separate bags if I remembered to bring bags.</p>
</li>
<li>
<p>The German Postbank has been best with accepting a lot of cash.</p>
</li>
</ol>
</li>
</ol>
<h3 id="auctions"><a class="toclink" href="#auctions">Auctions</a></h3>
<ol>
<li>
<p>Open submissions for in-kind donations months before the convention, maybe three to six months.</p>
<ol>
<li>
<p>Collect data such as:</p>
<ol>
<li>
<p>The name of the donor.</p>
</li>
<li>
<p>A description – in particular one that points out what makes the item valuable.</p>
</li>
<li>
<p>The estimated value – this will be the starting bid.</p>
</li>
<li>
<p>Additional notes – such as that it has been donated by a company on the condition that the company is credited during the auction.</p>
</li>
<li>
<p>Means of submission – whether they’ll send it by mail or will hand it in at the convention.</p>
</li>
<li>
<p>What they would like to have happen if it is not auctioned off – it can stay in the possession of the convention for the next year, the donor can reclaim it right after the event, or they can pay for the convention mailing it back.</p>
</li>
</ol>
</li>
<li>
<p>Asking for the estimated value is very important so that nothing is auctioned off below the production cost!</p>
</li>
<li>
<p>GalaCon usually only allows submissions at the convention until a certain time, so that they have enough time to prepare the auction sequence.</p>
</li>
<li>
<p>GalaCon also asks for a picture up front. If it is not good enough, they can still scrap it and take a new one.</p>
</li>
<li>
<p>GalaCon at times politely declines donations that are clearly of little value.</p>
</li>
</ol>
</li>
<li>
<p>GalaCon uses an Access database to maintain all submission data and automatically generate a presentation and silent auction sheets from it. This is probably a bit more work with Google Sheets.</p>
</li>
<li>
<p>The following assumes that the convention has received a lot of in-kind donations, potentially more than it can auction off. When a convention receives less, the returns are bound to be much lower, and less effort is warranted.</p>
</li>
<li>
<p>The sequence of auctions is currently one of three events: (1) silent auction, (2) small auction, and (3) main auction.</p>
<ol>
<li>
<p>The silent and small auctions help get rid of less valuable items because the time of the main auction is limited and extremely valuable.</p>
</li>
<li>
<p>Items that got hardly any bids in the silent auction are awarded to the winners; items that got a few more bids go into the small auction; more popular items are held back for the main auction.</p>
</li>
</ol>
</li>
<li>
<p>Silent auction:</p>
<ol>
<li>
<p>All items (or possibly minus the most valuable items) are on display, always guarded or behind glass.</p>
</li>
<li>
<p>They have numbers, and there are numbered sheets, one per item, that give donor and description of the item, list the rules, and let people enter their bids.</p>
</li>
<li>
<p>To bid, people need to get numbered donor cards by entering their contact data into a list and signing that they accept the rules.</p>
</li>
<li>
<p>Then they enter their number and bid into the respective sheet.</p>
</li>
<li>
<p>The rules can also be used to explain the other auctions. They stipulate:</p>
<ol>
<li>
<p>Who the auctions work, so much of this subsection.</p>
</li>
<li>
<p>The minimum increments of the bids.</p>
</li>
<li>
<p>The timeframe of the silent auction.</p>
</li>
<li>
<p>The means by which the winners will be announced.</p>
</li>
<li>
<p>The payment methods and currencies that we can accept.</p>
</li>
<li>
<p>That we do not give out items on credit (it worked a few times but we also lost €600 from someone who didn’t pay up after the first GalaCon).</p>
</li>
<li>
<p>Where the ATMs are, so people can fetch enough cash before or during the auction (we have allowed people to reserve items they won for a few minutes to go to an <span class="caps">ATM</span>).</p>
</li>
<li>
<p>That people who disrupt the auction can be removed from the hall.</p>
</li>
</ol>
</li>
</ol>
</li>
<li>
<p>Small auction:</p>
<ol>
<li>
<p>The small auction takes place in some context that distinguishes it clearly from the main auction auction so not to detract from its prestige.</p>
</li>
<li>
<p>The items are less valuable, but otherwise the same procedure applies as to the main auction.</p>
</li>
</ol>
</li>
<li>
<p>Main auction:</p>
<ol>
<li>
<p>The sequence of items is important because we don’t want people to hold back to save their money for a particular valuable item at the end only to then, overwhelmingly, not win it and take their money home again.</p>
</li>
<li>
<p>It may be useful to start with one or two small items, so that people can learn how the auction works in a low-stakes situation, but then we want to quickly move on to the valuable items.</p>
</li>
<li>
<p>The order should be set beforehand and the items arranged on stage accordingly to save time and prevent mistakes.</p>
</li>
<li>
<p>The auctioneer spends less than three minutes on the least valuable items and less than ten minutes on the most valuable items with the rest around five minutes each.</p>
</li>
<li>
<p>They wait a maximum of 20 seconds for a new bid before selling, often much less.</p>
</li>
<li>
<p>In addition to the auctioneer, GalaCon also has someone on stage to accept the payments, protocol them, and hand out the items, and several people to assist the two.</p>
</li>
<li>
<p>Behind the scenes, the main organizer gives instructions to the auctioneer via headphones and other people operate the projector.</p>
</li>
<li>
<p>GalaCon shows the photos on the projector to make sure that everyone in the audience (1000+ people) knows what is being auctioned off.</p>
</li>
<li>
<p>Sometimes guests of honor at the convention come on stage to animate the audience. This works well if they know how auctions work and that the time is very valuable. We limit these performances to valuable items and to 3–5 minutes each.</p>
</li>
</ol>
</li>
<li>
<p>A professional auctioneer – but one who is used to inexperienced audiences – makes a big difference, because they’ll be as fast as the audience allows. If in doubt, faster seems to be better.</p>
</li>
</ol>
<h2 id="charity-albums"><a class="toclink" href="#charity-albums">Charity Albums</a></h2>
<p><img alt="Donations to our Seeds of Kindness fundraisers 1–4" src="/images/seeds-of-kindness-donations.png"></p>
<p>Our greatest success and key to our online fundraising strategy with about €150,000 raised has been our series of charity albums. (The graph above, however, includes some offline donations from conventions.)</p>
<h3 id="resources"><a class="toclink" href="#resources">Resources</a></h3>
<ol>
<li>
<p>A few highly committed friends.</p>
</li>
<li>
<p>Access to a few news outlets/blogs that are read by the target demographic.</p>
</li>
<li>
<p>Contact to about 100 good amateur musicians.</p>
</li>
<li>
<p>3–5 friends who are very good with music as judges.</p>
</li>
<li>
<p>Someone who is good at cover art.</p>
</li>
<li>
<p>An account with something like CauseVox.</p>
</li>
<li>
<p>Friends at online radio channels for a big premiere.</p>
</li>
</ol>
<h3 id="process"><a class="toclink" href="#process">Process</a></h3>
<ol>
<li>
<p>About six months out, we ~make plans for pulling it all off in four months~ agree on a title.</p>
</li>
<li>
<p>We set a release date and a soft and hard deadline for musicians to submit tracks. If they submit after the soft deadline, we won’t be able to give feedback and allow them to resubmit an updated version.</p>
<ol>
<li>Spacing these out by one month each gives us some room to pester and wait for important (gifted or famous and gifted) contributors without stretching the patience of the musicians who submitted in time.</li>
</ol>
</li>
<li>
<p>Then we announce the project on our blog with all necessary details (<a href="http://broniesforgood.org/2016/10/announcement-acoh/">here a sample announcement</a>):</p>
<ol>
<li>
<p>We announce it on various pertinent forums where musicians are.</p>
</li>
<li>
<p>We email all musician we are in contact with or who have contributed in the past.</p>
</li>
<li>
<p>We send a shortened version of the announcement (with reference to the full version) out to all news sites we are in touch with.</p>
</li>
<li>
<p>We post public reminders every month or so.</p>
</li>
<li>
<p>There were or are several competing news sites in the fandom, so it was important to involve them all equally in the premiere lest the ones that feel excluded conclude that we must have sided with their competition in their conflicts that we want to have nothing to do with.</p>
</li>
</ol>
</li>
<li>
<p>As tracks come in, we have the judges decide whether they make the bar, and give feedback otherwise so that the musicians have a chance to resubmit an improved version.</p>
</li>
<li>
<p>One or two months out, we put together the release party and find an artist for the cover.</p>
<ol>
<li>There were several competing radio stations in the fandom, so it was important to involve them all equally in the premiere lest the ones that feel excluded conclude that we must have sided with their competition in their conflicts that we want to have nothing to do with.</li>
</ol>
</li>
<li>
<p>After the hard deadline, we start putting together the album:</p>
<ol>
<li>
<p>Encoding all tracks with the same codecs and settings, e.g., <span class="caps">FLAC</span> or <span class="caps">MP3</span> <span class="caps">VBR</span> 0.</p>
<ol>
<li>Most people expect some <span class="caps">MP3</span> format, but some prefer losslessly encoded files.</li>
</ol>
</li>
<li>
<p>Cleaning and filling in all <span class="caps">ID3</span> metadata.</p>
<ol>
<li>I first convert them all to <span class="caps">FLAC</span>, then fix the metadata with EasyTag, then convert them all to the compressed format.</li>
</ol>
</li>
<li>
<p>Normalize the volume of all tracks (one of our judges can usually do this).</p>
</li>
<li>
<p>Experiment with different orders of the tracks (our judges usually have strong opinions here).</p>
</li>
<li>
<p>Produce versioned release candidates while new songs come in, titles change, orders change, etc.</p>
<ol>
<li>These archives are huge, so I like to create a copy of the previous version on the server, rename it to the latest version, and then use rsync to only upload the diff between them.</li>
</ol>
</li>
</ol>
</li>
<li>
<p>We prepare a preview video of the top songs for the announcements.</p>
</li>
<li>
<p>We tell each musician whether their song has made it into the album and do so in time ahead of the premiere.</p>
</li>
<li>
<p>We prepare a submission schedule according to which every musician uploads their song or songs to YouTube on a particular day and submits it to the respective news sites. This avoids that all musicians submit at the same time increasing the standards for everyone, and only very few songs actually make the cut.</p>
</li>
<li>
<p>The “release party” livestream has its own section below.</p>
</li>
<li>
<p>A week or so after the release we follow it up with an announcement of the album fundraiser itself for everyone who couldn’t make it to the release party and then forgot about it. It gives an overview over what we’re planning for the year and also contains the album preview.</p>
</li>
</ol>
<h3 id="learnings"><a class="toclink" href="#learnings">Learnings</a></h3>
<ol>
<li>
<p>In our first fundraisers of this sort, we wanted to reward our top donors with donated gifts. People would offer in-kind donations for the top donors; months later, when the top donors could be selected, we asked these people to send their in-kind donations to the top donors. This was a mistake, because at that point, many of the these people could no longer be reached or did not possess the item anymore and we had to apologize to the top donors. We could’ve acted as escrow agent, but we decided to rather do away with the prizes altogether.</p>
</li>
<li>
<p>Fundraising platforms we tried had various problems:</p>
<ol>
<li>
<p>Betterplace makes it hard to reach donors again, in particular personally, and allows for little formatting, which makes instructions hard to follow.</p>
</li>
<li>
<p><span class="caps">AMF</span>’s fundraiser system is hard to use – you have to fill in a lot of data redundantly, and the WorldPay interface is highly confusing. We probably lost some donors who couldn’t figure it out.</p>
</li>
<li>
<p>CauseVox costs money in addition to the transaction fees. (This is our favored solution at this point.)</p>
</li>
</ol>
</li>
<li>
<p>We used to set a single deadline, but then many important contributors who were working on tracks missed it, and everyone who had submitted in time got impatient while we waited for the belated tracks. The musicians’ tardiness became our tardiness and blemished our image and standing among the musicians.</p>
</li>
<li>
<p>We relied heavily on a rather undiversified musician community that organized through one online forum. When the forum saw an exodus of talented musicians, we lost an important channel for reaching many of their kind.</p>
</li>
<li>
<p>Musicians often asked for their tracks to be freely available rather than just for a donation. The idea was that people would donate voluntarily since it had not been hard to circumvent our “security” all along. When we made them free (which was also necessitated by some technical limitations), other musicians complained about that decision.</p>
</li>
</ol>
<h2 id="album-premiere-livestreams"><a class="toclink" href="#album-premiere-livestreams">Album Premiere Livestreams</a></h2>
<p>We experimented with four livestream formats: our album premiere livestreams, art livestreams, panel livestreams, and gaming livestreams. This has been a hard one to get right, so we experimented with a few different formats.</p>
<h3 id="resources_1"><a class="toclink" href="#resources_1">Resources</a></h3>
<ol>
<li>
<p>Two or more friends (who can work together) with experience in livestreaming audio to several hundred listeners (or however many you expect).</p>
</li>
<li>
<p>A playlist of all tracks of the album as well as the tracks themselves.</p>
</li>
<li>
<p>Moderators for the listener chat.</p>
</li>
<li>
<p>Contact to all musicians.</p>
</li>
<li>
<p>A single page that integrates player, chat, and a donation button (ideally also a donation counter).</p>
</li>
</ol>
<h3 id="timeline"><a class="toclink" href="#timeline">Timeline</a></h3>
<ol>
<li>
<p>We usually started planning these streams rather late, maybe two weeks in advance, because we were busy with the album itself. In 2012, we still had interview guests at the premiere streams, but that would’ve become too much work the other years.</p>
</li>
<li>
<p>First we decided on the format of the stream:</p>
<ol>
<li>
<p>Do we want to prepare everything – intro, outro, and playlist?</p>
<ol>
<li>
<p>If we need to have several stations play the identical stream to avoid favoring one, then this option helps to keep it roughly in sync.</p>
</li>
<li>
<p>In this case, we can’t also have interview guests.</p>
</li>
</ol>
</li>
<li>
<p>Do we want to leave it to a moderator to introduce tracks and ask us to read the intro and outro live?</p>
<ol>
<li>This makes the stream a lot more lively. It’s possible when there are no competing radio stations, when they all work together well, or when we have streamers who are only associated with us.</li>
</ol>
</li>
</ol>
</li>
<li>
<p>We set up a website on a reliable server that can handle at least a hundred requests per second with ease. (In our case just a low-end <span class="caps">VPS</span> and static files served through Nginx.)</p>
<ol>
<li>Ours had the advantage that it would update automatically for all viewers when we changed something on the server, e.g., because we had to switch to another streamer. (Contact me if you need a copy.)</li>
</ol>
</li>
<li>
<p>Set a date and time that works for your prospective streamers and radio stations and is situated such that your core demographic is not at work during the stream.</p>
</li>
<li>
<p>Find people who are familiar with <span class="caps">IRC</span> moderation, register an <span class="caps">IRC</span> channel, set up chanserv to give op to the moderators and possibly voice to the musicians, set the topic with a link to the viewer page, and embed a client in the viewer page.</p>
<ol>
<li>A display of the name and artist of the current track is valuable.</li>
</ol>
</li>
<li>
<p>Find at least two streamer’s of whom one is ready to be the hot standby for the other in case something breaks. Embed the audio or video stream in the viewer page.</p>
<ol>
<li>
<p>We’ve had many internet connections failing and routers dying on our streamers during the streams. This is not just some rare contingency!</p>
</li>
<li>
<p>If you went with 2.a. and have several radio stations embedded in the page, viewers can easily switch between them if one has an outage, so this is less of a concern.</p>
</li>
</ol>
</li>
<li>
<p>Model the timeline of the stream in a spreadsheet (e.g., Google Sheets) so that everyone is clear on when who is responsible for moderation and streaming, and when interview guests are tuning in. This is probably less necessary for 1–2-hour album premiere streams, since one person (each) can fill each role for the whole time.</p>
</li>
<li>
<p>Send announcements to the news sites about one week out and the day before. You can send them earlier and ask the sites to schedule them for roughly that time so that they at least see them in time.</p>
<ol>
<li>
<p>Social media may help a little too, especially if all the musicians share/retweet/reblog you.</p>
</li>
<li>
<p>The announcements should ideally already contain the preview if it can be finished in time.</p>
</li>
</ol>
</li>
<li>
<p>Prepare the playlist and intro and outro if necessary.</p>
</li>
<li>
<p>During the stream, the artists can congratulate each other in the <span class="caps">IRC</span> chat and geek out about their methods and references.</p>
</li>
</ol>
<h2 id="art-livestreams"><a class="toclink" href="#art-livestreams">Art Livestreams</a></h2>
<p>We observed that there were two things that were popular in the fandom: artists livestreaming the creation of requested artwork and interviews (with moderated questions from the audience) with the professionals working on the show. So we combined the two. Our Kallisti livestreams were 24 hours long, enough time for interviews with some 12 guests.</p>
<h3 id="resources_2"><a class="toclink" href="#resources_2">Resources</a></h3>
<ol>
<li>
<p>Two or more friends (who can work together) with experience in livestreaming audio to several hundred listeners (or however many you expect)</p>
</li>
<li>
<p><span class="dquo">“</span>Voice” moderators or hosts of the stream.</p>
</li>
<li>
<p><span class="dquo">“</span>Text” moderators or “<span class="caps">IRC</span> envoys” for the listener chat.</p>
</li>
<li>
<p>A single page that integrates player, chat, and a donation button (ideally also a donation counter).</p>
</li>
</ol>
<h3 id="process_1"><a class="toclink" href="#process_1">Process</a></h3>
<ol>
<li>
<p>It has taken us two months and more to organize these. We discussed dates with our streamers first. Since we had 24 hours for interviews, it was easier to find time slots for our interview guests. It simplifies the process to set the date and rough timeframe among organizers only.</p>
</li>
<li>
<p>First we prepared a schedule (e.g., as Google Sheet). Our Kallisti <span class="caps">IV</span> one had the following columns: Time (<span class="caps">PDT</span>), Time (<span class="caps">EDT</span>), Time (<span class="caps">UTC</span>), Time (<span class="caps">UTC</span>+2), Interview Guests, Hosts, Streamers, <span class="caps">IRC</span> Envoys, Livestreaming Artists, Events. In it we noted:</p>
<ol>
<li>
<p>the times, as redundant and unambiguous as possible,</p>
</li>
<li>
<p>the interview guests and what special requirements they have (e.g., only one interviewer or special technical equipment to only reveal their number or handle to one person),</p>
<ol>
<li>we may distinguish community guests and show guests, so that we can schedule a community guest we trust to support our host in an interview,</li>
</ol>
</li>
<li>
<p>the hosts,</p>
</li>
<li>
<p>the streamers, main streamer and hot standby,</p>
</li>
<li>
<p>the <span class="caps">IRC</span> moderators,</p>
</li>
<li>
<p>the artists livestreaming at the time,</p>
</li>
<li>
<p>and any miscellaneous events, such as auctions or announcements.</p>
</li>
</ol>
</li>
<li>
<p>Some of the artists will be more involved than others, so that we could ask some of the more involved artists for artwork for press releases and the viewer page.</p>
</li>
<li>
<p>We typically sent two announcements to the news sites, one about two weeks out and one a few days ahead of the stream.</p>
<ol>
<li>
<p>We accompanied the announcements with an email text asking them to post it at a time when people are likely to read it and post it before the date of the stream, and we also included links to our embedded images in case of rare technical problems where the images got filtered by some security or privacy mechanism.</p>
</li>
<li>
<p>We put USPs in the title if at all possible, e.g., famous interview guests, well-known artists, etc.</p>
</li>
<li>
<p>We put important things like the date, the link to the viewer page, and the names of the interview guests in bold text.</p>
</li>
<li>
<p>We also included the times when the interview guests would join, because someone may only be interested in some of them, and rather than watching all 24 hours, rather not watch it at all.</p>
</li>
<li>
<p>We always named and credited (with link to their portfolio) the contributing artists.</p>
</li>
<li>
<p><a href="http://broniesforgood.org/2014/04/final-notice-the-kallisti-iv-charity-livestream/">Our Kallisti <span class="caps">IV</span> announcement</a> as example.</p>
</li>
</ol>
</li>
<li>
<p>The combined artistic streams:</p>
<ol>
<li>
<p>We and our partner convention already had contact to many artists. If we hadn’t had those, we’d have had to reach out to them. In the case of Kallisti <span class="caps">IV</span>, we had 23 artists signed up and up to 15 drawing in parallel.</p>
</li>
<li>
<p>We needed a whole bunch of streamers to ensure that someone was available at all times. It has happened that in the end enough routers caved in that one streamer had to stay awake some 48 hours to keep the stream running, but that is not optimal.</p>
</li>
<li>
<p>The artists streamed to each their own channels but asked their viewers to go to our page instead. The streamers tended to use a paid version of XSplit to capture these videos, usually via screen capture. (The APIs seemed to be much less reliable.)</p>
<ol>
<li>Running 10+ video streams in parallel, screen-capturing them, and encoding them all into one video stream takes many large screens and fast GPUs and CPUs. Stress-testing the equipment beforehand is important.</li>
</ol>
</li>
<li>
<p>In the week before the stream, we did a dress rehearsal of sorts. This is relatively easily done, since the streamer can just pick random streams and combine them while chatting with others on the team. All the while, one person can watch and listen to the end product on the public stream and notify the team of any problems.</p>
</li>
<li>
<p>People will make art requests during the stream, which need to be recorded by the chat moderators and others in a shared spreadsheet. Our last one had the following columns:</p>
<ol>
<li>
<p>Status (pending, reserved, in progress, or done), Artist, Character(s) to Draw, What to Draw, Appearance, Shipping, Art Style, Theme, Fandom, Requested By, Special Guest, Other Comments (like references), Link to Finished Work, Request No., Copypasta Text (the exact text from the chat), and various columns to keep track of whether we have a backup, where it still needs to be published, etc.</p>
</li>
<li>
<p>On Kallisti <span class="caps">IV</span>, we received 275 requests of which over 100 were completed.</p>
</li>
</ol>
</li>
</ol>
</li>
<li>
<p>The interviews:</p>
<ol>
<li>
<p>It was always necessary for us to start really early with this, about two months before the stream.</p>
</li>
<li>
<p>Our initial emails had a structure like this:</p>
<ol>
<li>
<p>Remind the potential guest of who we are.</p>
</li>
<li>
<p>Give them details on the format (audio-only).</p>
</li>
<li>
<p>Assure them that we filter questions, will not ask for potentially NDAed information, and can accommodate special requests.</p>
</li>
<li>
<p>Motivate them telling them about the impact of past streams.</p>
</li>
<li>
<p>Tell them where the proceeds from this stream are going.</p>
</li>
<li>
<p>Ask them if they want to participate, and if so, when they are available.</p>
</li>
<li>
<p>Keeping the work they need to do to a minimum (e.g., looking up things in their calendars) and keeping the emails short is probably advantageous</p>
</li>
</ol>
</li>
<li>
<p>We typically had a longer back and forth of emails to arrange details and often had to wait about a week for a reply. Hence the need for the liberal time allotment.</p>
</li>
<li>
<p>If possible, we tried to go through their agents to speed up the scheduling. They have highly organized agents so they don’t need to be highly organized themselves. So talking to the agents can be much more efficient. The agents, however, may be more or less altruistically minded than the guests themselves, so we fall back on direct emails when we don’t get a reply.</p>
</li>
<li>
<p>Some data from Kallisti <span class="caps">IV</span>:</p>
<ol>
<li>
<p>we sent out 26 interview invitations,</p>
</li>
<li>
<p>got 14 responses,</p>
</li>
<li>
<p>among them 9 positive and 4 negative responses, and</p>
</li>
<li>
<p>one maybe eventually did not work out.</p>
</li>
<li>
<p>In total, we sent 133 emails.</p>
</li>
</ol>
</li>
</ol>
</li>
<li>
<p>We often used the time between interviews to conduct little auctions in the chat, live singing (there are slight delays on all sides of the call, so can’t make music together very well), and general jolliness.</p>
</li>
<li>
<p>We live-tweeted the stream to the best of our ability and directed tweets at some multipliers in the fandom. We may’ve done the same on other social media. Some of the news sites were also ready to post reminders during the event. These have the advantages that people can tune in spontaneously without having to remember to do so at a later point.</p>
</li>
<li>
<p>Finally, we always published the gallery of all the scores of images that had been drawn during the stream, and often also video recordings of the interviews.</p>
</li>
</ol>
<h3 id="learnings_1"><a class="toclink" href="#learnings_1">Learnings</a></h3>
<ol>
<li>
<p>One mistake that we didn’t do was screw up communication about dates and times. Use <span class="caps">ISO</span> 8601 dates and <span class="caps">UTC</span> at all times to avoid confusion! (Always writing out the name of the month is also a viable hack.)</p>
<ol>
<li>You can additionally convert the times to some common timezones, but bear in mind that daylight saving time starts and end at different dates in different countries.</li>
</ol>
</li>
<li>
<p>We cannot have a guest of honor waiting for their turn to be interviewed during a livestream, and we would’ve felt obliged to discuss it with our guests if two of them might meet on air. Once we had learned this lesson, we left a lot of time between interviews, so that one guest can be late and we can overdraw the scheduled time for the interview without the next guest noticing anything.</p>
</li>
<li>
<p>Were were always very careful to have knowledgeable people conduct the interview and review all questions that the audience submitted:</p>
<ol>
<li>
<p>Otherwise interview guests may get asked questions they cannot answer for legal reasons; if this happened repeatedly, they’d reconsider whether they want to be interviewed by us again;</p>
</li>
<li>
<p>they may get annoyed by being asked the same questions over and over or by being asked dumb questions (imagine a voice actor being asked for the reasoning behind a plot line), or</p>
</li>
<li>
<p>they may see unseemly questions.</p>
</li>
</ol>
</li>
<li>
<p>It’s hard to schedule interviews for short streams because interview guests first have to be ready to get interviewed at all and on top of that have be available during a specific time window. When we wanted to have two guests at the same time, we asked them for their availability and then exploited any overlap we could find. Doodle would be perfect for that, but we were afraid it might come off as disrespectful.</p>
</li>
<li>
<p>It’s useful to have a few people of the team watch and listen to the public stream rather than the call, so that they can catch problem quickly when they arise.</p>
</li>
</ol>
<h2 id="panel-livestreams"><a class="toclink" href="#panel-livestreams">Panel Livestreams</a></h2>
<p>We started doing panel livestreams – Charity Chitchat 1 and 2 – as an alternative to the art livestreams that would be easier to organize. I will only mention in what aspects they are different rather than repeating much of the timeline above.</p>
<p>They need a lot less lead time, perhaps a month or so, since we only need two guests for them, but we wanted to pick two guests who would be interesting to have together – two guests, who we thought would enjoy each other’s company, have similar interests, etc. Therefore we selected and ranked several pairings and tried to get in touch with them.</p>
<p>In one case, our top pairing fell through, and we fell back on our second option, two people we had more rapport with and were more sure that they would agree if they have time. If only one of the people of the first pairing had agreed, we’d have to get creative, since we can’t possibly turn them down.</p>
<p>In each case, we proposed several (perhaps too many) possible dates to choose from, so that the guests could basically Doodle out when the stream would happen. Only when they had agreed on a common date did we announce the event.</p>
<p>While the shorter duration makes it much easier to organize, it is also less time for donations to come in, and the lack of the artistic component also diminishes the size of the audience that is interested in attending. Hence the returns have been consistently low.</p>
<h2 id="gaming-livestreams"><a class="toclink" href="#gaming-livestreams">Gaming Livestreams</a></h2>
<p>I was not involved in the organization of either of the gaming livestreams, and being swamped with something both times, was happy I didn’t have to think about it. I think the idea was that well-known extroverts from the fandom played and streamed a video game for a long time. It would get boring for the audience after a while, but they would not switch to a different game before a donation threshold was reached.</p>
<p>Neither of the streams was a really resounding success, unfortunately.</p>
<h2 id="comic-books"><a class="toclink" href="#comic-books">Comic Books</a></h2>
<p>We’ve only published one comic book, so this section is less of a composite than the others. The comic book is an anthology, to which each contributing artist added a comic of several pages or an individual image, each self-contained.</p>
<h3 id="resources_3"><a class="toclink" href="#resources_3">Resources</a></h3>
<ol>
<li>
<p>A publishing professional.</p>
</li>
<li>
<p>An art director.</p>
</li>
<li>
<p>Means of selling the book.</p>
</li>
</ol>
<h3 id="process_2"><a class="toclink" href="#process_2">Process</a></h3>
<ol>
<li>
<p>A cofounder of a partner convention conceived of the idea around August/September 2014; we started accepting preorders January to March 2015, and shipped the books starting April 2015.</p>
</li>
<li>
<p>Our art director (who is also a publishing professional) invited artists personally and vetted our nominations keep the art quality consistently high without having to turn down anyone.</p>
<ol>
<li>Alternatively we could’ve opened applications publically and picked the top artists, but that may’ve caused misgiving among the artists that we didn’t pick.</li>
</ol>
</li>
<li>
<p>We originally underestimated the initial costs by roughly half. Even so, it was beyond our combined donation budgets, and loaning the project private funds seemed more complicated than the alternative of preorders.</p>
<ol>
<li>There was probably no alternative to using something like PayPal for preorders since the money needed to remain at our disposal.</li>
</ol>
</li>
<li>
<p>We put together a Skype group chat to stay in touch with the artists and for them to get in touch with each other for collaborations.</p>
<ol>
<li>We tried using a forum at first, but the hurdle to posting must’ve been too high.</li>
</ol>
</li>
<li>
<p>We selected the printing company based on many factors such as price, location, and technology.</p>
<ol>
<li>
<p>The art director greatly preferred offset over laser, so that we couldn’t expect to get proofs. Choosing a printer in the state where the art director lived had some tax advantages and allowed him to have an eye on the printing process.</p>
</li>
<li>
<p>It was also meant to save shipping costs, but that didn’t work out because we underestimated demand in Europe and overestimated demand in the <span class="caps">US</span>.</p>
</li>
</ol>
</li>
<li>
<p>We first served the preorders, then sold the book at conventions, and eventually opened online sales.</p>
<ol>
<li>
<p>To keep an overview of the sales, we used a spreadsheet with the following columns: Date, Buyer’s Name, Quantity, Total Paid, Country, State, Comments / Notes, Status (different pending states and “shipped”), Total Sale (calculated), Shipping Paid (calculated), Postage Cost, Shipping Costs, Paypal Costs, Book Cost (calculated), Total Costs (calculated), Net Profit (calculated), Profit per Book (calculated), Cost per Book (calculated).</p>
</li>
<li>
<p>We shipped from the <span class="caps">US</span> and Germany depending on which was cheaper.</p>
</li>
</ol>
</li>
</ol>
<h3 id="learnings_2"><a class="toclink" href="#learnings_2">Learnings</a></h3>
<ol>
<li>I think we well trusting our art director on most decisions. When we compared our motivations, it turned out that he was terminally interested, first, in creating a beautiful work of art, and second in raising donations. The BfG team was either primarily or exclusively interested in raising donations.</li>
</ol>
<p>When it came to deciding on the number of pages in the comic book, I felt that the marginal value of extra pages would scale more slowly than the price (which already scaled sublinearly). Our art director was adamant that a sizable number of pages was necessary to meet expectation or not to disappoint. I had very low confidence in my hunch, so I trusted the expert despite his slightly different goals. I would generally recommend this approach.
2. We planned to have the books in time for a large <span class="caps">US</span> convention but failed largely because of one artist who overdrew the deadline greatly (and others to a lesser degree). This was probably a costly failure. Our art director also concluded that we should have invited more artists so that we’re less dependent on any tardy ones.</p>
<ol>
<li>
<p>We should’ve ordered some sort of coating for the covers because they rub off.</p>
</li>
<li>
<p>We tried to suggest an <span class="caps">EA</span>-related theme for the comics, but none of the artists used it. It may need to be figured out in collaboration with the artists.</p>
</li>
<li>
<p>Since we needed to use the preorder funds to pay for the printing and initial shipping, we couldn’t have people transfer them to the charity directly, e.g., through something like CauseVox. Instead we used PayPal, even though we knew that it was known for freezing accounts. PayPal froze our account luckily some time after the printing was paid for. We stopped online sales, and a half-year bureaucratic back-and-forth ensued. Eventually, we were allowed to transfer to <span class="caps">AMF</span> again.</p>
</li>
</ol>
<p>We continued using PayPal after the preorder phase because we wanted to keep our options open for creating a second <em>Bound Together</em> without having to take preorders. In hindsight, it would’ve been better to switch to CauseVox at that point.</p>
<p><a href="https://claviger.net/lessons-from-fandom-fundraising-recommendations.html">The third part</a> in this series gives recommendations on how one might try to reproduce our success or preferably exceed it.</p>Fandom Fundraising I: Chronology2017-02-05T12:00:00+00:002017-02-05T12:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2017-02-05:/lessons-from-fandom-fundraising-chronology.html<p>From 2011 to 2015, I’ve been involved in charity fundraising efforts that raised over $300,000 for several charities. I hope others can draw on some of the experiences documented here to repeat this success. This is the first of three articles and gives a chronological overview of our work.</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#introduction">Introduction</a><ul>
<li><a href="#motivation">Motivation</a></li>
<li><a href="#summary">Summary</a></li>
</ul>
</li>
<li><a href="#chronological-overview">Chronological Overview</a><ul>
<li><a href="#mid-to-late-2011">Mid to Late 2011</a></li>
<li><a href="#2012">2012</a></li>
<li><a href="#2013">2013</a></li>
<li><a href="#2014">2014</a></li>
<li><a href="#2015">2015</a></li>
<li><a href="#2016">2016</a></li>
</ul>
</li>
</ul>
</div>
<h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<h3 id="motivation"><a class="toclink" href="#motivation">Motivation</a></h3>
<p>I’ve long been driven by a need to do something with my life that I might, at its conclusion, reflect on thinking that it has made a difference that I have lived. I’ve also been compelled by the overwhelming suffering that I saw wherever I looked. Having dabbled in several other forms of activism, I found that my work for the poverty-focused charity <a href="https://yoursiblings.org">Your Siblings</a> and for <a href="http://broniesforgood.org">Bronies for Good</a>, the group this article series is about, was a particularly good fit for my skill profile.</p>
<p>Your Siblings introduced me to the burden of responsibility that comes with having to decide which project to support with the little funding we could raise and thus the burden of responsibility to decide who gets to live, and who, in turn, does not.</p>
<p>At Bronies for Good, we tried to give our donors all the information that has led us to selecting a particular charity or project to support so that our donors may form their own opinion. But often we found that they were ready to trust our recommendation without digging deeper. So the first burden of responsibility was joined by a second: our responsibility for living up to the trust that thousands of people put into us by giving hundreds of thousands of euros to our recommended charities.</p>
<p>We tried hard to do justice to these joint responsibilities. We certainly did better than many, but until 2014 our skill was still greatly limited. Eventually, however, I found out about <a href="https://en.wikipedia.org/wiki/Effective_altruism">effective altruism</a> and all the knowledge – some of it decades older – that people in the movement had amassed. It is well possible that all our years of effort prior to that date had the same positive impact as just a few hundred euros donated to the charity we support today.</p>
<h3 id="summary"><a class="toclink" href="#summary">Summary</a></h3>
<p>Starting in 2011, we have raised about €238,000 on a budget of less than €1,000 per person per year, less than €500 in my case. Our team consisted of about six core members at its peak with many others helping out in various special functions.</p>
<p>The budget is a rough extrapolation only from the expenses I noted as associated with the fundraising over the years and does not include such things as my cost of living that I would’ve had to pay for in any case. The funding came from my part time job as software developer.</p>
<p>There is a small chance that I may have completed my degree in computer science more quickly if I hadn’t engaged in these activities, incurring a cost in lost wages, but I think the chance is low due to the semester-wise inelasticity of the degree. Any remaining expected cost is probably offset by the experience I gained in the process.</p>
<p>The €238,000 is the aggregate of activities that thousands of donors, hundreds of musicians, scores of artists, and many other people have contributed to besides our core team. Most of it is the direct product of campaigns that we organized or co-organized; at most around 1% stems from fundraisers we merely inspired.</p>
<p>I think that the vast majority of the donations would have gone to much less effective organizations or nonaltruistic ends if we hadn’t influenced them – based on the perhaps roughly <a href="http://mdickens.me/2016/04/26/on_priors/">log-normal distribution</a> of cost-effectiveness across interventions; that few people in the community who I talked to had learned about effective altruism at all, through other channels than us, or even understood basic concepts; and that we inspired a number of people to donate for the first time in their lives based on their comments on our online fundraisers. Donations that are not counterfactually valid are donations that I made to our own fundraisers, again around 1–2% at most.</p>
<ol>
<li>
<p>The following section gives a chronological overview of our activities.</p>
</li>
<li>
<p><a href="https://claviger.net/lessons-from-fandom-fundraising-strategies.html">The second part</a> describes how we organized our various events and is mostly interesting for someone who wants to see if there is something they can borrow from our experiences.</p>
</li>
<li>
<p><a href="https://claviger.net/lessons-from-fandom-fundraising-recommendations.html">The third part</a> highlights our learnings and makes recommendations as to how I think projects like these might be replicated in the future.</p>
</li>
</ol>
<h2 id="chronological-overview"><a class="toclink" href="#chronological-overview">Chronological Overview</a></h2>
<p><img alt="The aggregate donations of our online and offline fundraising. Please disregard the downward turn of the trendline. A downward turn like that is impossible because the graph shows the aggregate." src="/images/lessons-from-fandom-fundraising-chronology/online-and-offline-donations.png"></p>
<p>I’ve written this chronology mostly from memory, aided by <a href="http://broniesforgood.org">our blog posts</a>, chat logs, and internal documents. Unfortunately, I don’t have statistics on everything that we have done. In cases where I have none, I would have to reconstruct the data from the chat logs, old email notifications, and third-party websites. If you have a particular interest in certain data, I can put in more effort to reconstruct them.</p>
<h3 id="mid-to-late-2011"><a class="toclink" href="#mid-to-late-2011">Mid to Late 2011</a></h3>
<p>In summer 2011, I became enamored of the <span class="caps">TV</span> series <em>My Little Pony: Friendship is Magic</em> that had premiered on October 10, 2010. I was studying computer science at the time and ran a poverty-focused charity, Your Siblings, and the show provided me with a welcome distraction from the harrowing knowledge I confronted in the second job.</p>
<p>In summer 2011, Bronies for Good was formed, a group of fans of the show (“bronies”), who wanted to funnel the fandom’s enthusiasm into charitable ends. At the time, I frequently read about charity-related efforts in the fandom, but I usually read about each once and then never again. Bronies for Good stood out to me, because I read about them with some consistency over the later months of 2011.</p>
<p>It seemed to me that the show selected strongly for sensitive and altruistic people. It also inspired great enthusiasm, so in December 2011, I ran a little marketing experiment where I invited artists to submit pony-themed art that conveyed themes of compassion, altruism, charity, etc. On the day the Equestria Daily news blog posted my quasi press release, my organization’s website received over 1,500 page views (as opposed to the usual 50). It was an order of magnitude less on the second day, for a roughly log-normal-looking distribution. <a href="http://yoursiblings.org/blog/2011/11/30/christmas-campaign-2011/">I received 54 submissions</a>. In many cases, unfortunately, the link to the theme was not apparent, and no one got sufficiently interested in my organization to ask about our work or donate. As my campaign concluded, I got in touch with Bronies for Good, who welcomed me with open arms.</p>
<p>Bronies for Good was at the time running a <a href="http://web.archive.org/web/20141221014652/http://www.globalgiving.org/fundraisers/6545">fundraiser</a> for the Children’s Cancer Association that had stood out to them on the charity comparison platforms they were aware of. The fundraiser was based on a music album, <em>Smile!</em>, that they had compiled, and reached almost $27,000. Then the platform awarded another €3,000 (a refund of the platform fees, I think) for being the most successful fundraiser of the year or some other time period.</p>
<h3 id="2012"><a class="toclink" href="#2012">2012</a></h3>
<p><img alt="Aggregate donations from Seeds of Kindness 1–4." src="/images/lessons-from-fandom-fundraising-chronology/seeds-of-kindness-donations.png"></p>
<p>I was involved in the conception of the second music album and fundraiser, <a href="http://broniesforgood.org/seeds-of-kindness/">Seeds of Kindness</a> (which I originally proposed to call “Seeds of Truth”), whose proceeds were to go to the projects that my charity supported, in particular a small medical clinic in a rural part of Uganda and close to a refugee camp. (More on the music album strategy in the second part of the series.)</p>
<p>The €7,214 that were still missing to fully fund the clinic were collected within days after the launch of the fundraiser, and it went on to raise over €40,000 in total. The shape has always been a logarithmic one, with most of the donations coming in within the first three months and only a fraction afterwards. An independent group contributed by running their own gaming fundraiser, which raised €581. (More information on that, again, in the second part of the series.)</p>
<p>Soon after, a cover of a song from the show, a collaboration of 11 soloists and over 160 choir singers, was released, the <a href="https://www.youtube.com/watch?v=UaG-JijAr-c">Massive Smile Project</a>. It was only loosely associated with our fundraiser (for legal trademark reasons), but it may have caused another bump in donations. Since donations were still coming in at a high rate at that point anyway, the impact of this release is particularly hard to estimate.</p>
<p>In July, a singer and entertainer well known in the fandom ran a <a href="http://broniesforgood.org/2012/07/mic-and-bfgs-steam-2012-charity-raffle-extravaganza/">fundraiser livestream</a> for us that was nightmarishly organized but attracted almost 1,000 viewers at the peak and washed about €3,700 toward our Seeds of Kindness fundraiser.</p>
<p>The Germany-based European convention <a href="https://www.galacon.eu/">GalaCon</a> was the first to host a <a href="http://broniesforgood.org/2012/08/the-galacon-2012-charity-auction/">charity auction</a> for our charity. It was highly successful, with over €7,500 raised from only just over 300 attendees, and the start of a very impactful cooperation. (More details, again, in the second part of the series.)</p>
<p>The third music album was released, <a href="http://broniesforgood.org/seeds-of-kindness-2/">Seeds of Kindness 2: </a><a href="http://broniesforgood.org/seeds-of-kindness-2/"><em>Faithful and Strong</em></a> (the format is [fundraiser name]: <em>[album title]</em>), and raised over €31,000. After €21,000, my charity’s room for funding was exhausted and we selected another beneficiary – the German branch of Engineers Without Borders – based on our “best-effort” prioritization, and technical and legal feasibility.</p>
<p>November saw the the first incarnation of the <a href="http://broniesforgood.org/2012/11/reminder-kallisti-a-massively-multiartist-charity-livestream/">Kallisti</a> <a href="http://broniesforgood.org/2012/11/kallisti-a-massively-multiartist-charity-livestream-photo-album/">livestream</a>, a 24-hour livestream with artists and interviews with show guests – writers, voice actors, composers, and other artisans who create the <span class="caps">TV</span> show. It raised €1,722. Because it was a collaboration between us and GalaCon, our part amounted to €861.</p>
<h3 id="2013"><a class="toclink" href="#2013">2013</a></h3>
<p><img alt="Sessions of our website visitors." src="/images/lessons-from-fandom-fundraising-chronology/bfg-website-sessions.png"></p>
<p>In January and June we organized two more <a href="http://broniesforgood.org/2013/01/kallisti-ii-art-charity-livestream-gallery/">Kallisti</a> <a href="http://broniesforgood.org/2013/06/kallisti-iii-interviews-with-amy-keating-rogers-andrea-libman-anneli-heed-m-a-larson-and-julie-basecqz/">fundraisers</a> following the same format, which raised €4,950 and €1,885 respectively, of which our charity received the usual 50%.</p>
<p>The fourth music album and fundraiser, <a href="http://broniesforgood.org/2013/08/welcome-to-seeds-of-kindness-3/">Seeds of Kindness 3: </a><a href="http://broniesforgood.org/2013/08/welcome-to-seeds-of-kindness-3/"><em>A Beautiful Heart</em></a>, launched and went on to raise €23,000.</p>
<p><a href="http://broniesforgood.org/2013/08/ys-galacon-charity-cooperation-orphanage-funded/">GalaCon 2013</a> raised €14,254 – the charity auction and a Bronies for Good table combined. At the table, we accepted donations and gave out thank yous to the donors, such as decorative pins or art prints.</p>
<p>There was a <a href="http://broniesforgood.org/2013/11/gaming-for-good-a-brony-gaming-livestream-fundraiser/">reboot</a> of the gaming fundraiser in November, but I was not involved in the organization. It raised around €230 toward Seeds of Kindness 3.</p>
<p>Christmas was marked by the <a href="http://broniesforgood.org/2013/12/charity-double-premiere-dash-of-christmas-and-a-christmas-carol/">premiere of two radio plays</a> that friends of Bronies for Good contributed. One used to be a storybook with images accompanying the narration and dialog, but Hasbro intervened and banned the project from being released in its audio-visual form for trademark reasons. Again I find it hard to reconstruct the bump in donations around the time, but I think it was not proportional to the effort that had gone into the projects and most likely below €700.</p>
<p>Shortly after Christmas, another independent project – a <a href="http://broniesforgood.org/2013/12/top-50-pony-music-countdown-2013-charity-event/">top fifty</a> of the best songs of 2013 according to a community vote – encouraged donations toward our fundraiser. The returns were part of Seeds of Kindness 3, and the bump is hard to reconstruct today – probably less than €400.</p>
<p>Despite the still sizable success of Seeds of Kindness 3, it was evident that the enthusiasm of the fandom at large was on a downward trajectory. Had we run Seeds of Kindness 3 with the same lackluster technical know-how, marketing, connections, brand awareness, etc. that we had at the time of the first Seeds of Kindness, it would’ve surely raised even less.</p>
<p>In the graph at the top of the article it looks as if online fundraising reached a tipping point only around early 2015, but in my experience, this downward turn became evident to us throughout 2013. (Please disregard the downward turn of the trendline. Since the graph shows the aggregate raised, a downward turn like that is impossible. The downward turn that I’m referring to is one of the first derivative.) It became harder and harder to reach people – livestreams that used to peak at 200–300 viewers now peaked at 80–120 viewers despite much more intensive advertisement. Our team increasingly thinned too, mostly due to university degrees and full-time work. As the remaining team members’ efforts increased and donations decreased, we eventually reached a point where we decided that certain types of events were no longer worthwhile. This point was reached about a year after the general downturn became evident to me; hence the delay in the graph.</p>
<h3 id="2014"><a class="toclink" href="#2014">2014</a></h3>
<p>In 2014, friends of the organization ran <a href="http://broniesforgood.org/2014/03/the-ministry-of-brony-sessions-one-charity-album/">their own album project</a> to benefit our fundraiser. The album was more lightly supervised and limited to electronic music while ours had no genre limitations. It raised €1,888.</p>
<p>In February and November we ran livestream events of a few hours – <a href="http://broniesforgood.org/2014/02/charity-chitchat-with-natasha-levinger-and-amy-keating-rogers/">Charity Chitchat 1</a> and <a href="http://broniesforgood.org/2014/11/charity-chitchat-2-donation-matching-with-nicole-oliver-and-ingrid-nilson/">Charity Chitchat 2</a> – whose unique feature was that we had two guests of honor for interviews, so that we could host something like a panel discussion. They raised a little over €600 and €700 respectively, counted as part of Seeds of Kindness.</p>
<p>This year we also published an essay <a href="http://broniesforgood.org/2014/02/on-bullying/">on bullying</a> and an audio panel discussion <a href="http://broniesforgood.org/2014/03/podcast-on-feminism-and-the-fandom/">on discrimination</a> against people with intersectional identities. These were topics dear to members of our team that had gained currency in the fandom around the time. I have no data on any impact they may have had on our activities except for one second-hand account that one donor allegedly discontinued their donations to either Your Siblings or the Against Malaria Foundation as a result of the second project. Others may have started donating to them for the same reason for all I know.</p>
<p>The <a href="http://broniesforgood.org/2014/05/kallisti-iv-art-gallery-quality-in-grandeur/">fourth Kallisti livestream</a> raised €3,420 (half of which benefitted our charity) and the fourth album, <a href="http://broniesforgood.org/2015/01/2015-seeds-of-kindness-year-against-malaria/">Seeds of Kindness 4: </a><a href="http://broniesforgood.org/2015/01/2015-seeds-of-kindness-year-against-malaria/"><em>Shine Together</em></a>, raised €9,552.</p>
<p>While the last Kallisti livestream was a surprise success, we shared the returns with our convention partner, so that it was only half as successful as it may seem at first glance. Seeing how we could only hope to raise around €1,500 with another Kallisti, how it took about two months to organize, and how our team had shrunk further and was now missing key people, we no longer felt that the returns were worth the effort and risks.</p>
<p>The year 2014 was special in that it was the year when I learned about the movement for effective altruism and thus gained access to a wealth of information on charity prioritization. Effective altruism was everything we’ve been trying to do all along. Now we saw that we’ve been doing a pretty haphazard job at it and corrected course by relying on the research of GiveWell for our charity recommendations.</p>
<p>At the time, the Against Malaria Foundation (<span class="caps">AMF</span>) was still demoted after GiveWell’s 2013 concerns about its scalability, but upon review, we didn’t share GiveWell’s concerns and viewed <span class="caps">AMF</span> as better than or on par with their other recommendations. Technical and legal reasons (further) tipped the balance, so that we decided to finish up all ongoing collaborations we were engaged in (as Your Siblings and Bronies for Good) and then focus on <span class="caps">AMF</span>.</p>
<p>Finally, in December, just in time for the launch-release of Shine Together, <span class="caps">AMF</span> was reinstated as GiveWell top charity. We partied!</p>
<h3 id="2015"><a class="toclink" href="#2015">2015</a></h3>
<p>The year 2015 was the year of the launch-release of <a href="http://broniesforgood.org/2015/11/2016-a-year-for-building-bridges/">Seeds of Kindness</a>: <a href="http://broniesforgood.org/2015/11/2016-a-year-for-building-bridges/"><em>Building Bridges</em></a>, our sixth charity album in the Seeds of Kindness series. Since “Seeds of Kindness <em>N</em>” was always associated with our <em>(N+1)</em>th charity album, which caused confusion, we started to drop the number from the fundraiser name and rather distinguished the fundraisers by the album titles. This Seeds of Kindness “5” has since raised €7,400.</p>
<p>It was also a year full of conventions: <a href="http://broniesforgood.org/2015/04/brony-fair-2015/">Brony Fair 2015</a> (€1,068), <a href="http://broniesforgood.org/2015/08/galacon-2015/">GalaCon 2015</a> (€20,027), BABSCon 2015 (€798), <span class="caps">EQLA</span> 2015 (€839), and <a href="http://broniesforgood.org/2015/09/czequestria-2015/">Czequestria 2015</a> (€3,865). GalaCon and Czquestria have been the most comprehensive and amicable collaborations as they rely on our (and thus GiveWell’s or <span class="caps">ACE</span>’s) recommendation entirely and dedicate their auctions to our charity in addition to offering us a free and strategically placed table at the event. The other conventions only offered a free table.</p>
<p>Eventually we also sent our first comic book to the presses, <a href="http://broniesforgood.org/bound-together/"><em>Bound Together</em></a>. (No <span class="caps">BDSM</span> reference intended.) We ordered 1255 copies (too many) for $6,167.58, funded from 294 preorders. Online we had a turnover of $11,142 minus cost for shipping of $3,380 for a profit of $7,762 for <span class="caps">AMF</span>. The comic book also sold well at European conventions (hardly at all at <span class="caps">US</span> conventions, contrary to our expectation). Adding up the tallies from the European conventions (251 books = €6275), we get a total of €13,644 thanks to <em>Bound Together</em>.</p>
<p>A the end of 2015, I officially took my leave from Bronies for Good and Your Siblings to focus on either full-time direct work or earning to give at other organizations or companies, though I continue to run the European convention division of Bronies for Good.</p>
<h3 id="2016"><a class="toclink" href="#2016">2016</a></h3>
<p>Last year I ran three convention programs for Bronies for Good: <a href="http://broniesforgood.org/2016/04/brony-fair-2016/">Brony Fair 2016</a> (€1,349), <a href="http://broniesforgood.org/2016/06/pony-congress-2016/">Pony Congress 2016</a> (€530), and <a href="http://broniesforgood.org/2016/08/galacon-2016/">GalaCon 2016</a> (€23,019). Pony Congress – next to GalaCon and Czequestia – also counts among the convention that collaborated with us fully, including the charity auction. The convention had too few attendees, though, and was set in Poland, which has very low wages, so that the return was meager.</p>
<p>In early 2016, right after Brony Fair, we agreed to switch our support to Animal Equality to boost our fundraising impact by another two or three orders of magnitude. About half a year later, however, one of us asked for the decision to be reversed for <span class="caps">US</span> fundraising.</p>
<p>We hope to release Seeds of Kindness: <em>A Change of Heart</em> early this year, but I’m hardly involved in the process anymore.</p>
<table>
<tr>
<th>Source</th>
<th>Total €</th>
<th>Total $</th>
</tr>
<tr>
<td>Smile</td>
<td>€ 23,163</td>
<td>$30,000</td>
</tr>
<tr>
<td>Seeds of Kindness</td>
<td>€ 40,887</td>
<td>$52,831</td>
</tr>
<tr>
<td>Elements of Charity</td>
<td>€ 581</td>
<td>$769</td>
</tr>
<tr>
<td>GalaCon 2012</td>
<td>€ 7,545</td>
<td>$9,273</td>
</tr>
<tr>
<td>Seeds of Kindness 2</td>
<td>€ 31,146</td>
<td>$40,588</td>
</tr>
<tr>
<td>Kallisti</td>
<td>€ 1,722</td>
<td>$2,201</td>
</tr>
<tr>
<td>Kallisti <span class="caps">II</span></td>
<td>€ 4,950</td>
<td>$6,616</td>
</tr>
<tr>
<td>Kallisti <span class="caps">III</span></td>
<td>€ 1,885</td>
<td>$2,419</td>
</tr>
<tr>
<td>GalaCon 2013</td>
<td>€ 14,254</td>
<td>$18,831</td>
</tr>
<tr>
<td>Seeds of Kindness 3</td>
<td>€ 23,005</td>
<td>$30,802</td>
</tr>
<tr>
<td>Ministry of Brony</td>
<td>€ 1,888</td>
<td>$2,597</td>
</tr>
<tr>
<td>Kallisti <span class="caps">IV</span></td>
<td>€ 3,420</td>
<td>$4,744</td>
</tr>
<tr>
<td>GalaCon 2014</td>
<td>€ 12,378</td>
<td>$16,621</td>
</tr>
<tr>
<td>Stable-Tec GalaCon</td>
<td>€ 915</td>
<td>$1,185</td>
</tr>
<tr>
<td>Czequestria 2014</td>
<td>€ 2,012</td>
<td>$2,606</td>
</tr>
<tr>
<td>Seeds of Kindness 4</td>
<td>€ 9,552</td>
<td>$11,682</td>
</tr>
<tr>
<td>Brony Fair 2015</td>
<td>€ 1,068</td>
<td>$1,163</td>
</tr>
<tr>
<td>BABSCon 2015</td>
<td>€ 798</td>
<td>$890</td>
</tr>
<tr>
<td>GalaCon 2015</td>
<td>€ 20,027</td>
<td>$22,051</td>
</tr>
<tr>
<td>Czequestria 2015</td>
<td>€ 3,865</td>
<td>$4,321</td>
</tr>
<tr>
<td><span class="caps">EQLA</span> 2015</td>
<td>€ 839</td>
<td>$940</td>
</tr>
<tr>
<td>Seeds of Kindness 5</td>
<td>€ 7,406</td>
<td>$8,128</td>
</tr>
<tr>
<td>Brony Fair 2016</td>
<td>€ 1,349</td>
<td>$1,539</td>
</tr>
<tr>
<td>Pony Congress 2016</td>
<td>€ 530</td>
<td>$712</td>
</tr>
<tr>
<td>GalaCon 2016</td>
<td>€ 23,019</td>
<td>$30,908</td>
</tr>
<tr>
<td></td>
<td></td>
<td></td>
</tr>
<tr>
<td>Totals</td>
<td>€ 238,204</td>
<td>$304,415</td>
</tr>
</table>
<p><a href="https://claviger.net/lessons-from-fandom-fundraising-strategies.html">The second part</a> of this series dives into detail on how we organized all these events.</p>Estimating the Harm North Korea Inflicts on its Citizens2016-08-27T12:00:00+00:002016-08-27T12:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2016-08-27:/estimating-the-harm-north-korea-inflicts-on-its-citizens.html<p>A quantitative analysis – using Guesstimate – of the harm the North Korean government inflicts on its general population and its prisoner population, and a comparison to the harm from malaria in Mozambique and Angola.</p><div id="pec-encrypted-content" style="display:none">E3vGoqEY4PGZtv5XCHkYhg==;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;^</div>
<div id="pec-decrypted-content">
<h4><i></i></h4>
</div>
<form id="pec-decrypt-form">
<p>
Old articles can be embarrassing. If you would like to access them anyway, you may
<a href="https://bit.ly/3jPN8tT" target="_blank">request access here</a>. Please indicate
who you are in case I don’t know you or don’t recognize you from your email address.
</p>
<input type="password" id="pec-content-password" placeholder="Password" />
<button type="submit" id="pec-decrypt-content">Decrypt</button>
</form>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/core.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/enc-base64.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/cipher-core.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/pad-nopadding.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/md5.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/aes.js"></script>
<script type="text/javascript">
(function () {
var strip_padding = function (padded_content, padding_char) {
/*
* Strips the padding character from decrypted content.
*/
for (var i = padded_content.length; i > 0; i--) {
if (padded_content[i - 1] !== padding_char) {
return padded_content.slice(0, i);
}
}
};
var decrypt_content = function (password, iv_b64, ciphertext_b64, padding_char) {
/*
* Decrypts the content from the ciphertext bundle.
*/
var key = CryptoJS.MD5(password),
iv = CryptoJS.enc.Base64.parse(iv_b64),
ciphertext = CryptoJS.enc.Base64.parse(ciphertext_b64),
bundle = {
key: key,
iv: iv,
ciphertext: ciphertext
};
var plaintext = CryptoJS.AES.decrypt(bundle, key, { iv: iv, padding: CryptoJS.pad.NoPadding });
try {
return strip_padding(plaintext.toString(CryptoJS.enc.Utf8), padding_char);
}
catch (err) {
// encoding failed; wrong password
console.log(err);
return false;
}
};
var init_decryptor = function () {
var decrypt_btn = document.getElementById('pec-decrypt-content'),
password_input = document.getElementById('pec-content-password'),
encrypted_content = document.getElementById('pec-encrypted-content'),
decrypted_content = document.getElementById('pec-decrypted-content'),
decrypt_form = document.getElementById('pec-decrypt-form');
decrypt_btn.addEventListener('click', function () {
// grab the ciphertext bundle
var parts = encrypted_content.innerHTML.split(';');
// decrypt it
var content = decrypt_content(
password_input.value,
parts[0],
parts[1],
parts[2]
);
if (content) {
// success; display the decrypted content
decrypted_content.innerHTML = content;
decrypt_form.parentNode.removeChild(decrypt_form);
encrypted_content.parentNode.removeChild(encrypted_content);
// any post processing on the decrypted content should be done here
}
else {
// ¯\_(ツ)_/¯
password_input.value = '';
}
});
};
document.addEventListener('DOMContentLoaded', init_decryptor);
})();
</script>LLIN Visualizer2016-08-18T12:00:00+00:002017-02-21T12:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2016-08-18:/llin-visualizer.html<p>Visualizing distributions of long-lasting insecticide-treated mosquito nets in the epidemiological context. My master’s thesis in computer science at the Department of Computer Science of the Freie Universität Berlin. Please read the <a href="https://claviger.net/assets/masters-thesis.pdf"><span class="caps">PDF</span> version of the thesis</a> as I haven’t fully recreated the formatting in Markdown.</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#executive-summary">Executive Summary</a></li>
<li><a href="#acknowledgements">Acknowledgements</a></li>
<li><a href="#source-code-and-demonstration">Source Code and Demonstration</a></li>
<li><a href="#ch:introduction">Introduction</a></li>
<li><a href="#ch:previous-work">Previous Work</a><ul>
<li><a href="#sec:previous-research-altruistic-coordination">Previous Research on Altruistic Coordination</a></li>
<li><a href="#subsec:wtp">Willingness to Pay for Increased Moral Goal Satisfaction</a></li>
<li><a href="#subsec:moral-preferences">Diversity of Moral Preferences</a></li>
<li><a href="#subsec:identifying-moral-preferences">Identifying Moral Preferences</a></li>
<li><a href="#subsec:values-spreading">Values Spreading</a></li>
</ul>
</li>
<li><a href="#correcting-ignorance">Correcting ignorance.</a></li>
<li><a href="#correcting-misinformation">Correcting misinformation.</a><ul>
<li><a href="#subsec:trust">Trust</a></li>
<li><a href="#sec:previous-research-altruistic-prioritization">Previous Research on Altruistic Prioritization</a></li>
<li><a href="#subsec:finding-best-charities">Finding the Best Charities</a></li>
<li><a href="#subsec:motivations-donors">Motivations of Donors</a></li>
<li><a href="#subsec:market-impact">A Market for Impact</a></li>
<li><a href="#sec:prioritization-research">Prioritization Research</a></li>
<li><a href="#subsec:constraints">Constraints</a></li>
<li><a href="#subsec:methodologies">Methodologies</a><ul>
<li><a href="#subsubsec:government-aid">Government Aid</a></li>
<li><a href="#subsubsec:foundations">Foundations</a></li>
<li><a href="#subsubsec:individual-donors">Individual Donors</a></li>
</ul>
</li>
<li><a href="#sec:strategy-amf">Strategy of the Against Malaria Foundation</a></li>
<li><a href="#sec:llin-distributions">Distributions of Long-Lasting Insecticide-Treated Mosquito Nets</a></li>
<li><a href="#subsec:epidemiology">Key Epidemiological Facts</a></li>
<li><a href="#subsec:history">Brief History</a></li>
<li><a href="#subsec:role-amf">Role of the Against Malaria Foundation</a></li>
<li><a href="#ch:requirements-and-modeling">Requirements and Modeling</a><ul>
<li><a href="#sec:abduction">Forming the Idea</a></li>
<li><a href="#sec:data-exploration">Data Exploration</a></li>
<li><a href="#subsec:survey-form">The Survey Form</a></li>
<li><a href="#subsec:survey-data">The Survey Data</a></li>
<li><a href="#subsec:forthcoming-data">Forthcoming Data</a></li>
<li><a href="#subsec:other-data">Other Data</a><ul>
<li><a href="#the-world-map">The World Map</a></li>
</ul>
</li>
<li><a href="#insecticide-resistance">Insecticide Resistance</a></li>
<li><a href="#incidence-and-endemicity">Incidence and Endemicity</a></li>
<li><a href="#net-gaps">Net Gaps</a></li>
<li><a href="#indicators">Indicators</a></li>
<li><a href="#sec:functional-requirements-net-analytics">Functional Requirements</a></li>
<li><a href="#subsec:original-set">The Original Set</a></li>
<li><a href="#subsec:extension-set">The Extension Set</a></li>
<li><a href="#sec:coordination-amf">Coordination with AMF</a></li>
</ul>
</li>
<li><a href="#ch:architecture-and-implementation">Architecture and Implementation</a><ul>
<li><a href="#sec:architecture">Architecture</a></li>
<li><a href="#subsec:models">Models</a></li>
<li><a href="#subsec:controllers">Controllers</a></li>
<li><a href="#subsec:views">Views</a></li>
<li><a href="#subsec:management-commands">Management Commands</a></li>
<li><a href="#sec:technology">Technology</a></li>
<li><a href="#subsec:core-technologies">Core Technologies</a></li>
<li><a href="#subsec:other-important-technologies">Other Important Technologies</a></li>
</ul>
</li>
<li><a href="#ch:evaluation-and-learnings">Evaluation and Learnings</a><ul>
<li><a href="#sec:functional-optimization">Functional Optimization</a></li>
<li><a href="#evaluation">Evaluation</a></li>
<li><a href="#sec:nonfunctional-optimization">Nonfunctional Optimization</a></li>
<li><a href="#evaluation_1">Evaluation</a></li>
<li><a href="#sec:learnings">Learnings</a></li>
<li><a href="#subsec:learnings-dates">Dates</a></li>
<li><a href="#subsec:learnings-health-areas">Health Areas</a></li>
</ul>
</li>
<li><a href="#ch:appendices">Appendices</a><ul>
<li><a href="#app:ima-2014-pdrs">IMA 2014 Predistribution Registration Survey</a></li>
</ul>
</li>
<li><a href="#endnotes">Endnotes</a></li>
<li><a href="#bibliography">Bibliography</a></li>
</ul>
</li>
</ul>
</div>
<h2 class="unnumbered" id="executive-summary"><a class="toclink" href="#executive-summary">Executive Summary</a></h2>
<p>The <span class="caps">LLIN</span> Visualizer makes transparent many of the factors that guide the
prioritization decisions of the Against Malaria Foundation. Thereby it
fosters trust between the foundation – one of the funders of
distributions of long-lasting insecticide-treated bednets for malaria
vector control – and altruistic investors who donate toward the
intervention, and thus addresses a coordination problem that poses a
hurdle toward a more efficient market for altruistic impact.</p>
<h2 class="unnumbered" id="acknowledgements"><a class="toclink" href="#acknowledgements">Acknowledgements</a></h2>
<p>I would like to express my gratitude to Rob Mather and Dr. Andrew Garner
of the Against Malaria Foundation, who supplied me with much data and
answered all my questions about net distributions, and my mentors and
supervisors for their feedback and encouragement. I would also like to
thank all my beta testers, Sara Nowak for her invaluable quality
assurance, Martin Wilson for English copyediting, and Djahbaï Chonville
for French copyediting. Finally I want to thank all that have shaped my
ethical and epistemic worldview, without whom I would not have had the
idea of writing this thesis.</p>
<h2 class="unnumbered" id="source-code-and-demonstration"><a class="toclink" href="#source-code-and-demonstration">Source Code and Demonstration</a></h2>
<dl>
<dt><span class="caps">PDF</span> version.</dt>
<dd>
<p>There is also a <a href="https://claviger.net/assets/masters-thesis.pdf"><span class="caps">PDF</span> version of this thesis</a>.
It potentially contains fewer formatting errors.</p>
</dd>
<dt>Source code.</dt>
<dd>
<p>The complete <span class="caps">LLIN</span> Visualizer project is open source and hosted on Bitbucket.<sup id="fnref:LLINVisualizerCode"><a class="footnote-ref" href="#fn:LLINVisualizerCode">58</a></sup>
At the time of writing, the latest commit was <code>fbdbf2e</code> (July 14, 2016).</p>
</dd>
<dt><span class="caps">LLIN</span> Visualizer demonstration instance.</dt>
<dd>
<p>I am maintaining a demonstration instance of the project on my server at
<a href="http://nets.claviger.net">http://nets.claviger.net/</a>.<sup id="fnref:LLINVisualizerInstance"><a class="footnote-ref" href="#fn:LLINVisualizerInstance">56</a></sup></p>
</dd>
<dt><span class="caps">LLIN</span> Visualizer demonstration video.</dt>
<dd>
<p>I have also produced a video demonstrating the use of the <span class="caps">LLIN</span> Visualizer.<sup id="fnref:LLINVisualizerVideo"><a class="footnote-ref" href="#fn:LLINVisualizerVideo">54</a></sup></p>
</dd>
</dl>
<h2 id="ch:introduction"><a class="toclink" href="#ch:introduction">Introduction</a></h2>
<blockquote>
<p>“Homines ad deos nulla re propius accedunt quam salutem hominibus
dando.” (In nothing do humans more nearly approach the gods than in
giving health to humans.) —Cicero, Pro Ligario, <span class="caps">XII</span></p>
</blockquote>
<p>Just as there is a market for monetary profit – which is an
agent-relative goal since one person’s profit is not automatically
another person’s profit – there could also be a market for altruistic
impact – for maximizing happiness, minimizing suffering, or any other
such agent-neutral goals that people find worthy of optimization. An
organization that manages to make efficient progress toward such a goal
would receive lavish investment until it reaches its limits of
scalability and its marginal cost-effectiveness drops below that of
another organization. Startups in this space would receive such lavish
investment if they promise rapid growth that will render them highly
cost-effective in the long run.</p>
<p>Unfortunately there are a few hurdles left to overcome toward such a
market. Some agents are not or not yet intent on investing in such a
market; failure modes of this category will be addressed briefly below.
They have been subject to research in social psychology and behavioral
economics that is outside the purview of this text. But even agents that
are eager to drive positive change in the world can face problems. Here,
two important ones:</p>
<p>When investors and recipients are two distinct groups, the investors
have an even harder time assessing the impact they have had than they
would have had anyway, so that, in the absence of rigorous impact
evaluations, vast inefficiencies result.</p>
<p>The investments that are most cost-effective in expectation require
highly risk-neutral investors, but the learned heuristics for the
diminishing marginal utility of money for personal investments skew
expectations for altruistic investments.</p>
<p>Only the first of these is subject to the <span class="caps">LLIN</span> Visualizer discussed in
this text. The second is addressed in my article on <a href="/concept-for-donor-coordination.html">donor coordination</a>. So let us consider some of the
inefficiencies – or injustices – that these and other problems allow to
continue unabaded.</p>
<p><img alt="Global income distribution." src="/images/llin-visualizer/global-income-distribution.png"></p>
<p>The organization Giving What We Can has compiled data from three sources
to get an idea of the global income distribution.<sup id="fnref:1"><a class="footnote-ref" href="#fn:1">1</a></sup> According to those
data, a person in Germany with an income of €42,000 is already within
the richest 1% of the world population.</p>
<p>This inequality leads to large numbers of poverty-related
deaths,[^macaskill2015doing 431] a rough lower bound of which can be
estimated at over 30,000 per day.<sup id="fnref:2"><a class="footnote-ref" href="#fn:2">2</a></sup></p>
<p>If just a fraction of the richest in the world redistributed a portion
of their wealth to the poorest – directly and by investing into research
how to alleviate poverty most effectively as well as scaling up the
solutions – this death toll could decrease drastically.</p>
<p>And in fact we are seeing significant flows of international aid and
philanthropy. In 2013, aid and philanthropy totaled \<span class="math">\(197
billion[^adelman2013index 6, 12], while the Brookings Institution
estimates that only \\)</span>66 billion per year can suffice to eliminate
extreme poverty globally.[^chandy2011poverty 13]</p>
<p>MacAskill also estimates that in view of purchasing power
parity, average income, and diminishing marginal returns to wealth, “for
those of us living in rich countries, you should expect to be able to do
at least one hundred times as much to benefit other people as you can to
benefit yourself.”[^macaskill2015doing 55] Yet, using the example of <span class="caps">US</span>
philanthropy in 2014, only 4% of donations were invested outside the
<span class="caps">US</span>.<sup id="fnref:GivingUSA2015"><a class="footnote-ref" href="#fn:GivingUSA2015">74</a></sup> These figures are similar for other developed
countries and would indicate that citizens of a country value their
compatriots about <span class="math">\(\frac{100\%}{4\%} \cdot 100 = 2\,500\)</span> times as highly
as a person in extreme poverty. Still, the guiding mantra of the Bill <span class="amp">&</span>
Melinda Gates Foundation – that every life has equal value – does not
commonly stir up controversy.</p>
<p>There are many reasons why we are seeing this odd allocation of
resources, and most of them are well without the purview of this text,
such as the psychology behind surprising investment decisions that seem
to indicate that some people value a week in space higher than saving
the lives of approximately 9,600 people,<sup id="fnref:BBCTito"><a class="footnote-ref" href="#fn:BBCTito">36</a></sup><sup id="fnref:GiveWellAMF"><a class="footnote-ref" href="#fn:GiveWellAMF">68</a></sup> or that
people value saving one dog over saving approximately 150 dogs or almost
290,000 other animals,<sup id="fnref:WeekHokget"><a class="footnote-ref" href="#fn:WeekHokget">144</a></sup><sup id="fnref:ACEImpactDonations"><a class="footnote-ref" href="#fn:ACEImpactDonations">32</a></sup> to cite just
two examples.<sup id="fnref:3"><a class="footnote-ref" href="#fn:3">3</a></sup></p>
<p>While the specific reasons are manifold, they can be roughly categorized
into three categories:</p>
<dl>
<dt>Moral differences.</dt>
<dd>
<p>The following chapters will give a brief insight into the research
into the different moral preferences people have, as well as the
debate over whether these are differences of taste or of belief.
Usually, these differences are fairly subtle, but that is not to say
that there are not also highly outré moral preferences.<sup id="fnref:4"><a class="footnote-ref" href="#fn:4">4</a></sup></p>
</dd>
<dt>Lacking education.</dt>
<dd>
<p>Educating oneself comes at a price that is usually paid in time.<sup id="fnref:5"><a class="footnote-ref" href="#fn:5">5</a></sup>
Hence it is rational for agents to forgo education if they expect
the value of information of the activity to be lower than the price
they would have to pay for it. However, such an estimate of the
expected value of information may be grossly inaccurate, in which
case agents may make decisions that are contrary to their moral
preferences simply out of ignorance. Apart from lacking information,
cognitive biases such as scope neglect,[^kahneman2000economic
212][^desvousges2010measuring 66] the identified victim
effect,<sup id="fnref:ErlandssonIVE"><a class="footnote-ref" href="#fn:ErlandssonIVE">60</a></sup><sup id="fnref:kogut2005identified"><a class="footnote-ref" href="#fn:kogut2005identified">94</a></sup><sup id="fnref:jenni1997explaining"><a class="footnote-ref" href="#fn:jenni1997explaining">90</a></sup><sup id="fnref:small2003helping"><a class="footnote-ref" href="#fn:small2003helping">137</a></sup>
and the bystander
effect,<sup id="fnref:hudson2004bystander"><a class="footnote-ref" href="#fn:hudson2004bystander">84</a></sup><sup id="fnref:darley1968bystander"><a class="footnote-ref" href="#fn:darley1968bystander">49</a></sup> are
often involved.</p>
</dd>
<dt>Failures of coordination.</dt>
<dd>
<p>Even when groups of agents are value aligned and well informed, they
can fail to minimize the overall total of the opportunity costs of
their investment decisions if they lack critical ingredients of
coordination, such as communication and trust.</p>
</dd>
</dl>
<aside>
The *identified victim effect* is closely
associated with the *proportion dominance effect* and *in-group
effect*[^ErlandssonIVE], and thus makes for a rich example of the
psychological phenomena that will not be covered any further. Three
factors that influence people’s readiness to help are their sympathy
with the victim, the perceived utility of their help, and their feeling
of responsibility.
Sympathy can be aroused by giving the name and possibly a picture of the
victim rather than citing a statistic – even though these donors most
likely know that each person in such a statistic is also an individual
with name and face.
The perceived utility is partially dependent on the reference class, so
that people are more excited about helping 20 people out of a group of
24 rather than 20 out of 400 – even though the value of each life does
not change as the reference class changes, and reference classes are
often chosen arbitrarily.
Feelings of responsibility are stronger the closer or more alike the
victim is. Erlandsson, Björklund, and Bäckström, for example, vary the degree of the family
relationship but other “in-groups,” such as nationality, ethnic group,
or species, are imaginable. This comes at a detriment to those who lack
any in-group members with the capacity to help them, for example many of
the poor in developing countries.
</aside>
<p>The contributions of this project only address agents that feel a
sufficient moral alignment with the goals of the Against Malaria
Foundation (<span class="caps">AMF</span>), and the problem of lacking education is being addressed
by several organizations of the effective altruism spectrum and hence
left outside the scope of this text. Rather, the contributions of this
project aim to improve trust into the prioritization decisions of <span class="caps">AMF</span>
through enhanced transparency, a crucial ingredient of the coordination
of altruistic efforts.</p>
<p>In the following sections of this chapter I will discuss these hurdles
in more detail in order to locate the contributions of this project
within their context, introduce methods of prioritization research, and
then focus on the specific prioritization decisions relevant to the
intervention that the <span class="caps">LLIN</span> Visualizer addresses: the distribution of
long-lasting insecticide-treated mosquito nets for malaria vector
control. On that foundation I will describe the process of conception
and implementation of the <span class="caps">LLIN</span> Visualizer: previous work that it builds
upon, the requirements it was conceived of to satisfy, the
considerations that have gone into its architecture and implementation,
and finally the evaluation of the results and important learnings drawn
from the process.</p>
<p>“I could have got more out. I could have got more. I don’t know. If I’d
just… I could have got more.”</p>
<h2 id="ch:previous-work"><a class="toclink" href="#ch:previous-work">Previous Work</a></h2>
<h3 id="sec:previous-research-altruistic-coordination"><a class="toclink" href="#sec:previous-research-altruistic-coordination">Previous Research on Altruistic Coordination</a></h3>
<h3 id="subsec:wtp"><a class="toclink" href="#subsec:wtp">Willingness to Pay for Increased Moral Goal Satisfaction</a></h3>
<dl>
<dt>Moral preferences.</dt>
<dd>
<p>A basic metric in economic analysis is the willingness to pay (<span class="caps">WTP</span>) of an
agent. Agents have different preferences, and as these preferences are
met, the marginal benefit of further preference satisfaction along the
same dimension decreases, which is termed the <em>principle of decreasing
marginal benefit</em>: As the marginal benefit of a good to an agent
decreases, so does the agent’s <span class="caps">WTP</span> for that good.[^parkin2008economics 40]</p>
</dd>
</dl>
<p>Usually agents consider some of these preferences <em>moral preferences</em> –
some typical examples are a preference for feeling beings to suffer as
little as possible or a preference for people not to lie. I will leave
it to the intuition of the reader to decide what should constitute a
<em>moral</em> preference as opposed to any other preference, as any definition
I have seen leaves a significant gray area and would not be necessary
for comprehension of this text.</p>
<p>Rather I will distinguish goals according to the degree to which they
are shared by agents. Here agent-relative goals (such as personal
profit, neighborly help, patriotism, etc.) have an inherent upper bound
on the extent to which other agents can share them, while agent-neutral
goals (reducing suffering, maximizing justice, maximizing the number of
paperclips in the universe, etc.) do not have such a limit.</p>
<p>To aid comprehension, however, I will make an artificial distinction of
moral <em>preferences</em> and moral <em>goals</em> that becomes meaningful in the
case of agent-relative preferences: two people with a personal profit
motive share the same preference for profit but their goals are
different ones since they are different agents. If they also share the
agent-neutral preference for minimizing global suffering, then they also
share the same goal of reducing it.</p>
<p>At least agent-neutral goals are pure public goods in that they are
<em>nonrival</em> and <em>nonexcludable</em>: One agent’s enjoyment of achievement of
a moral goal does not lessen another agent’s enjoyment, and it is
impossible (or prohibitively costly) to bar someone from enjoying
it.[^hall2009microeconomics 477]</p>
<p>Analogously, I will not distinguish donations from purchases or
investments, except insofar as I view them all as different types of investments.</p>
<dl>
<dt>Measuring the willingness to pay.</dt>
<dd>
<p>In some contexts it is hard to tease apart the influence of
agent-neutral and agent-relative motives on someone’s revealed
preference. People may be concerned about climate change because of the
potential harm it can cause for all sentient life or because they own
property in coastal regions.</p>
</dd>
</dl>
<p>One area where researchers have found this to be more feasible is in
farmed animal welfare. Norwood and Lusk determined the revealed <span class="caps">WTP</span> of
a fairly representative sample of informed consumers for (1) pork and
eggs from different systems of animal farming[^norwood2011compassion
266] and (2) changes of systems of animal farming for animals whose
produce the subjects could not consume.[^norwood2011compassion 294]
Particularly the second part is of interest here, as such welfare
increases constitute a pure public good.</p>
<p>The authors could not simply study point of sales data and infer
people’s preferences and <span class="caps">WTP</span> from these because most people are uninformed
of practices in animal agriculture. This parallels a wider problem in
altruistic investment that will be the subject of the following subsections.</p>
<p>Norwood and Lusk thoroughly informed a sample of consumers about
advantages and disadvantages for animal welfare of production systems
such as battery cages and the free range system for hens, took measures
to minimize social incentives, and then auctioned animal welfare changes
affecting set numbers of animals at university farms far from the
research locations. The study relied on actual payments rather than
hypothetical questions, which usually yield exaggerated valuations. Thus
the influence of environmental concerns associated with animal
agriculture is minimized since the number of animals in question is
unaffected, and selfish interests, such as health benefits, are
minimized since the subjects are highly unlikely to ever get in touch
with the farms’ produce.</p>
<p>Nonetheless, two thirds of the participants were willing to pay for a
single hen to be moved from the cage to the free range system for an
average (including the one third of bids of zero) of
\$1.08.[^norwood2011compassion 298]</p>
<p>Results like these show that moral goals, absent selfish interests, have
real monetary value to a majority of people.</p>
<h3 id="subsec:moral-preferences"><a class="toclink" href="#subsec:moral-preferences">Diversity of Moral Preferences</a></h3>
<p>As will be detailed in section [subsec:motivations-donors], the
benefits from seemingly altruistic actions like donating are for many
people either dominated by other benefits than moral preference
satisfaction or their moral preferences are hard to distinguish from the
self-interested preferences of other people. For other people their
moral preferences are paramount. In either case moral preferences are of
different nature and importance for different people.</p>
<p>The reasons for this disagreement are subject of the philosophical
discipline of metaethics. The two metaethical views that are typically
distinguished are moral realism and moral antirealism. According to the
first doctrine, there are objective moral facts while according to the
second morality is subjective.</p>
<p>Singer argues that this distinction, while fundamental,
is yet of little practical relevance. While moral realists would
attribute the diversity of moral views to differences in beliefs about
the world, antirealists attribute it to differences about personal
preferences. Consequently, realists have failed to establish how any set
of moral goals follows from circumstances of the observable world while
antirealists, by nature of their claim, fail to establish why any set of
moral goals should be binding for anyone.<sup id="fnref:singer1973triviality"><a class="footnote-ref" href="#fn:singer1973triviality">135</a></sup>
<sup id="fnref:peter-singer-disclaimer"><a class="footnote-ref" href="#fn:peter-singer-disclaimer">136</a></sup> Since neither view has been substantiated
by evidence, there is instead a considerable debate over which view bears
the burden of proof in the first place.</p>
<p>As these doctrines use different language to describe the same facts, I
am forced to choose lest I need to present every argument from two
perspectives. My weakly held perception is that antirealism is the more
parsimonious theory, so that I will use the language of antirealism
throughout this paper. I want to be clear, however, that this choice is
motivated not by evidence for the view but by an absence of evidence for
the competing view. Either view of course acknowledges that differences
in revealed moral preferences are real.</p>
<p>Often these differences are due to insufficient information or biases. A
person may, for example, be willing to pay significant amounts for the
building of schools in a region, but may then discover that the
availability of schools is not the bottleneck to achieving better
education in that region. Plausibly, for example, the children’s health
or the parents’ and teachers’ beliefs about the benefits of education
may be bottlenecks.[^banerjee2012poor 79] Now the person may be willing
to pay much less for more schools and much more for improved health and
improved access to information in the same region. The preference for
more schools was not an informed preference and thus not indicative of a
<em>terminal</em> moral goal.</p>
<p>Still somewhat unpragmatically, a terminal moral goal can be defined as
any moral goal that a person would still hold even if they were
arbitrarily intelligent, perfectly rational, and had available to them
all relevant information.<sup id="fnref:6"><a class="footnote-ref" href="#fn:6">6</a></sup> In the following I will use <em>moral goal</em>
(or <em>moral preference</em>) to mean <em>terminal moral goal</em> unless stated otherwise.</p>
<h3 id="subsec:identifying-moral-preferences"><a class="toclink" href="#subsec:identifying-moral-preferences">Identifying Moral Preferences</a></h3>
<p>As in the case of animal agriculture, we again face the problem that we
cannot simply study donation statistics to infer people’s moral
preferences since most of these revealed preferences are likely far from
being terminal moral preferences.</p>
<p>The Money For Good <span class="caps">II</span> study provides some quantification of the problem.
It found that only 6% of donations from individual donors (as opposed to
advisors and foundations) are informed by any comparative research, and
33% are informed by any research, with the additional 17 pp. of donors
only doing research on the one charity they seek to donate
to.[^MFGIIStudy 14] Without doing comparative research, these donors are
not informed of whether there are charities that realize their moral
goals much more directly and efficiently, and regarding the motivations
that Money For Good <span class="caps">II</span> found for donations, we will also see that many
are unrelated to moral goals.</p>
<p>One limited means for approximating terminal moral goals are surveys of
philosophers such as that of the PhilPapers Foundation, following the assumption
that philosophers are likely to have had to question their moral
intuitions for some time. The survey reveals that the three most popular
classes of moral theories, deontology (25.9%), consequentialism (23.6%),
and virtue ethics (18.2%), share the space with no single one clearly
dominating the others.<sup id="fnref:PhilPapersSurvey"><a class="footnote-ref" href="#fn:PhilPapersSurvey">142</a></sup> These do not allow inferences
to the popularity of individual moral goals, but they do make
implications about the relative popularity of classes of moral goals at
least within the surveyed cohort.</p>
<p>One caveat is that this survey appears to have a fairly informal
character, which makes it hard to assess its quality. Another caveat is
that philosophers of domains other than the branches of ethics may have
no informed opinion on moral questions but that limiting the sample to
ethicists greatly reduces the sample size. Finally there is some
evidence to contest the assumption that education about ethical
questions reduces a person’s susceptibility to
bias.<sup id="fnref:schwitzgebel2012expertise"><a class="footnote-ref" href="#fn:schwitzgebel2012expertise">132</a></sup></p>
<p>Another interesting branch of research is that associated with Moral
Foundations Theory.<sup id="fnref:graham2012moral"><a class="footnote-ref" href="#fn:graham2012moral">106</a></sup> Researchers in this space try to
find moral foundations – which may be analogous to what I refer to as
moral goals – with high predictive power for explaining cultural, often
political, differences. The researchers that proposed the theory
proceeded from a set of first five, later six, and in the future
possibly more moral principles. The original set of five principles
derived from early work by Shweder[^graham2012moral 16] and has since
been expanded. More recent papers tested the principles’ predictive
power through self-report surveys, measures of reaction time, facial
micro-expressions, event-related potentials, neuroimaging, and text
analysis.[^graham2012moral 15] In all cases that I am aware of, the set
of principles was predetermined, and in the case of the surveys the
researchers had developed questions that aimed to isolate single
principles in a hypothetical situation that was simple enough not to
require further education of the subject.</p>
<p>Here is the latest set of foundational principles according to the
researchers’ website:<sup id="fnref:MoralFoundations"><a class="footnote-ref" href="#fn:MoralFoundations">77</a></sup></p>
<p>Each culture then constructs virtues, narratives, and institutions on
top of these foundations, thereby creating the unique moralities we see
around the world, and conflicting within nations too. The five
foundations [plus one “very good candidate”] for which we think the
evidence is best are:</p>
<dl>
<dt>Care/harm</dt>
<dd>
<p>This foundation is related to our long evolution as mammals with
attachment systems and an ability to feel (and dislike) the pain
of others. It underlies virtues of kindness, gentleness,
and nurturance.</p>
</dd>
<dt>Fairness/cheating</dt>
<dd>
<p>This foundation is related to the evolutionary process of
reciprocal altruism. It generates ideas of justice, rights,
and autonomy. Note: In our original conception, Fairness included
concerns about equality, which are more strongly endorsed by
political liberals. However, as we reformulated the theory in 2011
based on new data, we emphasize proportionality, which is endorsed
by everyone, but is more strongly endorsed by conservatives.</p>
</dd>
<dt>Loyalty/betrayal</dt>
<dd>
<p>This foundation is related to our long history as tribal creatures
able to form shifting coalitions. It underlies virtues of patriotism
and self-sacrifice for the group. It is active anytime people feel
that it’s “one for all, and all for one.”</p>
</dd>
<dt>Authority/subversion</dt>
<dd>
<p>This foundation was shaped by our long primate history of
hierarchical social interactions. It underlies virtues of leadership
and followership, including deference to legitimate authority and
respect for traditions.</p>
</dd>
<dt>Sanctity/degradation</dt>
<dd>
<p>This foundation was shaped by the psychology of disgust
and contamination. It underlies religious notions of striving to
live in an elevated, less carnal, more noble way. It underlies the
widespread idea that the body is a temple which can be desecrated by
immoral activities and contaminants (an idea not unique to
religious traditions).</p>
</dd>
<dt>Liberty/oppression</dt>
<dd>
<p>This foundation is about the feelings of reactance and resentment
people feel toward those who dominate them and restrict
their liberty. Its intuitions are often in tension with those of the
authority foundation. The hatred of bullies and dominators motivates
people to come together, in solidarity, to oppose or take down
the oppressor. We report some preliminary work on this potential
foundation in this paper, on the psychology of libertarianism
and liberty.</p>
</dd>
</dl>
<p>While the hypothetical scenarios used in at least one study did seem
sufficiently simple to me to readily comprehend their implications, the
isolation of the principles was less successful in several cases. For
example, one question that tests for readiness to subvert authority asks
what monetary incentive would be sufficient for the respondents to
“curse [their] parents, to their face,” answers to which may just as
soon be indicative of the respondents’ readiness or reluctance to
inflict harm.</p>
<p>The theory has been criticized for being less parsimonious than
established taxonomies that rely only on
harm<sup id="fnref:doi:10.1080/1047840X.2012.651387"><a class="footnote-ref" href="#fn:doi:10.1080/1047840X.2012.651387">78</a></sup> or only on
authority.<sup id="fnref:KuglerPoliticalOrientation"><a class="footnote-ref" href="#fn:KuglerPoliticalOrientation">95</a></sup> Conversely I could not find any
mention of whether principles such as the proliferation of knowledge or
beauty can be reduced to the known principles or whether they might
represent principles of their own right. The question is suggested by
the significant levels of philanthropy to support universities and the
arts. Perhaps the research is forthcoming.</p>
<h3 id="subsec:values-spreading"><a class="toclink" href="#subsec:values-spreading">Values Spreading</a></h3>
<p>Agents interested in one or several of the same moral goals will soon
notice that they can reach these more easily when they cooperate on
realizing them, convince more agents to join their cause, avoid zero-sum
competition, and possibly even trade with groups with different goals. I
have put these modes of values spreading into the following order so
that the cheapest method (requiring minimal resources and sacrifice) has
first priority, while the rest require increasing effort and
compromising on one’s goals.</p>
<p><strong>Cooperation.</strong> Agents can agree on some of their (not necessarily
terminal) moral goals and cooperate toward their realization.</p>
<p><strong>Education.</strong> Education comes in two subforms, both of which can
generally be thought of as cooperative behaviors.</p>
<p><strong>Correcting ignorance.</strong> Educating other agents on topics they had not
considered or did not know enough about.</p>
<p><strong>Correcting misinformation.</strong> Helping other agents realize that some of
their assumptions about the world are mistaken.</p>
<p><strong>Trade.</strong> When it has either been established that both sides disagree
on preferences rather than beliefs or when both sides find that the
other is resistant to education to the point that it would make it
prohibitively costly, then both sides can usually gain from avoiding a
zero-sum fight in favor of compromise.</p>
<p>These are explained in more detail below:</p>
<aside>
Stag hunt.
: The stag hunt is a scenario where at least two agents have the
option to cooperate on a common goal that would be optimal for both
of them or defect in favor of some smaller gain, thereby also
ruining the chances of the other to attain more than the
smaller gain. The payoff matrix of all options (cells) for both
players (tuple elements):
————————————————————————————————————
————————————————————————————————————
Prisoner’s dilemma.
: The prisoner’s dilemma is different from the stag hunt in that the
cases in which exactly one agent defects are optimal for the
respective agent. Since the outcome of both defecting is slightly
better than cooperating while the other defects, there is a double
incentive to defect. The objectively best outcome, however, is that
of both agents cooperating. The payoff matrix of all options (cells)
for both players (tuple elements):
————————————————————————————————————
————————————————————————————————————
Nash equilibrium.
: In our examples, the Nash equilibria are the scenarios in which
neither agent has anything to gain by changing their own strategy
when they learn the strategy of the other agent.
</aside>
<dl>
<dt>Cooperation.</dt>
<dd>
<p>When agents have several moral goals each, the situation resembles a
stag hunt in that there are two Nash equilibria – when the agents
cooperate and when they defect. (In the prisoner’s dilemma, for
comparison, the only Nash equilibrium is the case of both agents defecting.)</p>
</dd>
</dl>
<p>To cooperate, they will need to agree on pursuing the moral goal that
they share. A concrete example may be the following: One activist is
interested in both the proliferation of knowledge and the reduction of
suffering. Another activist is focused solely on the reduction of
suffering. Through dialogue, they can come to agree on the shared goal
of reducing suffering. A variation is the situation where at least one
agent shares only an instrumental goal with the other agent. Thus, for
example, a queer activist and an anti-<span class="caps">HIV</span>/<span class="caps">AIDS</span> activist can both agree to
support secularization efforts in Uganda. Of course, the agents can
enter into several such cooperative efforts if their working
relationships and resources allow.</p>
<dl>
<dt>Education.</dt>
<dd>
<p>Education comes in two different forms depending on whether the
recipient is mistaken or merely oblivious.</p>
</dd>
</dl>
<h1 id="correcting-ignorance"><a class="toclink" href="#correcting-ignorance">Correcting ignorance.</a></h1>
<dl>
<dd>When people do not seem to share any moral goals, it may be the case
that they simply never considered some key question. They may not be
aware that there is a decision that they have constantly been making by
omission because they were unaware of it. Telling such people about the
options they have may greatly enrich their thinking and their lives. It
may also come at a benefit to the internal consistency of their moral
system or their identity and thus reduce cognitive dissonance. Some
people may never have considered philanthropy, for example, so simply by
telling them about this possibility one can create new cooperation partners.</dd>
</dl>
<h1 id="correcting-misinformation"><a class="toclink" href="#correcting-misinformation">Correcting misinformation.</a></h1>
<dl>
<dd>One variation on this theme is the case where one party has badly chosen
instrumental goals but compatible terminal goals, whether they are aware
of them (thinking their instrumental goals are well chosen) or not
(mistaking their instrumental goals for terminal goals). In such cases,
one cost-effective and cooperative way of winning them for the
cooperative effort would be to provide them with the information they
are lacking.</dd>
</dl>
<p>For example, a person might think that “charity begins at home,” and
that when everyone supports their neighbors, friends, and family,
everyone is supported and even efficiently so, because we know how to
effectively support those close to us. But then they may learn that in
some countries poverty is so rampant, and the rich are so well insulated
from the poor, that any neighbors, friends, and family of the person in
need are just as much in need themselves. They may learn that there are
neglected interventions whose success is much more certain and
cost-effective than that of any interventions addressing common
afflictions in our society. Realizing this, the person may update and
abandon the mantra.</p>
<dl>
<dt>Trade.</dt>
<dd>
<p>Other people, however, may have the same convictions as the misinformed
person, but may not actually lack information. They may be well aware of
all one’s arguments and still not be swayed by them. Here the education
approach will fail. What remains is trade.<sup id="fnref:7"><a class="footnote-ref" href="#fn:7">7</a></sup></p>
</dd>
</dl>
<p>When these people have a strong interest in spreading their values, we
encounter a situation that resembles the prisoner’s dilemma. Much
research has been invested into this scenario. Axelrod for
example, has boiled down his insights from a competition of algorithms
in the repeated prisoner’s dilemma into four
rules:<sup id="fnref:axelrod2009evolution"><a class="footnote-ref" href="#fn:axelrod2009evolution">34</a></sup></p>
<p>Don’t be envious.</p>
<p>Don’t be the first to defect.</p>
<p>Reciprocate both cooperation and defection.</p>
<p>Don’t be too clever.</p>
<p><img alt="image" src="/images/llin-visualizer/gains-from-trade.png"></p>
<p>One very simple algorithm that implemented this behavior was <em>Tit for
Tat</em>, which cooperated unless the other agent chose to defect in the
last round. Thus it quickly forgave defection, cooperated by default,
punished and rewarded defection and cooperation respectively, and was
perfectly predictable.</p>
<p>Tomasik has investigated what this could mean for altruists.
He compares the case of (1) deep ecologists, which make a distinction
between humans and other animals and care terminally about diversity of
life and noninterference, and (2) animal welfarists, which care about
the well-being of all individual sentient beings.<sup id="fnref:TomasikCompromise"><a class="footnote-ref" href="#fn:TomasikCompromise">146</a></sup>
These two are often opposed. Humans have been able to reduce the
suffering of humans and domesticated animals due to disease, starvation,
adverse climate, and many other sources by developing technical,
medical, and economic remedies. Other animals are still exposed to all
these to maximal degrees and cannot change these conditions for
themselves, centrally due to their lower intelligence. Hence, improving
animal welfare will require intervention in ecosystems, presumably by
humans directly or proximately.</p>
<p>When these opposed factions both want to control the future to implement
their moral preferences, they would have to fight each other, a fight
they would win with some probabilities less or equal to one that sum up
to one in this simplified example.</p>
<p>However, both sides know aspects of the future that they care more
strongly about than about others, while the opposing faction does not
necessarily care more about the same aspects. Animal welfarists, for
example, may value a world with few animals suffering greatly over a
world with many animals suffering, but deep ecologists may be
indifferent to the number of animals per species so long as diversity is maintained.</p>
<p>Based on their estimates of how likely each faction would be to
overpower the other in a fight and the value they put on different
aspects of their utopias, they can arrive at a compromise that is better
with certainty for both than the expected value of a fight.</p>
<p>This is already happening. One example is the cooperation between <span class="caps">VEBU</span>
(“Vegetarian Union”) and the sausage producer Rügenwalder on creating
vegetarian products. While both sides had to make concessions,
Rügenwalder was able to achieve financial success and <span class="caps">VEBU</span> reduced the
degree of animal suffering experienced by the company’s
livestock.<sup id="fnref:WeltRügenwalder"><a class="footnote-ref" href="#fn:WeltRügenwalder">51</a></sup><sup id="fnref:VebuRügenwalder"><a class="footnote-ref" href="#fn:VebuRügenwalder">139</a></sup></p>
<p>When the factions are less homogeneous than a company or association,
then there is the additional risk of subfactions forming and obviating
the conditions of the original compromise. Such considerations would
complicate the situation further than necessary for this overview since
most of the users of the <span class="caps">LLIN</span> Visualizer will likely be individuals or
tight-knit groups.</p>
<h3 id="subsec:trust"><a class="toclink" href="#subsec:trust">Trust</a></h3>
<p><img alt="“A woman poses for a picture in front of her net.” Photo and caption courtesy of AMF." src="/images/llin-visualizer/20245-distribution-photo.jpg"></p>
<p>The practicalities of a proliferation of this approach and related forms
of trade based on moral differences have been explored in . One central
problem that is discussed in the paper is the problem of trust, which
takes two forms, factual and counterfactual trust.</p>
<p>To use an example of particular relevancy to the <span class="caps">LLIN</span> Visualizer: there
are factories that produce <span class="caps">WHOPES</span> <span class="caps">II</span>–approved nets as the Against
Malaria Foundation (<span class="caps">AMF</span>) uses<sup id="fnref:AMFNets"><a class="footnote-ref" href="#fn:AMFNets">29</a></sup> in Vietnam, Thailand, China, and
Tanzania<sup id="fnref:AMFFactories"><a class="footnote-ref" href="#fn:AMFFactories">31</a></sup>. These factories are run by for-profit companies.</p>
<p>When altruists want to deploy a number of nets in the fight against
malaria, they pay <span class="caps">AMF</span> to do so. In practice they pay <span class="caps">AMF</span> to pay some of these
factories to deliver nets, and <span class="caps">AMF</span> then cooperates with a local partner in
their distribution. But for simplicity let us view <span class="caps">AMF</span> as a black box: the
donor pays <span class="caps">AMF</span> to deploy nets because <span class="caps">AMF</span> is better at it than the donor.</p>
<p>Factual trust now means that the donor can actually trust <span class="caps">AMF</span> to deploy the
right number of high-quality nets. <span class="caps">AMF</span> can prove this by providing extensive
documentation in the form of photos, videos, statistics, and reports
from independent supervisors.</p>
<p>Counterfactual trust is more intricate; it means proving that the same
effect would not have occurred anyway, without the donors’
contributions. This is the central reason randomized controlled trials
use randomized control groups, namely to know what would have happened
without the intervention. In the case of <span class="caps">AMF</span>, for example, one of my
worries was that a distribution by the foundation may only replace
another distribution by another funder who instead invests into a less
cost-effective intervention. As we will see in section
[sec:strategy-amf], the intervention is too neglected for that to be a
major concern.</p>
<p>Trust is central to the <span class="caps">LLIN</span> Visualizer as it gives an overview over the
prioritization decisions <span class="caps">AMF</span> has made as well as their results as evinced by
successfully completed distributions. The specifics of how this is
achieved are the matter of chapter [ch:architecture-and-implementation].</p>
<h3 id="sec:previous-research-altruistic-prioritization"><a class="toclink" href="#sec:previous-research-altruistic-prioritization">Previous Research on Altruistic Prioritization</a></h3>
<p>Let us assume that the problem of value alignment is solved. For
example, one may be allocating only one’s own donation budget or that of
a foundation with a value-aligned team, or one might be deciding on
one’s own career path. In these cases the next hurdle to take is that of
deciding on the most viable path to take to realize one’s moral goals.</p>
<p><img alt="Gas prices across the US according to GasBuddy." src="/images/llin-visualizer/gasbuddy-2015-12-07.png"></p>
<p>Here, first, three very unlike entities require an introduction,
GasBuddy, disability-adjusted life years (DALYs), and the <em>Disease Control Priorities in Developing Countries</em> study.</p>
<p>GasBuddy is a website that tracks gasoline prices across the <span class="caps">US</span> and
Canada. At the time of writing, the lowest average price across all
states of the <span class="caps">US</span> was \<span class="math">\(1.796 per gallon and the highest \\)</span>2.802 per
gallon (<span class="math">\(\sigma = 0.226\)</span>). Admittedly, some extreme prices might have
gotten normalized by the averaging within each state, but even the range
of prices across all cities the website tracks is just slightly greater
(<span class="math">\(\sigma = 0.279\)</span>). The relevance of this observation will become
apparent later.</p>
<p>For many people it is intuitive to care about their own well-being and
that of others. This is evident, for example, in the wide acceptance of
government-funded health services in many countries. Health heavily
impacts well-being, as does freedom, intellectual edification, the
social environment, relationships, and many more factors. Researchers in
medicine often employ quality-adjusted life years (QALYs) and
disability-adjusted life years (DALYs) as measures for health and thus as
proxies, however imperfect, of well-being. DALYs are the sum of years lived
with disability (<span class="caps">YLD</span>) and years of life lost (<span class="caps">YLL</span>).
Sassi gives a more comprehensive introduction to the two
measures, including their historical background, and contrasts
them.<sup id="fnref:sassi2006calculating"><a class="footnote-ref" href="#fn:sassi2006calculating">131</a></sup></p>
<p><em>Disease Control Priorities in Developing Countries</em>, second edition, (<span class="caps">DCP2</span>)
uses <span class="caps">DALY</span> measures for the study of the
cost-effectiveness of 309 medical interventions that are typically seen
as highly cost-effective by developed-world medical
standards.<sup id="fnref:jamison2006disease"><a class="footnote-ref" href="#fn:jamison2006disease">88</a></sup> It is a four-year-long collaboration of
over 350 specialists from around the world, made possible by funding and
staff time from the Bill <span class="amp">&</span> Melinda Gates Foundation, the World Bank, the
World Health Organization, and the Fogarty International Center of the
National Institutes of Health.<sup id="fnref:GatesDCPP2002"><a class="footnote-ref" href="#fn:GatesDCPP2002">38</a></sup></p>
<h3 id="subsec:finding-best-charities"><a class="toclink" href="#subsec:finding-best-charities">Finding the Best Charities</a></h3>
<p><img alt="DCP2 cost-effectiveness estimates of interventions related to high-burden
diseases in sub-Saharan Africa.<span
data-label="fig:dcp2"></span>" src="/images/llin-visualizer/dcp2-results.cropped.png"></p>
<p>It is studies like <span class="caps">DCP2</span> as well as more recent ones<sup id="fnref:8"><a class="footnote-ref" href="#fn:8">8</a></sup> that GiveWell is
greatly interested in. GiveWell is an organization that has been looking
for the most effective charities in the world since 2007. Eight years
into its search, it has found four charities that meet their criteria.
One of these criteria is that the cost-effectiveness of the intervention
the charity implements has to exceed that of unconditional cash
transfers, the reasoning being that simply giving cash to people was
such a simple and scalable intervention that any more complex and thus
risky programs need to prove that they are at least better than
unconditional cash transfers in expectation.</p>
<p>Several such programs exist, but only two – the distribution of
long-lasting insecticide-treated bed nets and mass deworming in schools
– have implementing charities that GiveWell can confidently
recommend.<sup id="fnref:9"><a class="footnote-ref" href="#fn:9">9</a></sup> These charities have funding gaps of over \$100 million
for 2016–2018, meaning that tens of millions of people will end up
untreated or unprotected if the funding goals cannot be met. If the
Against Malaria Foundation could reach everyone who needs a bednet, that
funding gap could easily be five times as high.<sup id="fnref:GiveWellMarcyErskine"><a class="footnote-ref" href="#fn:GiveWellMarcyErskine">66</a></sup></p>
<p>According to GiveWell’s estimates, both these priority programs are
about ten times as cost-effective as unconditional cash transfers.
Expressed in dollars per <span class="caps">DALY</span> averted, the authors of <span class="caps">DCP2</span> estimate that in
sub-Saharan Africa, averting one <span class="caps">DALY</span> from malaria through distributions of
insecticide-treated bednets costs \<span class="math">\(5–17; averting one DALY from HIV/AIDS through
condom promotion and distribution costs \\)</span>52–112; and averting another
<span class="caps">DALY</span> from <span class="caps">HIV</span>/<span class="caps">AIDS</span> through antiretroviral therapy costs \$350–1,494.<sup id="fnref:10"><a class="footnote-ref" href="#fn:10">10</a></sup> Even
though the cost-effectiveness estimates span three orders of magnitude,
GiveWell has at least considered all of these interventions, and there
are charities implementing them.</p>
<p>For completeness it should be noted that GiveWell’s latest estimates at
the time of writing put <span class="caps">LLIN</span> distributions at \$2,838 per life
saved.<sup id="fnref2:GiveWellAMF"><a class="footnote-ref" href="#fn:GiveWellAMF">68</a></sup> People who die of malaria usually die at age five
or younger before they can build natural immunity.</p>
<p>For simplicity I will assume that QALYs can be converted to DALYs in the
following sample calculations, that the life expectancy is around 57
years,<sup id="fnref:WorldBankData"><a class="footnote-ref" href="#fn:WorldBankData">145</a></sup> that discounting for age and uncertainty can be
disregaded, and that the average life incurs less than 0.1 <span class="caps">YLD</span> per
year.<sup id="fnref:GBDCompare"><a class="footnote-ref" href="#fn:GBDCompare">86</a></sup> Then I arrive at about \<span class="math">\(58 per DALY
(\)</span>\sigma = 8$).<sup id="fnref:MAWAMFGuesstimate"><a class="footnote-ref" href="#fn:MAWAMFGuesstimate">55</a></sup></p>
<h3 id="subsec:motivations-donors"><a class="toclink" href="#subsec:motivations-donors">Motivations of Donors</a></h3>
<p><img alt="A Fermi estimate (using Guesstimate) for comparison of AMF and the
Make-A-Wish foundation that assumes that aniticipation and memory of a
granted wish are on average as good as not suffering hearing loss and
the wish itself is on average as desirable as being spared severe chest
injury.<span
data-label="fig:maw-model"></span>" src="/images/llin-visualizer/maw-model.png"></p>
<dl>
<dt>Make-A-Wish.</dt>
<dd>
<p>But there are also charities like the Make-A-Wish Foundation that invest
\$7,500 on average to make a wish come true for a child with a
life-threatening (but not necessarily terminal) illness.<sup id="fnref:MakeAWishFAQ"><a class="footnote-ref" href="#fn:MakeAWishFAQ">98</a></sup>
Based on the information in the organisation’s <span class="caps">FAQ</span>, I have created a simple
estimate (using the Guesstimate app) as an example of the rough Fermi
calculations that would help almost any donor improve their giving decisions.</p>
</dd>
</dl>
<p>In this model, I assume that the desirablility of being granted a wish
is somewhere on the spectrum of the <span class="caps">WHO</span> disability weights up to being
relieved of severe multiple sclerosis (disability weight of
0.707,<sup id="fnref:WHO2004disabilityweights"><a class="footnote-ref" href="#fn:WHO2004disabilityweights">153</a></sup> one of the most severe disabilities
listed). The estimate puts the cost for one <span class="caps">DALY</span> averted for the Make-A-Wish
Foundation very roughly somewhere in the area of a million dollars.</p>
<p>Since such quantitative estimates are rough and murky, one can instead
do a thought experiment to compare the intervention of Make-A-Wish to a
life-saving one like that of the Against Malaria Foundation. One can
imagine that there is a mother with a dying child. She has only \$7,500
and can spend it either to grant her child a last wish or to try an
experimental treatment that might save the child’s life. How unlikely
would the success of the treatment have to be before the mother would
choose to grant the wish instead? The probability could probably be so
small as to be intuitively indistinguishable from being infinitesimal.</p>
<p>Yet, in 2014 and in the <span class="caps">US</span> alone, people chose to donate almost \<span class="math">\(66
million to Make-A-Wish.[^ProPublicaMakeAWish] \\)</span>66 million that would
have easily filled the funding gap of all GiveWell top charities in
2014, and \$66 million that had the same impact as a donation of a few
thousand dollars to <span class="caps">AMF</span> would have had, so that they have come at the
opportunity cost of more than 23 thousand lives lost to malaria, lives
that could have been saved.</p>
<p>Make-A-Wish is just one example. As noted in the introduction, the
impact of 96% of <span class="caps">US</span> donations<sup id="fnref2:GivingUSA2015"><a class="footnote-ref" href="#fn:GivingUSA2015">74</a></sup> could have been multiplied
by an approximate factor of one hundred by investing them in a
developing country.</p>
<dl>
<dt>Types of donors.</dt>
<dd>
<p>The Money For Good <span class="caps">II</span> study<sup id="fnref:MFGIIStudy"><a class="footnote-ref" href="#fn:MFGIIStudy">83</a></sup> provides some quantification
and explanation of the problem. The authors investigated the motivations
of donors and clustered them into six categories, which they describe
with some key statements from the respective donors:</p>
</dd>
<dt>“repayer” (23%)</dt>
<dd>
<p>“I give to my alma mater”; “I support organizations that have had an
impact on me or a loved one.”</p>
</dd>
<dt>“casual giver” (18%)</dt>
<dd>
<p>“I give to well known nonprofits because it isn’t very complicated.”</p>
</dd>
<dt>“high impact” (16%)</dt>
<dd>
<p>“I support causes that seem overlooked”; “I give to nonprofits I
feel are doing the most good.”</p>
</dd>
<dt>“faith based” (16%)</dt>
<dd>
<p>“We give to our church”; “We only give to organizations that fit
with our religious beliefs.”</p>
</dd>
<dt>“see the difference” (13%)</dt>
<dd>
<p>“I think it’s important to support local charities”; “I give to
small organizations where I feel I can make a difference.”</p>
</dd>
<dt>“personal ties” (14%)</dt>
<dd>
<p>“I give when I am familiar with the people who run an organization.”</p>
</dd>
</dl>
<p>For each of these categories, they investigated the driving motivations
of the donors for specific donations.</p>
<p><img alt="Key drivers of donations.<span
data-label="fig:key-drivers"></span>" src="/images/llin-visualizer/mfg2-donor-drivers.png"></p>
<p>These categories suggest links to the moral foundations that
Graham et al. have found. For example, the faith-based and personal
ties donations may be motivated by feelings of loyalty. This may also
explain the strong propensity of people to donate to causes within their
country despite the higher price. The high percentage of donations made
as quasi repayment may be an artifact of the high evolutionary value of
not defecting in prisoner’s dilemma–like situations, only that here
these donations come at a high opportunity cost, leading to a tragedy of
the commons.</p>
<p>That said, the categories leave some questions unanswered, for example
what portion of “faith-based” giving goes to organizations that work on
furthering religious ends, such as churches, and what percentage goes to
any charity. The breakdown does make clear that no other factor that was
studied had significant influence on these donors’ choices,[^MFGIIStudy
134] but “fit with religious beliefs” is the only attribute that
establishes this category, so it may be that a significant number of
them donated to any nondescript charity that happened not to conflict
with their religious beliefs.</p>
<p>Similarly, the “personal ties” category leaves open what portion of
these people is doing a favor to friends and family by donating and what
portion is merely so risk-averse that they only trust organizations that
are run by people they know personally. The leading attribute, “Familiar
with org/leadership,” indicates the second while the runner-up
attribute, “Friend/Family asked me,” indicates the first. A more
interesting grouping may be one into “personal ties” and “risk averse,”
where the first only captures the factors “Friend/Family asked me” and
“Try to support friends’ charities.”</p>
<dl>
<dt>The “GiveWell market.”</dt>
<dd>
<p>The study also found that only about 6% of donors do any comparisons
between charities before donating. On a market where the
cost-effectiveness of products varies by several orders of magnitude,
this indifference is striking. The top reasons for not conducting
research (any research, not just comparative research) are that donors
felt they were familiar with the organization, that it is a well-known
organization, that they are involved with the organization, that the
organization was recommended to them by a trusted person, that it is a
religious institution, and that the donation was small. Few of these
could double as arguments against comparative research. Inaccessibility
or absence of information was only cited as a problem by few
donors.[^MFGIIStudy 46] On the other hand they also identified
effectiveness as the key unmet need of individual donors that were
interested in conducting research.[^MFGIIStudy 50] Taking into account
that better research would sway some donors toward not donating and some
toward donating to better organizations, the researchers estimated that
better information could move about \<span class="math">\(10 billion per year from
individual donors toward higher-performing organizations (about \\)</span>15
billion in total, taking foundations and advisers into account).</p>
</dd>
</dl>
<p>GiveWell requested the raw data of the study to determine in a more
fine-grained fashion the size of the market that it serves. It presents
three estimates. Broadly, the donations from all donors already
interested in comparing and evaluating multiple organizations total
\<span class="math">\(4.1 billion according to its estimates. When it narrowed the facet
down by excluding donors that ranked what it considered to be factors
irrelevant to impact higher than relevant factors, the total shrunk only
slightly to \\)</span>3.8 billion. Only when they narrowed the facet to donors
that both fit the first group and ranked expected impact highest did the
amount shrink to \$554 million.</p>
<p>However, it also notes that the market may be still much bigger since
successful businesses often create solutions to problems that only a
fraction of its future customers had previously been aware of, or in the
author’s terms, “for most successful businesses I can think of, it’s
still the case that most people aren’t customers of them.”<sup id="fnref:GiveWellMFG"><a class="footnote-ref" href="#fn:GiveWellMFG">73</a></sup></p>
<h3 id="subsec:market-impact"><a class="toclink" href="#subsec:market-impact">A Market for Impact</a></h3>
<p>Once a significant number of people are explicitly donating where they
see the greatest potential for impact, something else will happen: a new
market will emerge. The reason the prices that GasBuddy tracks are as
close to each other as they are is that there are market pressures that
do not allow some companies to sell the same product at a multiple of
the price – no one would buy it. Conversely there is a strong incentive
for companies to become more cost-effective in their production so that
they can hand down the lower prices to their customers to attract more
of them.</p>
<p>Analogously, as soon as charities notice that there is money to be made
in becoming more effective, be it even just millions rather than
billions of dollars, they will be incentivized to become more
cost-effective too. With 86% of all donations coming from loyal
donors,[^MFGIIStudy 140] there is great value not only in redirecting
donations to more effective organizations but also in nudging
established organizations to become more effective.</p>
<p>One variation of this idea is to simply pay people, organizations, or
countries for the impact they have already achieved, thus incentivizing
impact and providing means for further impact, which, in turn, can be
purchased. Countries like Britain and Norway, and the Gates Foundation
are experimenting with this approach,<sup id="fnref:EconomistImpactPurchase"><a class="footnote-ref" href="#fn:EconomistImpactPurchase">141</a></sup> and
impact certificates are an example of such an effort by individual
funders.<sup id="fnref:TheImpactPurchase"><a class="footnote-ref" href="#fn:TheImpactPurchase">85</a></sup></p>
<h3 id="sec:prioritization-research"><a class="toclink" href="#sec:prioritization-research">Prioritization Research</a></h3>
<p>Having established the importance of prioritization in the introduction,
this chapter will introduce to methodologies, constraints, and the
interest groups of prioritization research.</p>
<h3 id="subsec:constraints"><a class="toclink" href="#subsec:constraints">Constraints</a></h3>
<dl>
<dt>Purview of prior research.</dt>
<dd>
<p>One early effort at creating a consolidated overview of
cost-effectiveness measures for priorities setting was the first edition
of (<span class="caps">DCP1</span>) published in 1993. It surveyed 52 interventions, or more than 100
when grouped by cause, as some have multiple objectives. For this
research they drew on studies dating back into the
1970s.<sup id="fnref:jamison1993disease"><a class="footnote-ref" href="#fn:jamison1993disease">89</a></sup></p>
</dd>
</dl>
<p>The second edition expanded this set to 319 interventions whose
cost-effectiveness measures were compiled by a staff of 350, more than
four times the staff of <span class="caps">DCP1</span>.<sup id="fnref2:jamison2006disease"><a class="footnote-ref" href="#fn:jamison2006disease">88</a></sup> The authors are quick to
acknowledge the specificity of their purview:</p>
<blockquote>
<p>A frequent, often justified, criticism of cost-effectiveness analyses is
that they address only one of many criteria that could be used to
evaluate health interventions. Epidemiological, medical, political,
ethical, and cultural factors often also play important roles in the
decision to allocate resources to a specific health condition or inter-
vention; however, determining how one might weigh cost- effectiveness
ratios alongside these other considerations when setting priorities for
spending is difficult. Musgrove (1999) shows how to take some of these
connections into account, including circumstances in which cost-
effectiveness is an adequate criterion by itself. One approach is for
the policy maker to think of cost- effectiveness ratios as the relative
“price” of purchasing a unit of health (a <span class="caps">DALY</span>, for instance) using
different inter- ventions. These costs, along with the budget
constraint, can help determine the optimal allocation of resources among
a given set of interventions.</p>
</blockquote>
<p>This clarification omits what exactly is meant by cost-effectiveness
analysis in this context. If, for example, cultural factors make condom
distributions infeasible, all this would mean is that a policy maker
would incur high costs due to the marketing necessary to change cultural
norms or that they would incur low adoption rates and thus low
effectiveness. In either case the effect could again be expressed in
terms of cost-effectiveness. The implementing party might also lose
acceptance among the population, incurring a different type of cost in
addition to the monetary one. The effectiveness might suffer, too, in
that the intervention may be cut sort after the next election so that it
will have less time to earn back some <span class="math">\(O(1)\)</span> startup costs. Or if an
epidemic is about to break out, a treatment for the disease that used to
be less cost-effective could become highly cost-effective in view of the
rapid rate of future infections that containment of the disease can forestall.</p>
<p>But that is not what is meant by cost-effectiveness analysis in the
context of <span class="caps">DCP2</span>. Cost-effectiveness analyses of communal, medical
interventions are hard to create, and usually require randomized
controlled trials (RCTs) at scales much smaller than the interventions
policy makers want to implement. It is RCTs like these that <span class="caps">DCP2</span> compiles and
cost-effectiveness analyses like these that it is referring to throughout.</p>
<p>As GiveWell has found repeatedly, such estimates can be
flawed.<sup id="fnref:GiveWell2011Errors"><a class="footnote-ref" href="#fn:GiveWell2011Errors">70</a></sup> Even more importantly, however, their
external validity is highly limited. A recent paper published by the
IDinsight team – and endorsed even by a staunch critic of RCTs in
development economics<sup id="fnref:Pritchett2015"><a class="footnote-ref" href="#fn:Pritchett2015">122</a></sup> – introduces the taxonomy of
knowledge-focused evaluations (KFEs), “those primarily designed to build
global knowledge about development interventions and theory,” and
decision-focused evaluations (DFEs), those “driven by implementer
demand, tailored to implementer needs and constraints, and embedded
within implementer structures,” that serves as basis for their
recommendation: “Where the primary need is for rigorous evidence to
directly inform a particular development programme or policy, DFEs will
usually be the more appropriate tool. KFEs, in turn, should be employed
when the primary objective is to advance development theory or in
instances when we expect high external validity, <em>ex ante</em>. This
recalibration will require expanding the use of DFEs and greater
targeting in the use of KFEs.”<sup id="fnref:Shah2015"><a class="footnote-ref" href="#fn:Shah2015">134</a></sup></p>
<dl>
<dt>Measures at scale.</dt>
<dd>
<p>While RCTs have low external validity, self-evaluations of <span class="caps">NGO</span> that are
scaling programs cannot feasibly measure their impact.<sup id="fnref:11"><a class="footnote-ref" href="#fn:11">11</a></sup> Doing so
would be very costly, they could only serve half the population because
of the randomized control group, the results would likely be of poor
quality because of the tremendous effort it would take to monitor the
operation, and while randomization of individual intervention on village
level may avoid problems of inequity,<sup id="fnref:haushofer2015your"><a class="footnote-ref" href="#fn:haushofer2015your">81</a></sup> it may also
produce less reliable results.</p>
</dd>
</dl>
<p>What NGOs do instead is that they measure proxies that the prior
research allows them to translate to impact fairly reliably. <span class="caps">AMF</span>, for
example, measures net condition and use in intervals of six months after
each distribution. These are data that can then be used to extrapolate
what the impact in terms of DALYs averted may be.</p>
<dl>
<dt>Limited value of information.</dt>
<dd>
<p>Another factor that will crop up repeatedly in this text is that the
expected value of information sometimes does not warrant the attention
of the researchers. If the same time can be used to find interventions
that are more effective or have larger funding gaps at a sufficient
level of effectiveness, then these should be prioritized within the
prioritization process. There may be more hurdles of the same kind
depending on the context, as the next section will explain.</p>
</dd>
</dl>
<h3 id="subsec:methodologies"><a class="toclink" href="#subsec:methodologies">Methodologies</a></h3>
<h4 id="subsubsec:government-aid"><a class="toclink" href="#subsubsec:government-aid">Government Aid</a></h4>
<p>Priorities setting in the political domain is a highly complex process
that emerges as a multitude of agents, some willing to compromise to
greater degrees than others, struggle to realize their preferences,
whether agent-neutral or agent-relative.</p>
<p>arrives at an unfavorable verdict on the impact of governmental
development aid over the past decades.<sup id="fnref:moyo2011dead"><a class="footnote-ref" href="#fn:moyo2011dead">107</a></sup> At the same time,
many invaluable research efforts (such as many cited in this text) were
partially funded by governmental or government-funded institutions, and
similar institutions also implement interventions with strong
evidentiary bases. Whether all the government activities over the past
decades add up to a net plus or minus is impossible to determine.</p>
<p>Instead of trying to answer such sweeping questions as whether
multilateral development aid is beneficial or harmful, Banerjee and Duflo,
as well as the many outputs of the Abdul Latif Jameel Poverty Action Lab
in general, make many concrete and highly specific recommendations for
interventions that require action from governments or large funders but
remain relatively neglected.<sup id="fnref:banerjee2012poor"><a class="footnote-ref" href="#fn:banerjee2012poor">35</a></sup></p>
<p>Even though was published in 2011 and many of the studies it cites much
earlier still, at least some of the interventions – such as school-based
deworming – remain underfunded. Hauck and Smith investigated the
concrete reasons for the hold-ups, which resulted in an analyses of
typical problems of the political process itself. They identified the
following explanatory models in the literature:<sup id="fnref:hauck2015politics"><a class="footnote-ref" href="#fn:hauck2015politics">80</a></sup></p>
<p>Models of competing interest groups, based on the assumption that
powerful interest groups – for example provider organizations – may seek
to skew decisions in their own favour at the expense of less organized stakeholders.</p>
<p>Voting models, such as the median voter model, (Ahmed and Greene 2000),
which assert that political decision-makers will seek to develop
policies that attract particular voter groups, in an effort to maximize
political support. The implication of this insight for priority setting
is that the size and contents of a public benefits package may be skewed
towards the preferences of key voting groups.</p>
<p>Bureaucratic decision-making models, which assert that ‘bureaucrats’ may
make decisions in their own interests, rather than the interests of the
population as a whole.</p>
<p>Institutional economics to explain how governmental institutions and
more generally the organisation of the political system influence decisions</p>
<p>Decentralization and contracting-out to non-governmental organizations,
and implications on provision of public services.</p>
<p>Since the processes of priorities setting are so deeply embedded in a
such a much more complex system, it is not the ideal context in which to
explain any methodologies.</p>
<h4 id="subsubsec:foundations"><a class="toclink" href="#subsubsec:foundations">Foundations</a></h4>
<p>The Good Ventures foundation cooperates closely with the Open
Philanthropy Project for prioritization research. It is highly
transparent about its processes and thinking, and I am familiar with its
work, two advantages that make it a good exemplar to explain one
methodology that foundations employ to allocate
funding.<sup id="fnref:OpenPhil2016Process"><a class="footnote-ref" href="#fn:OpenPhil2016Process">112</a></sup></p>
<p>Given the findings of that only 39% of foundations seek to maximize
their impact with their grants and 36% never question whether some set
of cause areas they are associated with is in fact the most impactful
focus of their grants (often, perhaps, because their by-laws stipulate
it),[^MFGIIStudy 64] the methodology of the Open Philanthropy Project is
probably not representative of foundations in general, but it is
informed by interviews with scores of other foundations, and, crucially,
it is closely aligned with the thinking that <span class="caps">AMF</span> applies in its context.</p>
<dl>
<dt>Learning from other foundations.</dt>
<dd>
<p>One basic first step is to tap the knowledge that already exists out
there. Since other foundations may not be as transparent about their
decision making but may well have much valuable experience in the field,
co-funding of projects is one opportunity to access that experience.</p>
</dd>
<dt>Exploring potential causes.</dt>
<dd>
<p><em>Focus areas</em>, as the Open Philanthropy Project calls them, or cause
areas are often fixed for foundations by decree of the founder or
funder. Unless this person has invested significant research into this
decision and updates it as the changing circumstances require, it is a
matter of luck whether the cause area harbors any impactful investment opportunities.</p>
</dd>
</dl>
<p>The founder of <span class="caps">AMF</span> has boiled this down to a credo to the effect that the
goal of every charity should be to shut down,<sup id="fnref:12"><a class="footnote-ref" href="#fn:12">12</a></sup> because it is founded
to solve a specific problem, and needs to solve it even while that means
that it will make itself obsolete.</p>
<p>Foundations should take the same approach to choosing cause areas. Hence
the Open Philanthropy Project tries to gain an overview of the whole
spectrum of cause areas before them and ranks them according to three
core criteria:</p>
<dl>
<dt>Importance.</dt>
<dd>
<p>How many individuals does this issue affect, and how deeply?</p>
</dd>
<dt>Neglectedness.</dt>
<dd>
<p>All else equal, we prefer causes that receive less attention from
others, particularly other major philanthropists.</p>
</dd>
<dt>Tractability.</dt>
<dd>
<p>We look for clear ways in which a funder could contribute
to progress.</p>
</dd>
</dl>
<p>Given the sheer number of such fields, the funder needs to conduct many
fairly shallow investigations. These investigations aim to answer the
questions of what the problem is and how important it is; what avenues
there are for addressing it; and who else is working on it.</p>
<p>Having gained this cursory overview, some causes will stand out and thus
warrant deeper investigation: interviews with scores of experts of the
field, extensive reviews of the academic literature, and so-called
learning grants that open doors to the networks of grantees.</p>
<dl>
<dt>Selecting and prioritizing cause areas.</dt>
<dd>
<p>Having investigated these top cause areas, more factors are taken into
account to filter and rank them, such as the capacities and aptitudes of
the staff at the time.</p>
</dd>
<dt>Hiring and grantmaking.</dt>
<dd>
<p>Finally a process of delegation is initiated whereby one or several
domain experts are hired to assume responsibility of the area in the
long term. Thanks to sometimes decade-long experience in the field,
these people bring knowledge to the organization that by far exceeds
that of the foundation staff even after their deep investigations of the
areas. These new hires are gradually integrated into the organization
with the goal of them eventually running the research processes in their
domain nearly autonomously.</p>
</dd>
</dl>
<p>These are the stages that the Open Philanthropy Project has completed
for at least<sup id="fnref:13"><a class="footnote-ref" href="#fn:13">13</a></sup> two focus areas so far. Within these areas the domain
experts then recommend grants to the funders, especially Good Ventures.</p>
<p>When organizations receive grants and in response scale up their
operation, there is a lock-in happening that makes the organization
dependent on continued grants. Hence it is important to research
individual grants rigorously and, if that should become necessary, to
phase out giving with exit grants.<sup id="fnref:OpenPhil2015Giving"><a class="footnote-ref" href="#fn:OpenPhil2015Giving">110</a></sup> In any case, the
opportunity costs of a less than optimal decision are high, but a less
than optimal selection of cause areas would incur the combined
opportunity cost of all grants within it, so it is a much greater wager.</p>
<h4 id="subsubsec:individual-donors"><a class="toclink" href="#subsubsec:individual-donors">Individual Donors</a></h4>
<p>The methodology of the Open Philanthropy Project is a useful template
for understanding the processes for individual donors, but it is joined
by a few problems that lead to key characteristics that are probably
hard to avoid.<sup id="fnref:GiveWellProcess"><a class="footnote-ref" href="#fn:GiveWellProcess">72</a></sup></p>
<p>One key problem is that donors transfer their level of risk aversion
from investments into highly agent-relative preferences to investments
into agent-neutral investments.</p>
<p>Foundations may fund hundreds of projects hoping that one or two will
succeed to a degree that makes their investment cost-effective. In areas
where experience is scarce, this is often the only process of knowledge
creation there is.</p>
<p>Individual donors are usually only able to fund a fraction of one such
project and when it fails, as it is likely to, such a donor would be
disillusioned and possibly cease donating all together. This is unlikely
to change, so the recommendations prioritization research can make to
individual donors are limited to comparatively low risk–low return
opportunities.<sup id="fnref:14"><a class="footnote-ref" href="#fn:14">14</a></sup></p>
<dl>
<dt>Strong focus on individual organizations.</dt>
<dd>
<p>While foundations can think in terms of making grants toward cause
areas, individual donors are much more inclined to think in terms of
donating to individual charities.<sup id="fnref:15"><a class="footnote-ref" href="#fn:15">15</a></sup> For the foundation the most
important part of the research is in selecting the cause areas to focus
on, but for individual donors who cannot devote much time to the
process, having to review charities even within one cause area is not
feasible. Hence organizations in this space – such as GiveWell, Giving
What We Can, and Animal Charity Evaluators – need to make
recommendations of specific charities and make them with high confidence.</p>
</dd>
<dt>Funding gaps.</dt>
<dd>
<p>Foundations are often accountable to a small set of highly engaged
donors that exert control over them. In these cases they are not
accountable to thousands of donors, all with their own opinions or
preferences for how their money should be used, the way most other
charities are. However, when a foundation wants to invest its funding
highly effectively, and one opportunity it has found is one that many
individual donors are also donating to, it runs into two problems.</p>
</dd>
</dl>
<p>First, it cannot fill the funding gap of that giving opportunity
completely or it would deprive the individual donors of their giving
opportunity. Some may not respond by doing research and finding a new
target for their donations but simply by stopping to donate, lost
donations that may as well count as costs the foundation incurred though
its grant. Hence the foundation needs to make sure that it projects the
donation influx of the charity correctly and then grants only so much
and so often that its grants do not fill the charity’s funding gap.</p>
<p>Second, however, even that is not enough since such projections are by
necessity mostly based on past data, so that donations that the charity
receives at any time will influence future projections upward and thus
foundation grants downward. This relationship is not a very direct one,
but it is direct enough for GiveWell to go to some lengths to avoid it,
because individual donations would become partially fungible with
foundation grants, so that the foundation suddenly morally ought to be
accountable to those donors. That is not an option for a foundation that
wants to make experimental grants without a complex democratic decision
process, or, vice versa, drive away donors that are attentive enough to
notice this dynamic.<sup id="fnref2:OpenPhil2015Giving"><a class="footnote-ref" href="#fn:OpenPhil2015Giving">110</a></sup></p>
<p>Harmonizing grants and funding gaps such that other donation flows do
not influence them and there are still no contingencies of over or
underfunding an intervention is a delicate balancing act.</p>
<dl>
<dt>Marketing.</dt>
<dd>
<p>A foundation rarely needs to worry about its public perception, but an
organization that makes public recommendations has a strong interest in
its own growth, be it in terms of donors, in terms of money moved to its
recommended charities, or any other such metric. Hence a foundation has
considerable freedom in its grant making which the first organization lacks.</p>
</dd>
</dl>
<p>The result is that the tractibility criterion (cf. the taxonomy of
importance, neglectedness, and tractibility in section
[subsubsec:foundations]) is heavily influenced by how uncontroversial
the recommendations are in the current cultural or political discourse
of the donors. That is a significant limitation.</p>
<h3 id="sec:strategy-amf"><a class="toclink" href="#sec:strategy-amf">Strategy of the Against Malaria Foundation</a></h3>
<p><img alt="“The process includes awareness through banners, speeches and performances on the perils of malaria.” Photo and caption courtesy of AMF." src="/images/llin-visualizer/20070-distribution-photo.jpg"></p>
<p><span class="caps">AMF</span>’s prioritization strategy is a consequence of its hybrid position both
as a fundraiser that receives donations from a wide range of small and
major donors and as a grant maker. Donors to <span class="caps">AMF</span> are, by necessity, agnostic
about the location of the net distribution that they fund. <span class="caps">AMF</span> needs to
first hold significant funding before distribution partners will engage
with them in the lengthy proposal process. Hence donors can only
afterwards know what distribution their donation has been assigned to.</p>
<p>The decision to focus on the prevention and thus elimination of malaria
via the most cost-effective means available at the time and now is one
that must have been made prior to the founding of the foundation in
2004. It may have been that the process that led to the decision was as
powerful as that of GiveWell and thus yielded similar results, but it is
probably more parsimonious to assume that there was an amount of luck
involved and that I would be writing about another organization now had
that luck not favored the founders of <span class="caps">AMF</span>.</p>
<p>Apart from this problem of cause selection, however, many of the themes
introduced above also mark the decision processes of <span class="caps">AMF</span>. The founder has
provided me with a prioritized list of the factors that they routinely
consider as part of this process. I will explain them in order.</p>
<aside>
There are large, multilaterally endowed
funders of net distributions, such as the Global Fund, and to an
individual donor it is not obvious that such a funder will not observe
that a region where it planned to conduct a distribution has already
been served by <span class="caps">AMF</span> so that it will cancel its plans.
This would not be a problem if the donor could assume that, in turn, the
use of the Global Fund’s money in the case where it does not benefit the
distribution is roughly as good as <span class="caps">AMF</span>’s use of the donation. But while the
Global Fund is highly transparent and invests into a portfolio of
research-backed medical interventions that are considered highly
cost-effective by developed-world standards, it yet contains
interventions that are markedly less cost-effective than <span class="caps">LLIN</span> distributions
even under the unrealistically favorable conditions of randomized
trials.
</aside>
<dl>
<dt>Malaria burden.</dt>
<dd>
<p>Of course, in order to be considered, a region needs to have a high
malaria burden. <span class="caps">AMF</span> used to demand from its distribution partners that they
collect data on the rates of malaria cases from health centers in the
region, but the quality of the data, when it was even available, was
consistently low and the effort of collecting it high, so that <span class="caps">AMF</span> has more
recently made exceptions to this requirement. GiveWell has also ceased
to track these data.[^GiveWellAMF bullet point “Malaria case rate data”]
There is no question that malaria needs to be a “material problem for
public health”<sup id="fnref:16"><a class="footnote-ref" href="#fn:16">16</a></sup> in a region for <span class="caps">AMF</span> to consider it, but the value of
information of the most precise data available on the rate does not
warrant its collection.</p>
</dd>
<dt>Net gap.</dt>
<dd>
<p>The second most important criterion is the magnitude of the net gap, the
number of nets needed in a region. Here <span class="caps">AMF</span> distinguishes two categories of
distributions, strategic and tactical ones. Strategic means that the
national malaria control program of a country seeks funders for
distributions two to four years down the line; tactical means that a
distributions by other funders have left regions unprotected, usually
for lack of funds, so that <span class="caps">AMF</span> fills these gaps to provide closer to
universal coverage. These distributions are more short term.</p>
</dd>
<dt>Accountability.</dt>
<dd>
<p>Another critical criterion is whether the national malaria control
program welcomes the enhanced accountability measures that <span class="caps">AMF</span> demands. For
a discussion of a related problem, see the paragraph on “trusted
partners” below.</p>
</dd>
<dt>Non-net costs.</dt>
<dd>
<p>Transport and distribution of the nets is of course associated with
significant costs, not even taking into account the cost of the nets
themselves. In the experience of the foundation, it is psychologically
beneficial to ask donors to pay for something very concrete like the
nets, even though the non-net costs may well be higher leverage given
that they are essential and yet constitute but a fraction of the net
costs. Hence the website of <span class="caps">AMF</span> is geared toward fundraising only for nets –
when making a donation one can choose whether to enter a monetary value
and have the number of nets it funds calculated or whether one wants to
enter a number of nets and have the monetary value calculated.</p>
</dd>
</dl>
<p>The founder told me that this has not been a problem for them in the
past as distribution partners are often able and willing to carry
non-net costs and because the foundation also contacts major donors and
individually seeks permission to convert their contributions into
unrestricted funding; these donors are usually eager to give their permission.</p>
<p>Certainly since the unrestricted donations of several tens of millions
of dollars from Good Ventures, <span class="caps">AMF</span> has the means to cover non-net costs when
its distribution partners can not, but historically, this has been a
critical factor. Since the recommendation by GiveWell the decision to
fundraise for restricted grants has sometimes been criticized as veiled
complicity in perpetuating the overhead myth,<sup id="fnref:17"><a class="footnote-ref" href="#fn:17">17</a></sup> but <span class="caps">AMF</span> avoids any more
explicit endorsement of it.</p>
<dl>
<dt>Trusted partners.</dt>
<dd>
<p>The last major factor is whether there is a distribution partner in the
country with whom <span class="caps">AMF</span> is comfortable working. In 2013 this posed a problem
when <span class="caps">AMF</span> tried for the first time to finalize a distribution of several
millions of nets, an endeavor for which they had to cooperate with
several other parties, some of whom did not fully comply with <span class="caps">AMF</span>’s
requirements during the proposal phase. The distribution fell
through.<sup id="fnref:AMFDelays2013"><a class="footnote-ref" href="#fn:AMFDelays2013">30</a></sup> At the time of writing, however, talks for
distributions of similar scale are progressing well.</p>
</dd>
</dl>
<p><img alt="“Distribution team sets an iron ledge over unstable bridge to help people cross by foot.” Photo and caption courtesy of AMF." src="/images/llin-visualizer/20068-distribution-photo.jpg"></p>
<dl>
<dt>Factors of lesser importance.</dt>
<dd>
<p>Besides the above factors, I encountered lesser factors that could
potentially influence prioritization decisions. Rob Mather, however,
told me that they have little overall influence.</p>
</dd>
</dl>
<p>First, since natural immunity can be lost after years of not being
exposed to malarial mosquitoes, I imagined that redistributions in
regions where it can be assumed, after more than three years, that most
nets are in bad condition might take precedence over fresh distributions
in regions where no nets had previously been distributed. But according
to <span class="caps">AMF</span> this has not been an issue that they considered so far. Natural
immunity does not fully account for the drop in the fatality rate around
that age. Rather it is due to many factors “including better
physiological reserves, immunity, and access to
treatment,”<sup id="fnref:GWLengeler2015"><a class="footnote-ref" href="#fn:GWLengeler2015">67</a></sup> so that the malaria case rate may increase
(though the consensus view is that it tends not to) but without an
accompanying increase in mortality.</p>
<p>Second, the problem with counterfactual validity that I also consider in
section [subsec:trust] only becomes relevant when there is significant
funding available compared to the funding gap. For a hypothetical
explanation, see the aside on counterfactual validity.</p>
<p>Following that reasoning, I considered that it might also be relevant
for <span class="caps">AMF</span> to make sure that the regions it serves would not have been served
by any other funder had <span class="caps">AMF</span> not intervened.</p>
<p>However, one important assumption mentioned above, that the funding is
significant compared to the funding gap, is not given in the case of <span class="caps">LLIN</span>
distributions, so that <span class="caps">AMF</span> considers this a minor factor even in strategic
distributions. (That it is not a factor in tactical distributions holds
by definition.) <span class="caps">AMF</span> estimates the annual funding need at \<span class="math">\(5 billion and the
available funding from all funders at \\)</span>2.5 to \$3 billion,<sup id="fnref:18"><a class="footnote-ref" href="#fn:18">18</a></sup> so it
is unlikely that such a replacement effect will occur. <span class="caps">AMF</span>, however, notes
that this might happen at lower geographic levels, for example within
countries. Funding, however, is usually budgeted at the large funders
that fund net distributions, so that even if such replacement takes
place, the funds would be spent with roughly equivalent
cost-effectiveness on another distribution.</p>
<p>Third, insecticide resistance has not yet been a factor to decide where
to conduct a distribution or whether to conduct one in a specific region
but rather to decide on the type of net to use. Rob Mather specifically
pointed out that in regions with higher levels of insecticide resistance
they use nets treated with piperonyl butoxide (<span class="caps">PBO</span>) in addition to the
insecticide, but also noted the experimental nature of this very recent
technology.<sup id="fnref:WHO2015"><a class="footnote-ref" href="#fn:WHO2015">155</a></sup></p>
<blockquote>
<p><span class="caps">PBO</span> is not an insecticide by itself, but acts by inhibiting certain
metabolic enzymes (e.g., mixed-function oxidases) within the mosquito.
These enzymes detoxify or sequester insecticides before they can have a
toxic effect on the mosquito. Therefore, a <span class="caps">PBO</span> <span class="caps">LLIN</span> would in theory have an
increased killing effect on malaria vectors that harbour such resistance
mechanisms compared to a pyrethroid-treated <span class="caps">LLIN</span>. However, the entomological
and epidemiological impact of <span class="caps">PBO</span> LLINs is expected to vary according to the
bioavailability and retention of <span class="caps">PBO</span>, and the level, intensity and
mechanisms of insecticide resistance for local vectors, as well as
across different transmission settings.</p>
</blockquote>
<p><span class="caps">PBO</span> is crucial because while there are different insecticides (notably
permethrin and deltamethrin), both of them are insecticides of the same
class, pyrethroid, and mosquitoes typically form resistance to a class
of insecticide rather than to individual chemicals. Research on a new
class of insecticide is ongoing.</p>
<p>Fourth, lacking infrastructure in a country has not thwarted
distributions in the past. A good partner, good planning, and patience
are key. “For example, the Democratic Republic of Congo (<span class="caps">DRC</span>) is one of the
most challenging places in the world to work given ten years of war in
recent times, the scale and geography of the country amongst other
reasons. However, we recently demonstrated with considerable (but not
perfect) success, that it is possible to use smartphone technology to
collect data from 250,000 households and produce accurate data to
support a net distribution, including knowing the location of all
households to within 6 metres.”</p>
<p>Fifth, <span class="caps">AMF</span> agreed that the political situation in a country can at times
convince a partner that it would not be safe for them to conduct a
distribution, so that it needs to be postponed.</p>
<aside>
Some people object to <span class="caps">AMF</span>’s program on the
basis that saving lives in general anywhere in the world is, supposedly,
of negative net value.
It is highly unclear whether the expected negative effect, if any, of an
additional person on all other sentient beings is sufficient to offset
any positive net value the life has for the person, and the relative
importance of human overpopulation is doubtful as
well.[^MonbiotOverpopulation] But these critics are correct in that an
increase in population may result – an increase, however, that is likely
to be smaller and less certain than intuitively apparent.
Roodman has conducted a metastudy of fourteen studies
of the influence of life-saving interventions on population
developments. His conclusion is that: “The best interpretation of the
available evidence is that the impact of life-saving interventions on
fertility and population growth varies by context, above all with total
fertility, and is rarely greater than 1:1.”[^RoodmanMortalityFertility]
Other highly effective interventions, such as deworming, improve rather
than save lives, and are thus immune to this criticism.
</aside>
<h3 id="sec:llin-distributions"><a class="toclink" href="#sec:llin-distributions">Distributions of Long-Lasting Insecticide-Treated Mosquito Nets</a></h3>
<p>This section will give a brief overview of the intervention the Against
Malaria Foundation implements.</p>
<h3 id="subsec:epidemiology"><a class="toclink" href="#subsec:epidemiology">Key Epidemiological Facts</a></h3>
<p>Malaria is caused by the parasite Plasmodium, of which the form
Plasmodium falciparum is the most common one in Africa. It is not
contagious but rather is spread almost exclusively by female Anopheles
mosquitoes, which bite mainly during the night. While there are strains
of malaria that affect other animals and humans can be infected with
these strains in rare occasions, there are no known cases of these
zoonotic strains then spreading to another human.<sup id="fnref:WHOMalaria"><a class="footnote-ref" href="#fn:WHOMalaria">151</a></sup>
Furthermore, the Anopheles species that are the primary malaria vectors
in Africa are strongly anthropophilic, that is, have a strong preference
for human blood over the blood of other animals.<sup id="fnref:CDCMalaria"><a class="footnote-ref" href="#fn:CDCMalaria">43</a></sup></p>
<p>The most at-risk groups are pregnant women and children under the age of
five because in both cases children are in danger that have not yet
developed natural immunity to malaria. This immunity usually forms
around the age of five and leads to few lethal infections at older ages.
It is, however, important to note that this immunity can also be lost
over time when no reinfections occur.<sup id="fnref:webb2009long"><a class="footnote-ref" href="#fn:webb2009long">149</a></sup></p>
<p>While there are a few chemicals that can be used as prophylaxis against
malaria, there is no market-ready vaccine to date. The development of
vaccines against malaria has proved challenging for numerous reasons,
among them the high degree of polymorphism of P. falciparum proteins and
the high rate of reproduction, which leads to a rapid evolutionary
development of resistance.<sup id="fnref:crompton2010advances"><a class="footnote-ref" href="#fn:crompton2010advances">48</a></sup> As a result, the
history of malaria control has largely been a history of malaria vector control.</p>
<h3 id="subsec:history"><a class="toclink" href="#subsec:history">Brief History</a></h3>
<p><img alt="Many developed countries were formally malarious. (Image courtesy of Rigg et al.)<span
data-label="fig:malaria-history"></span>" src="/images/llin-visualizer/world-map-of-past-and-current-malaria-prevalence-world-development-report-2009.png"></p>
<aside>
Critics sometimes object that killing mosquitos to
save humans is speciesist, especially when enough mosquitoes have to die
to save one human that the years of life lost for the mosquitos combined
outweigh the years of human life saved. According to a quick
“back-of-the-envelope” estimate based on data from the <span class="caps">CDC</span>, GiveWell,
and, <span class="caps">WHOPES</span>[^GuesstimateMosquitoYLL] this is typically the case, with the years
of mosquito life lost outweighing human life years saved by three orders
of magnitude – and that is only counting the directly affected first
generation, not their lost offspring.
What the view above ignores, however, is that mosquitoes may not, on
balance, have positive lives, because they are very short and usually
end painfully. Tomasik an advocate for wild animals
including insects, thinks that the jury is still out on whether
insecticides as well as more human methods of vector control are net
good as they reduce populations and thus spare future generations of the
species a net negative life, or whether the direct suffering and life
lost are worse. Hence, he argues for the usage of *humane*
insecticide.[^TomasikInsecticides]
</aside>
<p>Malaria has existed for at least one hundred thousand years, but it only
spread to humans a few thousand years ago. It was mentioned in ancient
Chinese texts of 2550 <span class="caps">BCE</span>, in ancient Egyptian texts of 1325 <span class="caps">BCE</span>, and ancient
Indian texts of around 500 <span class="caps">BCE</span>. The last author already correctly
identified insects as the vectors of malaria. Hippocrates disagreed and
thought weather, dirty drinking water, or poor health led to malaria.</p>
<p>The ancient Romans finally gave rise to the term <em>malaria</em> that we use
today. They correctly associated malaria with marshes and swamplands,
but thought it was the fetid gases from these areas that caused the
disease, so that their term for “bad air,” <em>mala aria</em>, gained currency.
Malaria took a heavy toll on the Roman empire, and by weakening Rome’s
soldiers and civilians, may have contributed to its
collapse.[^goldsmith2010battling 9–13]</p>
<p>In the 1940s and 1950s, large-scale and well-funded campaigns eliminated
malaria in the <span class="caps">US</span> and much of Europe. They involved the drainage of
stagnant water, installation of window screens, and spraying of
<span class="caps">DDT</span>.[^marcus2009malaria 72][^rigg2009world 117] <span class="caps">DDT</span> was later deemed too
dangerous and banned, but mosquitoes also developed resistance to it, so
even a lifting of the ban would have limited positive
effects.[^marcus2009malaria 68]</p>
<p>In the following, I want to focus on the one form of vector control
relevant to this thesis, the distribution of long-lasting
insecticide-treated bed nets. Nets have been used for protection against
mosquitoes since around 500 <span class="caps">BCE</span>, but only during World War <span class="caps">II</span> the practice
of impregnating nets gained currency.</p>
<p>Mosquitoes would often crawl underneath the rims of the nets, find rents
in them, bite body parts that were touching the net, or would even be
trapped underneath the net from the start. Insecticide-treated nets do
not have these disadvantages because mosquitoes would either die or at
least be repelled by the insecticide, and in the latter case the
insecticide may curtail their life expectancy to the point that the
malaria parasite will not have time to fully develop before the death of
its host.</p>
<p>These insecticide-treated nets formed a key part of the Roll Back
Malaria program of the World Health Organization (<span class="caps">WHO</span>), the United
Nations Children’s Fund (<span class="caps">UNICEF</span>), the World Bank, and the United Nations
Development Programme (<span class="caps">UNDP</span>).[^packard2010making 228–229] Yet they had
the problem that they needed frequent retreatment with fresh insecticide.</p>
<p>This situation was greatly improved when LLINs, long-lasting
insecticide-treated bednets were developed and the <span class="caps">WHO</span> shifted their
investments toward the new models in starting in 2002. These nets stay
impregnated for their full lifetime of about two to three years. This
improvement instated them clearly as the currently most cost-effective
means of malaria prevention.<sup id="fnref:Masum2010"><a class="footnote-ref" href="#fn:Masum2010">102</a></sup></p>
<p>In 2011, however, the Global Fund, one major distributor of LLINs, was hit
by the global financial crisis as well as its discovery that \$25
million had vanished from some of its community
programs.<sup id="fnref:TerraDaily2011"><a class="footnote-ref" href="#fn:TerraDaily2011">140</a></sup> Such fluctuations are particularly dangerous
because the protection adults had enjoyed up until then may have
undermined their natural immunity, so that timely redistributions are critical.</p>
<p>Another important development was that originally the <span class="caps">WHO</span> recommended
targeted distributions where the groups most at risk of malaria,
pregnant women and children under five, would receive nets in
distributions. In 2007, however, the <span class="caps">WHO</span> made the switch to recommending
universal distributions.<sup id="fnref:WHO2007"><a class="footnote-ref" href="#fn:WHO2007">154</a></sup> In addition to the protection LLINs lend
to the individual, the <span class="caps">WHO</span> also views their distribution as a vector
control intervention with considerable communal effects. Humans can be
seen as bait that lures mosquitoes to the protected sleeping spaces
where the insecticide kills them. Surviving mosquitoes, at least the
anthropophilic species, who avoided contact with LLINs or have developed
resistance to the insecticide, may still die of lack of feeding. Thus
universal distributions can reduce the mosquito population below some
critical threshold that will bar the parasites from completing their
life cycle, thus eliminating malaria. The mechanism requires that almost
all people in a region be protected by LLINs. Hence, universal distributions
have to be conducted with great attention to detail.</p>
<p>For further reading I recommend as the most comprehensive volume on
recent malaria elimination efforts that I have found during my
research.<sup id="fnref:webb2014long"><a class="footnote-ref" href="#fn:webb2014long">150</a></sup> gives a wider perspective on the history of the
disease.<sup id="fnref:packard2010making"><a class="footnote-ref" href="#fn:packard2010making">117</a></sup></p>
<h3 id="subsec:role-amf"><a class="toclink" href="#subsec:role-amf">Role of the Against Malaria Foundation</a></h3>
<p><img alt="“GPS coordinates of each house that received nets were recorded and uploaded each day.” Photo and caption courtesy of AMF." src="/images/llin-visualizer/20179-distribution-photo.jpg"></p>
<p><span class="caps">AMF</span>, founded in 2004, has been a rather minor funder of LLINs, but it is now
rapidly scaling up. Its small scale has enabled it to pilot a monitoring
methodology that was unprecedented by distributions outside scientific
trials. This is a crucial innovation for three reasons.</p>
<p>First, adults tend to underestimate the harm of the disease; since it is
rarely deadly for them, they neglect prevention and rather buy
treatment, which is more expensive in expectation. The psychology of
these nonoptimal decisions has been analyzed in .[^banerjee2012poor 59
ff.] Hence, education is a critical part of every net distribution, and
so is regular monitoring of the actual behavior of the recipients
afterwards to detect weaknesses in the education model.</p>
<p>Second, corruption is widespread in many countries that suffer from a
great malaria burden. Agreements on strong monitoring regulations ward
off corruption by making the distribution unappealing to corrupt
decision makers in potential partner organizations.</p>
<p>And third, the rigorous monitoring yields comprehensive measurements on
many proxy variables for impact – data with high informational value for
prioritization research.</p>
<p>“As our own species is in the process of proving, one cannot have
superior science and inferior morals. The combination is unstable and self-destroying.”</p>
<h2 id="ch:requirements-and-modeling"><a class="toclink" href="#ch:requirements-and-modeling">Requirements and Modeling</a></h2>
<h3 id="sec:abduction"><a class="toclink" href="#sec:abduction">Forming the Idea</a></h3>
<p>The motivation behind the specific approach I applied to deciding on the
topic of this thesis was that I wanted to find the idea that provided
the best fit for the requirements of the university while having the
maximal positive impact on the world. To this end, I contacted twelve
organizations working on pressing and neglected problems and described
my circumstances to them so they could determine whether there were any
projects on their backlogs that might fit my requirements.</p>
<dl>
<dt>Relevant attributes.</dt>
<dd>
<p>The first step toward forming the idea for this thesis was to define the
relevant attributes of any proposal along which they would be
prioritized. The final set was one consisting of six factors, whose
estimation was necessarily highly informal, informed by the proposal of
the organization and my own critical assessment. The estimates
distinguish three levels indicating high fit, low fit, and uncertainty.</p>
</dd>
<dt>Impact.</dt>
<dd>
<p>The potential for positive impact that I thought the proposal
promised after hearing the arguments of the proposing charity.</p>
</dd>
<dt>Replaceability.</dt>
<dd>
<p>The ease with which the charity could realize the project without
my help. If they would be able to do it themselves at a relatively
small cost, it would be incompatible with a high estimate
for impact.</p>
</dd>
<dt>Solidity.</dt>
<dd>
<p>A measure for how certain the project is to succeed. A software
development project is comparatively likely to succeed while an
exploratory data analysis might yield no interesting insights and a
project that depends on forming a community of users might easily
fail to materialize.</p>
</dd>
<dt>University fit.</dt>
<dd>
<p>The degree to which the proposal fits the requirements of
the university.</p>
</dd>
<dt>Personal fit.</dt>
<dd>
<p>The degree to which the proposal fits with my abilities.</p>
</dd>
<dt>Personal potential.</dt>
<dd>
<p>The degree to which the proposal fits with my future plans.</p>
</dd>
</dl>
<p>Four states emerged from the correspondence with the charities:</p>
<dl>
<dt>No reply.</dt>
<dd>
<p>Two of the organizations did not reply despite reminders.</p>
</dd>
<dt>Replied.</dt>
<dd>
<p>Four of the organizations replied, but the discussion did not
advance further. One, for example, told me they had a project with
good fit in their pipeline, but asked me to answer a few
questions first. I did, but the contact person left the organization
before replying, which I found out when I tried to contact other
people of the organization later on.</p>
</dd>
<dt>No need.</dt>
<dd>
<p>Three organizations discussed the idea in their team but found that
they had no projects they could use my help for.</p>
</dd>
<dt>Delivered.</dt>
<dd>
<p>Three organization had proposals for me and invited my feedback.</p>
</dd>
</dl>
<p>One hurdle had been that most of these organizations work on problems
where volunteers are generally not worth the staff time necessary to
train them, and that insofar as they do draw on volunteer work, it is
for clearly delimited tasks that require either very little training or
training that has been captured in writing.</p>
<p>Among the three responses was one pure software project that would not
have satisfied university requirements. The most interesting proposals
came from Animal Charity Evaluators and the Against Malaria Foundation.</p>
<p>The first group presented me with eight ready proposals that they were
looking for volunteers to realize. Three of them scored low on solidity
as they relied on data that would have to be obtained from corporate
entities. It seemed highly uncertain to me whether I would be able to
obtain these data even after extensive correspondence with a wide range
of them. Three more were proposals for studies with human subjects and
required significant amounts of contact with these subjects. Since
important decisions will be based on the results and I have great
respect of the difficulties of this study format combined with little
prior experience, I felt a low personal fit. Finally two studies scored
high on all counts, a study of historical influence of advocacy
campaigns on legislation and a study on the effectiveness of specific
types of social media advocacy. Below the gists:</p>
<p><strong>Legislation:</strong> Study the impact of legislative advocacy by determining
the total amount of funding spent lobbying and advertising for
pro-animal legislation in successful and unsuccessful campaigns and
comparing this to the estimated amount of suffering saved by successful
campaigns and the proportion of funding that went to successful campaigns.</p>
<p><strong>Social Network Spread:</strong> Track the flow of common animal advocacy
messages through online or in-person social networks to compare speed of
dispersal. Possible axes of comparison include positivity/negativity,
system/individual focus, and specificity.</p>
<p>The Against Malaria Foundation (<span class="caps">AMF</span>) provided the proposal for analysis of
their electronically collection distribution data, whose nature – in its
developed state – will become evident from the remainder of this section.</p>
<p>Additionally, I generated an idea for a <a href="/concept-for-donor-coordination.html">donor coordination system</a> that
piqued the interest of the Centre for Effective Altruism, a <span class="caps">UK</span> nonprofit.</p>
<p>I investigated all three remaining proposals. With the study on
legislative changes I identified the following problems:</p>
<p>It was unclear to me what degree of legal knowledge would be necessary
to assess the impact of legislative changes.</p>
<p>Even if the required legal knowledge was minimal, assessing levels of
animal well-being is fraught with uncertainties from domains including
psychology, biology, medicine, economics, and philosophy. I feel like I
have a sufficient grasp on the extent of these uncertainties not to make
premature judgments on related issues, but conversely the uncertainties
will only in rare cases allow me to make confident judgments. Since most
legal changes that I am aware of from the recent past were of small,
incremental nature, finding the ones that do allow for unequivocal
assessments with small margin of error would be hard and might not be
achievable in a reasonable timeframe.</p>
<p>Even if I manage to locate unequivocally successful changes in
legislation that I can quantify with low margins of error, the process
of attributing these to specific advocacy campaigns will be hard or
impossible in most cases since there are usually different groups
working towards subtly different ends with different levels of
efficiency, all of which would require assessing.</p>
<p>In those cases where all previous requirements are met, including a
relatively clear attribution to one or few campaign efforts, I would
still need to obtain data on the funding that benefited the campaign or
campaigns, data that may be confidential or, after a longer time, hard
to reconstruct.</p>
<p>Even if there are campaigns that meet all these requirements,
conclusions from this small and highly selective set would probably not
extrapolate well to future campaigns.</p>
<p>The study on social media spread seemed easier to realize. The only
problems I discovered were:</p>
<p>There is already a significant amount of data and research on social
media spread that might extrapolate fairly well to the specific types of
messaging that the study would have prioritized.</p>
<p>Running the study would require access to social media accounts with
sufficient reach and a representative base of followers. The
organization might be able to provide me with the first, but its
followers are by no means representative of the larger population.</p>
<p>The following are concerns I had about the proposal of <span class="caps">AMF</span>:</p>
<p>It addresses a highly specific set of problems, which limits my optimism
about its impact.</p>
<p>Many of its features can only be used if there is sufficient data at a
level that so far few local partners of <span class="caps">AMF</span> have agreed to collect. If this
highly detailed level of data collection should remain an exception in
the future, many features of the software will find little use.</p>
<p>In extension of the previous point, some features are still not in use
because the data is not yet available.</p>
<p>My own idea has the following disadvantages, the reason I am only
developing the plan pending such a time when I have the support of an
established organization to drive it further:</p>
<p>It depends on the continued cooperation of a sizable community of donors.</p>
<p>It depends on the continued cooperation of many nonprofits.</p>
<p>Some of its failure scenarios are silent, meaning that it is hard to
know whether they have occurred.</p>
<p>Given these considerations, I decided to prioritize the project of <span class="caps">AMF</span> and
complement it with my plans for the <a href="/concept-for-donor-coordination.html">donor coordination solution</a>.</p>
<h3 id="sec:data-exploration"><a class="toclink" href="#sec:data-exploration">Data Exploration</a></h3>
<p><img alt="image" src="/images/llin-visualizer/odk-gps.png"></p>
<p>Having obtained the first set of distribution data from <span class="caps">AMF</span>, I investigated
the nature of the data with a specific focus on its collection, format,
peculiarities (possible errors), and unequivocally erroneous entries.</p>
<p>The first set of data was predistribution registration survey (<span class="caps">PDRS</span>) data of
the distribution of 676,000 long-lasting insecticide-treated nets (LLINs) in
Kasaï Occidental, Democratic Republic of the Congo (<span class="caps">DRC</span>), funded by <span class="caps">AMF</span> in
August to November 2014. Their local partner that conducted the
distribution, <span class="caps">IMA</span> World Health, collected the survey data using an
OpenDataKit-based electronic system.</p>
<p><span class="caps">AMF</span> conducts universal distributions. (The distinction between universal and
targeted distributions is explained in section [subsec:history].) To
gauge the demand of nets per region and thus the nets that need to be
available at each distribution hub, they have their local partner
organizations conduct predistribution registration surveys, which
determine the number of sleeping spaces per household and the number of
existing and usable LLINs.</p>
<h3 id="subsec:survey-form"><a class="toclink" href="#subsec:survey-form">The Survey Form</a></h3>
<p>In the present case, the distribution partner asked for a number more
data points in the survey, which are not of primary interest to <span class="caps">AMF</span>.
Nonetheless these are interesting data, and <span class="caps">AMF</span> informed me that the
government has indicated interest in the data to augment its dated
census data.</p>
<p>I performed my initial analyses in R, but soon transferred my findings
to Python, where they are now implemented as part of cleaning and
analysis routines.</p>
<dl>
<dt>Language.</dt>
<dd>
<p>The survey data consists of 57 variables, of which at least 10 are
determined automatically. The French versions are authoritative but
since I do not know French and also <span class="caps">AMF</span> prefers to work with English labels,
they created a spreadsheet of translations; many of the English labels
seemed automatically translated to me, so I asked a French native
speaker to kindly correct them for me. The original French versions and
the English translations are given in appendix [app:ima-2014-pdrs].</p>
</dd>
</dl>
<p>A native speaker confirmed that “Est ce que votre ménage a au moins une
Moustiquaire” asks whether there is “at least one mosquito net” as
opposed to “more than one.” In the spreadsheet, “moins_une_MILD_HH”
is translated as “More than one net,” but the same native speaker said
that “moins_une_MILD_HH” alone is too compressed to make sense either
way. The meaning of this variable is hence almost certainly “at least one.”</p>
<dl>
<dt>Missing questions.</dt>
<dd>
<p>In several cases the questions associated with certain variables are not
given in the spreadsheet, which may be due to them being answered in
advance by the interviewer rather than the household representative. The
interviewer may know what to answer from a preparatory briefing.</p>
</dd>
<dt>Restrictive response options.</dt>
<dd>
<p>The final problem that I see with the survey is that it asks for the
source and the brands of the preexisting nets, but if a household has
several preexisting nets from different sources, then they may also be
of different brands, and both questions will become hard to answer. This
is true of 106,824 responses in the sample.</p>
</dd>
</dl>
<p>Oddly, the question asking for the sex of the respondent had response
options in English rather than French in the original and did not allow
for intersex options.<sup id="fnref:19"><a class="footnote-ref" href="#fn:19">19</a></sup> It also only asked of the representative of
the household, so that the sex of the other members remains unknown.</p>
<h3 id="subsec:survey-data"><a class="toclink" href="#subsec:survey-data">The Survey Data</a></h3>
<p><img alt="The software tries to prevent, catch, or ask for confirmation for
unusual inputs.<span
data-label="fig:preventing-errors"></span>" src="/images/llin-visualizer/odk-preventing-errors.png"></p>
<p>Although there were some problems with the data, as detailed below, most
of the data were of sufficient quality. The government of the <span class="caps">DRC</span> even
indicated interest in the data as they are more up-to-date than the only
census data that it has available.</p>
<dl>
<dt>Invalid rows.</dt>
<dd>
<p>I found that when date, one of the <span class="caps">GPS</span> coordinates, the confirmation of
the legal script, and the day of the distribution were missing from
rows, that they were either empty or invalid.</p>
</dd>
</dl>
<p>Most of the 9938 rows fell into the first category, but four seemed to
be aggregates of other rows, claiming to be samples of households with
several hundreds of thousands of tenants. I filter all of these rows
during the import.</p>
<dl>
<dt>Dates.</dt>
<dd>
<p>The dates of the rows are broken in so many ways that it seems likely
that several problems coincided here to produce them. My findings are
detailed in section [subsec:learnings-dates].</p>
</dd>
</dl>
<p>[16]<img alt="image" src="/images/llin-visualizer/odk-sleeping-spaces.png"></p>
<dl>
<dt>Opt-outs.</dt>
<dd>
<p>Only 516 (0.2%) of the respondents opted out of the survey and thus
probably the distribution.</p>
</dd>
<dt>Age.</dt>
<dd>
<p>The age distribution was significantly skewed upwards in comparison to
that of the <span class="caps">DRC</span> overall<sup id="fnref:CIA2015"><a class="footnote-ref" href="#fn:CIA2015">44</a></sup> with a median age of 33 rather than 18,
surely because the respondents were the people the household sent to
represent them and not a random sample.</p>
</dd>
<dt>People per household.</dt>
<dd>
<p>There are a few errors up to almost 241,000 people per household, so the
mean is skewed upward, but the median of the people per household is 5
with a median absolute deviation of 3.</p>
</dd>
</dl>
<h3 id="subsec:forthcoming-data"><a class="toclink" href="#subsec:forthcoming-data">Forthcoming Data</a></h3>
<p>Planning the <span class="caps">LLIN</span> Visualizer, I already made provisions for
postdistribution check-up (<span class="caps">PDCU</span>) data to be imported into the database, but
<span class="caps">AMF</span> has not yet been able to send me the raw data.</p>
<p>The data is intended to shed light on the actual usage of the nets. To
this end, PDCUs are conducted in six-month intervals after every
distribution. Five percent of the recipients are randomly selected and
surveyed, a process during which condition and usage of the nets are determined.</p>
<p>In order to find the households and to match them to the respective rows
in the <span class="caps">PDRS</span> data set, the houses are marked. This matching may be imperfect
if the marks get removed or the village moves due to changes in seasonal
labor. A separate household relation in the <span class="caps">LLIN</span> Visualizer is meant to
provide a flexible fuzzy mapping between survey rows to link responses
of the same household at different points in time.</p>
<h3 id="subsec:other-data"><a class="toclink" href="#subsec:other-data">Other Data</a></h3>
<p><img alt="image" src="/images/llin-visualizer/net-analytics-hulls-view.png"></p>
<h4 id="the-world-map"><a class="toclink" href="#the-world-map">The World Map</a></h4>
<p>All the data in the <span class="caps">LLIN</span> Visualizer is shown on a map whose data is
provided by one of two sources that the user can choose from, the
standard map of OpenStreetMap and the humanitarian version of it.<sup id="fnref:20"><a class="footnote-ref" href="#fn:20">20</a></sup></p>
<h3 id="insecticide-resistance"><a class="toclink" href="#insecticide-resistance">Insecticide Resistance</a></h3>
<p>The vector species Anopheles has a quick reproductive cycle, so that it
can produce resistance to insecticide quickly. The mosquitoes never
cross any wide, barren areas, so that resistance forms in independent
localized areas, but it is nevertheless a significant and insufficiently
solved problem.</p>
<p>Various organizations conduct tests for insecticide resistance, and the
website <span class="caps">IR</span> Mapper<sup id="fnref:IRMapper"><a class="footnote-ref" href="#fn:IRMapper">87</a></sup> provides an overview of the last sixty
years of such tests. the <span class="caps">LLIN</span> Visualizer displays this data and allows
for filtering by country, insecticide class, and year range, since <span class="caps">AMF</span>
only uses insecticides of the Pyrethroids class, and because insecticide
resistance tests that were older than a decade at the time of the
distribution are probably obsolete.</p>
<h3 id="incidence-and-endemicity"><a class="toclink" href="#incidence-and-endemicity">Incidence and Endemicity</a></h3>
<p>The <span class="caps">LLIN</span> Visualizer is indebted to <span class="caps">IR</span> Mapper also for the data on
Plasmodium falciparum and Plasmodium vivax incidence rates and the
entomological inoculation rate – the number of infective bites per
person per time. <span class="caps">IR</span> Mapper, in turn, obtained these data from the
Malaria Atlas Project (<span class="caps">MAP</span>). The data has a high resolution even though
the project could draw on fewer measurements. It achieved this by
augmenting the measurements with a model that took into account
“environmental covariates [such as] rainfall, temperature, land cover,
and urban/rural status.”<sup id="fnref:MalariaAtlasProjectZambia"><a class="footnote-ref" href="#fn:MalariaAtlasProjectZambia">99</a></sup></p>
<h3 id="net-gaps"><a class="toclink" href="#net-gaps">Net Gaps</a></h3>
<p><img alt="image" src="/images/llin-visualizer/net-analytics-net-gaps.png"></p>
<p>The net gap is the difference between the number of nets that are needed
in a given country and year and the number of nets that are already
funded for that country and year. These are laboriously compiled by
entities such as the Roll Back Malaria collaboration<sup id="fnref:RBMNetGap"><a class="footnote-ref" href="#fn:RBMNetGap">128</a></sup>.</p>
<p><span class="caps">AMF</span> provided me with two of these tables, one with estimates from 2013
for 2013–2016 – including nets needed, funded, and their difference –
and one newer one from 2015 with estimates for 2015–2017 but for fewer
countries and only including net gaps.</p>
<p>The <span class="caps">LLIN</span> Visualizer allows for the essential filtering by year, but it
also provides two different view modes. Malaria does not respect country
borders, so from the medical point of view, the aggregation by country
is arbitrary, so usually what users will be interested in are the
absolute net gaps, answering for example questions as to how much bigger
the net gap was than a distribution in a country in a certain year. But
malaria eradication is also fought in the political area, and
distribution partners of <span class="caps">AMF</span> are only active in some countries, so the
relative gaps of countries are also important. Countries with a larger
population will tend to have greater absolute net gaps, but maybe it is
also better served by NGOs or by its government, so that the relative
net gap is not unusually high. The populous Nigeria and Ethiopia have
large absolute and relative net gaps, for example, but according to the
2015 data, no nets at all were funded for Malawi resulting in a net gap
of about 1.7 million which would not have been noticeable on the
absolute overview. That net gap is tiny in comparison to Nigeria and
Ethiopia, but at the time of the estimate, <span class="caps">AMF</span> had never conducted a
distribution of even half that size. Yet they have a particularly
well-trusted partner in Malawi, Concern Universal, so that they are in a
perfect position to close this gap – in a much better position, surely,
than in Nigeria, where they conducted a single tiny distribution in
early 2008 with a partner they have only cooperated with once.</p>
<h3 id="indicators"><a class="toclink" href="#indicators">Indicators</a></h3>
<p>One piece of feedback I have gotten on the <span class="caps">LLIN</span> Visualizer is that users
would like to compare the endemicity and <span class="caps">LLIN</span>-related data with generic,
frequently-used indicators. In particular the mean <span class="caps">GDP</span> per capita was
cited as an interesting one, but I have found many more interesting
measures recorded or calculated by the World Bank, Gallup, the Center
for Global Development, the World Happiness Report, and others. In
response, I extended the <span class="caps">LLIN</span> Visualizer to display
them.<sup id="fnref:GallupWorldPoll"><a class="footnote-ref" href="#fn:GallupWorldPoll">64</a></sup><sup id="fnref:WorldHappinessReport2016Update"><a class="footnote-ref" href="#fn:WorldHappinessReport2016Update">148</a></sup><sup id="fnref:WorldBankGini"><a class="footnote-ref" href="#fn:WorldBankGini">152</a></sup><sup id="fnref:GallupLifeLadder"><a class="footnote-ref" href="#fn:GallupLifeLadder">65</a></sup><sup id="fnref:MedianIncome"><a class="footnote-ref" href="#fn:MedianIncome">52</a></sup><sup id="fnref:MedianGDP"><a class="footnote-ref" href="#fn:MedianGDP">82</a></sup></p>
<dl>
<dt>Confidence in government</dt>
<dd>
<p>A measure of the people’s confidence in their respective national
government (not government in the abstract).</p>
</dd>
<dt>Delivery quality</dt>
<dd>
<p>Delivery of government services, comprising indicators of government
effectiveness, regulatory quality, rule of law, and control
of corruption.</p>
</dd>
<dt>Democratic quality</dt>
<dd>
<p>Quality of the democratic process, comprising indicators of voice
and accountability, and political stability and absence of violence.</p>
</dd>
<dt>Freedom to make life choices</dt>
<dd>
<p>Answers to the Gallup World Poll question “Are you satisfied or
dissatisfied with your freedom to choose what you do with your life?”</p>
</dd>
<dt>Generosity</dt>
<dd>
<p>Gallup World Poll data on donation behavior in proportion to <span class="caps">GDP</span>, so
that poorer people are not penalized for being proportionally
less giving.</p>
</dd>
<dt>Gini index</dt>
<dd>
<p>The Gini index is a measure commonly used to measure income
inequality.<sup id="fnref2:WorldBankGini"><a class="footnote-ref" href="#fn:WorldBankGini">152</a></sup> It provides an important addition to the
<span class="caps">GDP</span> as explained below.</p>
</dd>
<dt>Life expectancy at birth</dt>
<dd>
<p>It goes without saying that the life expectancy is not necessarily
an age at which many people die, because especially in many
sub-Saharan countries frequent child deaths skew the life expectancy
downward even though people are much less likely to die once they
cross a certain age threshold (around age five).</p>
</dd>
<dt>Life Ladder</dt>
<dd>
<p>The Life Ladder or Cantril Self-Anchoring Striving Scale is one type
of well-being assessment tool used by Gallup,<sup id="fnref2:GallupLifeLadder"><a class="footnote-ref" href="#fn:GallupLifeLadder">65</a></sup> who
provide the data for the whole past decade an almost all
recognized countries.</p>
</dd>
<dt>Mean <span class="caps">GDP</span> per capita</dt>
<dd>
<p>The mean gross domestic product per capita is such a well-known
metric that it hardly needs introduction. The data imported into the
<span class="caps">LLIN</span> Visualizer at the time of writing are the latest World
Bank estimates.</p>
</dd>
<dt>Median <span class="caps">GDP</span> per capita</dt>
<dd>
<p>One criticism of the <em>mean</em> <span class="caps">GDP</span> per capita is that it both hides
income and consumption inequeality,<sup id="fnref:21"><a class="footnote-ref" href="#fn:21">21</a></sup> and skews the result
upward.<sup id="fnref2:MedianIncome"><a class="footnote-ref" href="#fn:MedianIncome">52</a></sup><sup id="fnref2:MedianGDP"><a class="footnote-ref" href="#fn:MedianGDP">82</a></sup> The latter is a problem that the
median avoids. For a response to the first problem, see the Gini
index above. The Median <span class="caps">GDP</span> per capita data are available for all
the same countries and years that the mean <span class="caps">GDP</span> per capita figures
are also available for.</p>
</dd>
<dt>Most people can be trusted</dt>
<dd>
<p>The proportion of people who think that most people can be trusted
as opposed to “Can’t be too careful.”</p>
</dd>
<dt>Negative affect</dt>
<dd>
<p>The affect measures comprise self-reported worry, sadness, and anger
on the previous day.</p>
</dd>
<dt>Perceptions of corruption</dt>
<dd>
<p>The mean of questions from the Gallup World Poll about perceived
corruption in different parts of society.</p>
</dd>
<dt>Positive affect</dt>
<dd>
<p>The affect measures comprise self-reported happiness, laughter, and
enjoyment on the previous day.</p>
</dd>
<dt>Social support</dt>
<dd>
<p>Responses to “If you were in trouble, do you have relatives or
friends you can count on to help you whenever you need them, or
not?” from the Gallup World Poll.</p>
</dd>
<dt>World Happiness Report score</dt>
<dd>
<p>The World Happiness Report publishes a score that is the weighted
aggregate of many indicators selected to give an idea of the degree
of happiness of the denizens of 157 countries. The current figures
are the ones from the 2016 update, which also cover earlier years.</p>
</dd>
</dl>
<h3 id="sec:functional-requirements-net-analytics"><a class="toclink" href="#sec:functional-requirements-net-analytics">Functional Requirements</a></h3>
<h3 id="subsec:original-set"><a class="toclink" href="#subsec:original-set">The Original Set</a></h3>
<p>The architecture decisions underlying the <span class="caps">LLIN</span> Visualizer where shaped
by the following requirements:</p>
<ol>
<li>
<p>Any user will be able to view aggregated data of the net distributions.</p>
</li>
<li>
<p>Privileged users will be able to view maximally detailed net
distribution data.</p>
</li>
<li>
<p>Users will be able to view relevant statistics of the aggregated data.</p>
</li>
<li>
<p>Administrators will be able to import new data into the system.</p>
</li>
<li>
<p><span class="caps">AMF</span> will be able to embed the software into its website.</p>
</li>
</ol>
<h3 id="subsec:extension-set"><a class="toclink" href="#subsec:extension-set">The Extension Set</a></h3>
<p>I extended this set by the following optional requirements:</p>
<ol>
<li>
<p>Users will be able to view past distributions even when only aggregated
data is available on them.</p>
</li>
<li>
<p>Users will be able to put distribution data into the context of
geographical patterns of malaria incidence, insecticide resistance, and
unmet need for nets (net gaps).</p>
</li>
<li>
<p>Users will be able to filter distributions, incidence data, insecticide
resistance tests, and net gaps by their date and various data
type–specific attributes.</p>
</li>
<li>
<p>Maintainers will be able to extend the statistical analyses.</p>
</li>
</ol>
<p>My reasoning was the following:</p>
<ol>
<li>
<p>Even just superficial meta data of distributions – such as their date,
size, and geographic location – are helpful to understand prioritization
decisions of <span class="caps">AMF</span> and also how these changed over the years as the
organization gradually scaled up its operations.</p>
</li>
<li>
<p>Malaria incidence, insecticide resistance, and net gaps are important
elements in prioritization decisions. There are other factors that are
harder to get data on, such as the degree to which <span class="caps">AMF</span> trusts various
potential partner organizations in the respective countries (see section
[sec:strategy-amf] for more details), but visualizing these three
factors already provides value to the user.</p>
</li>
<li>
<p>These filters are important because time tends to gradually drive these
data toward obsolescence. <span class="caps">IR</span> tests that, at the time of the
distribution, were decades old should have little bearing on the
distribution. Similarly, the chemical classes of the insecticides used
have changed and will continue to change. As a result, resistance to one
class of insecticide has little bearing on the feasibility of a
distribution if <span class="caps">AMF</span> uses a different one.</p>
</li>
<li>
<p>The analyses that I implemented are those that seemed interesting to me
as well as the participants of a straw poll I conducted among interested
users, but it is likely that over time new insights will be desired.
Then the maintainer (likely me or someone at <span class="caps">AMF</span>) needs to be able to
implement these in some fashion that is perpendicular to most of the
<span class="caps">LLIN</span> Visualizer, so that performing these changes will not require
extensive knowledge of the code base.</p>
</li>
</ol>
<h3 id="sec:coordination-amf"><a class="toclink" href="#sec:coordination-amf">Coordination with <span class="caps">AMF</span></a></h3>
<p>I tried to minimize the time that my questions diverted from the staff
of the foundation, and so collected my questions so they could answer
them all at once or, even better, I found answers to them myself before
I even sent them off.</p>
<p>The questions fell into the categories of (1) questions about the data
that I had already received, (2) technical questions about the
realization of the software, and (3) content questions about topics that
I needed to address in this text.</p>
<p><img alt="“The distributors and local children pose for a photo after a successful distribution and hanging of the nets.” Photo and caption courtesy of AMF" src="/images/llin-visualizer/20280-distribution-photo.jpg"></p>
<dl>
<dt>Questions about the data.</dt>
<dd>
<p>The question about the odd behavior of the health area variable
addressed in section [subsec:survey-data] was one topic of discussion.
A related problem was the reverse one, that in some cases names, such as
village and health area names, were misspelled. I proposed a metric
based on geographic distance of the point clusters and edit distance of
the names, but the problem was not critical and we did not pursue it
further. Another factor was that we saw no way to obtain any canonical
spellings of the village names; in many cases there probably are no
agreed-upon versions.</p>
</dd>
</dl>
<p>Other topics concerned the availability of further data that I could
import or rather the reasons for the delays they are facing. Finally,
there was a data analysis task that we discussed so that I could perform
and integrate it into my work, but <span class="caps">AMF</span>’s head of technology eventually
completed the analysis himself.</p>
<dl>
<dt>Technical questions.</dt>
<dd>
<p>One discussion topic was that of the distinction of access groups; we
decided that three groups, anonymous visitors, logged-in users, and
administrators were sufficient.</p>
</dd>
</dl>
<p>The distinction between anonymous visitors and logged-in users is
necessary so that only an anonymized version of the data is publicly
available but <span class="caps">AMF</span> can manually give special access rights to employees or
people it trusts. In terms of the anonymization I proposed two methods,
either just barring anonymous users from accessing the most fine-grained
household-level layer for anonymous users or doing the first and making
sure that <span class="math">\(k\)</span>-anonymity and <span class="math">\(l\)</span>-diversity are maintained on the cluster
level (we did not discuss <span class="math">\(t\)</span>-closeness). <span class="caps">AMF</span> confirmed that prohibiting
access to the household layer was sufficient.</p>
<p>Maintenance was another topic of discussion. There were two possible
scenarios the choice between which we had initially left open: (1) <span class="caps">AMF</span> takes
over maintenance of the software and runs it on its servers, or (2) I
continue to maintain it and run it on one of my servers. Option 1 has
the advantages that the <span class="caps">AMF</span> staff – mostly their head of technology – has
full control over the software and can embed or adapt it in any way they
see fit without being dependent on me for it. The main disadvantage is
that the head of technology is not familiar with the software framework.
Option 2 avoids this disadvantage at the price of some level of
dependence. In the end it was option 2 that we agreed on.</p>
<p>Finally I made a number of suggestions for features that I envisioned
for the <span class="caps">LLIN</span> Visualizer and asked for critique, prioritization, and
additional ideas. The results are listed as part of section [sec:functional-optimization].</p>
<dl>
<dt>Content questions.</dt>
<dd>
<p>In this category of question I have mostly corresponded with the founder
and <span class="caps">CEO</span> of <span class="caps">AMF</span> on the topic of the foundation’s strategies for prioritization. I
have already covered the insights from these emails in section [sec:strategy-amf].</p>
</dd>
</dl>
<h2 id="ch:architecture-and-implementation"><a class="toclink" href="#ch:architecture-and-implementation">Architecture and Implementation</a></h2>
<blockquote>
<p>“We’ll continue tomorrow – if I live.” —Paul Erdős</p>
</blockquote>
<h3 id="sec:architecture"><a class="toclink" href="#sec:architecture">Architecture</a></h3>
<p><img alt="The architecture of the LLIN Visualizer. The arrows indicate what code
uses which resources. Gray components are used as black boxes; blue
components are concerned with data access; and green components are
concerned with data manipulation.<span
data-label="fig:architecture"></span>" src="/images/llin-visualizer/net-analytics-architecture.png"></p>
<p>the <span class="caps">LLIN</span> Visualizer exploits the model-view-controller design pattern to
maximally separate logic, so that each layer can be discussed in turn.</p>
<h3 id="subsec:models"><a class="toclink" href="#subsec:models">Models</a></h3>
<p><img alt="The models of the LLIN Visualizer. Some columns are omitted.<span
data-label="fig:models"></span>" src="/images/llin-visualizer/net-analytics-models.analytics.select.png"></p>
<dl>
<dt>Authentication.</dt>
<dd>
<p>The authentication system is provided by the Django framework. It allows
for fine-grained access management and for easily administering the
permissions of users or groups of users. Finally, management commands
provide functionality that is only run explicitly by the administrator
or by a cron daemon.</p>
</dd>
<dt>Survey responses.</dt>
<dd>
<p>The analytics app is the core of the <span class="caps">LLIN</span> Visualizer. The SurveyResponse
relation retains all the data of all surveys that have been imported
into the system. These data never change but reads need to be fast to
provide a good experience for the user, so I chose not to fully enforce
the third normal form on this table with regard to the columns that
describe the location (since one location, however defined, functionally
determines village name, health area name, district name, etc.). Doing
so would have been fraught with error anyway, as we will see in section [subsec:survey-data].</p>
</dd>
</dl>
<p>The location data was “four dimensional” in the sense that it comprised
latitude, longitude, altitude, and accuracy. At first I tried to use all
of these coordinates for the geometric component of the table, but
storing the accuracy in that fashion seemed to be unusual and for
2.5-dimensional data, including the altitude, I would have had to
continually override defaults throughout the code, so I found it simpler
to restrict the geometric component to two dimensions and store the rest
as floating point number columns.</p>
<dl>
<dt>Countries.</dt>
<dd>
<p>The only location variable that allowed for normalization according to
the third normal form and proved necessary for other performance-related
reasons<sup id="fnref:22"><a class="footnote-ref" href="#fn:22">22</a></sup> is the country. It is separated out to its own relation.</p>
</dd>
<dt>Clusters.</dt>
<dd>
<p>Since anonymous users must only see aggregated information for reasons
of data privacy, the Cluster relation provides such an aggregation on a
flexible level. Depending on the data it can sometimes be useful to
provide aggregations on the level of the village, the district, or
country-specific regions, such as the health areas of the Democratic
Republic of the Congo.</p>
</dd>
</dl>
<p>Using the Cluster relation, nested aggregations on several such levels
can be prepared, each with geographic information describing its
location and expanse.</p>
<dl>
<dt>Households.</dt>
<dd>
<p>The household relation only serves to link survey responses the belong
to the same household at different points in time. It is less a result
of the data management than the code layout, since it can neatly contain
any logic that will be necessary to match these entries.</p>
</dd>
</dl>
<p><img alt="image" src="/images/llin-visualizer/net-analytics-all-distributions.png"></p>
<dl>
<dt>Distributions.</dt>
<dd>
<p>Every survey response belongs to a distribution, but there are many
historical distributions that do not have any survey responses
associated with them. Just like clusters, distributions also carry
geographic information about their location and expanse as well as
various additional meta data.</p>
</dd>
<dt>Distribution campaigns.</dt>
<dd>
<p>When <span class="caps">AMF</span> conducts distributions, they usually make one contract with a local
partner stipulating several parallel or consecutive distributions in the
same country or region. By “distribution campaign” I refer to the
overall contract that comprises one or several distributions.
Distribution campaigns have corresponding pages on the website of <span class="caps">AMF</span>.</p>
</dd>
<dt>Statistics.</dt>
<dd>
<p>Finally the statistics relation is a flexible aggregation of data that
can be visualized for the user. The specifics of this module will be
described in section [subsec:management-commands].</p>
</dd>
</dl>
<p>All of the data in all relation can be imported or generated by
management commands but also individually administered through a
web-based backend.</p>
<h3 id="subsec:controllers"><a class="toclink" href="#subsec:controllers">Controllers</a></h3>
<p>The controllers (called “views” in Django) are the most minimal layer of
the software since most computations are so time-consuming and
unchanging that they are prepared once by management jobs. The results
are written to the database and merely served by the controllers.</p>
<p>Most controllers implement <span class="caps">JSON</span> <span class="caps">API</span> endpoints that are read by the JavaScript frontend.</p>
<h3 id="subsec:views"><a class="toclink" href="#subsec:views">Views</a></h3>
<p>The views (called “templates” in Django) are centrally based upon the
Bootstrap framework, CoffeeScript that compiles to JavaScript, the
Leaflet map software, and the D3 data visualization library.</p>
<p>A light scaffold provides the basis for the application code and the
tests. Upon this scaffold hang the NetLayer classes, each representing
one layer of the data aggregation. The first, the distribution layer,
represents all distributions, with or without survey data. Those
distributions that have survey data associated with them allow for a
drill-down to the flexible cluster layer – for example health zones in
the case of the <span class="caps">DRC</span>. At the same time the corresponding data on insecticide
resistance tests for the same country is displayed; it is represented by
a layer of its own right. Finally, authenticated users can drill down to
the household level and view individual survey responses.</p>
<p>The data on insecticide resistance tests and malaria incidence is
provided by the Insecticide Resistance Mapper.</p>
<h3 id="subsec:management-commands"><a class="toclink" href="#subsec:management-commands">Management Commands</a></h3>
<p>Most of the logic of the <span class="caps">LLIN</span> Visualizer is in the views and in the
management commands. These are jobs that are executed very occasionally,
for example, when new distribution data is imported. Most of the data
whose processing is discussed below have been introduced in section [subsec:other-data].</p>
<dl>
<dt>Distribution import.</dt>
<dd>
<p>An <span class="caps">API</span> on the website of <span class="caps">AMF</span> provides distribution data as <span class="caps">XML</span> dialect in a
semistructured fashion. The job parses the <span class="caps">XML</span>, extracts the structured
data, and then uses a heuristic approach to extract more data from the
embedded description.</p>
</dd>
</dl>
<p>Any dates are either given as one pair of month and year, as a range of
months and a year, or as a range of two pairs of month and year. A
significant part of the code is concerned with date parsing.</p>
<p>Finally the data are written to the database atomically for performance
and so no clean-up of an incomplete, aborted import is necessary.</p>
<p>This file, however, lacks some data, which are then scraped from the
<span class="caps">HTML</span> of the <span class="caps">AMF</span> website itself, which again necessitates some heuristic parsing.</p>
<dl>
<dt><span class="caps">IR</span> Mapper import.</dt>
<dd>
<p><span class="caps">IR</span> Mapper<sup id="fnref3:IRMapper"><a class="footnote-ref" href="#fn:IRMapper">87</a></sup> provides its data as <span class="caps">JSON</span>, but unfortunately it is
still not perfectly structured. Many numeric values are saved as
strings, sometimes without apparent reason, but sometimes also because
some values are in different units than others, significantly sometimes
in absolute units and sometimes in percent. The import job tries hard to
resolve these inconsistencies and save the data in the same unit.</p>
</dd>
</dl>
<p>It also maps idiosyncratic names for countries to their official spellings.</p>
<p>The data on Plasmodium falciparum and Plasmodium vivax incidence rates
and the entomological inoculation rate, the number of infective bites
per person per time, are also sourced from an <span class="caps">API</span> provided by <span class="caps">IR</span> Mapper in
the form of raster layers. The data themselves are the product of the
.<sup id="fnref:MalariaAtlasProject"><a class="footnote-ref" href="#fn:MalariaAtlasProject">100</a></sup></p>
<dl>
<dt>Net gap import.</dt>
<dd>
<p>The net gap data (see subsection [subsec:other-data]) have come in
tables that can be converted to <span class="caps">CSV</span> files. A dedicated job imports data
on nets needed, financed, and net gaps severally. The user needs to
specify the type of data (out of those three types) and the year of the
estimate, so that the <span class="caps">LLIN</span> Visualizer can choose the newer estimate if
there are several for a given year, country, and type.</p>
</dd>
<dt>Indicator import.</dt>
<dd>
<p>The indicator data described in subsection [subsec:other-data] needs
to be imported from various sources in several different formats. The
indicator import handles this task. The user supplies the data in <span class="caps">CSV</span>
files and specifies the type of data they are importing, and the job
augments the variables known for each country-year tuple with the new data.</p>
</dd>
<dt>Survey import.</dt>
<dd>
<p>The survey import job is probably the module that was most
time-consuming to implement.</p>
</dd>
</dl>
<p>First, it needs to be flexible. The data <span class="caps">AMF</span> has provided up to this point
have come in two different formats, but there is no limit to the number
of different raw data formats the <span class="caps">LLIN</span> Visualizer needs to be able to
handle. Hence, a <span class="caps">CSV</span> parser class provides a generic interface to all the
cleaning methods I have written while individual <span class="caps">CSV</span> profiles inherit these
and determine to which fields they need to be applied. At the same time
they provide the flexibility of mapping fields of the input data to
arbitrary database fields, not only one to one, but also by merging
fields or splitting them up.</p>
<p>Second, the cleaning of the data requires flexibility too. Each
resulting field needs to have custom cleaning routines associated with
it but there also need to be higher-level cleaning routines that can
decide whether the data from several fields in combination make sense or
whether the whole row is invalid. Due to some intricacies of the order
in which type conversions are applied, I could not draw on the form
validation tools that Django provides. The key difference was that these
escalate field-level errors to the user to correct or re-enter the data.
In our case, this is either impossible or the job of the cleaning routines.</p>
<p>Many of the cleaning routines preform type conversions but there are a
few more sophisticated operations involved as well. The date parsing
draws on contextual data from the <span class="caps">AMF</span> website to sanity-check the dates and
discard them if them seem erroneous. Many cleaning routines apply
translations that convert fields with values of “oui” and “non” to
Boolean types, and normalize various strings to key values that are
standardized within the <span class="caps">LLIN</span> Visualizer. Finally, names of locations are
normalized using a sequence of heuristics that are optimized to merge
many different spellings of the same name while introducing few errors.</p>
<p>The module allows the administrator fine-grained control over its
behavior by specifying the parser and distribution to be used,
specifying some properties of the <span class="caps">CSV</span> file, and specifying the mode in which
the database writer should operate. There is also a mode that allows for
the testing of new parser profiles.</p>
<p>The writer is resilient in that it can recover after a failed insertion
and continue processing. This is only possible when it commits after
every row, so while the atomic writer has a significant performance
advantage, it is not applicable when the data contain errors. It is not
used by default, but when the user knows that the data contain no
errors, the option can be activated.</p>
<dl>
<dt>Cluster generation.</dt>
<dd>
<p>The cluster generation job at the same time generates the statistics
associated with the clusters and runs the same calculations for
distributions, which are structurally identical to clusters. In detail:</p>
</dd>
</dl>
<ol>
<li>
<p>The first job groups all responses to the surveys by their
respective distributions and then uses PostGIS functions to
calculate the centroids and convex hulls of these points. These data
are then used to augment the distribution relation.</p>
<p>Below a simplified <code>select</code> query in relational algebra using a
syntax based on<sup id="fnref:RelaXHelp"><a class="footnote-ref" href="#fn:RelaXHelp">93</a></sup> by way of illustration of the actual
<code>update</code> query used in the management command. Note that the
relational algebra syntax resembles pivoting more than the
<code>group by</code> syntax I chose in the code, so that the structure is
slightly different from the <span class="caps">SQL</span> version.</p>
<p>
<div class="math">$$\begin{array}{rl}
\pi & _{ \mathtt{distribution\_id},~
\mathtt{st\_centroid}(\mathtt{collection}) ~→~ \mathtt{centroid},~
\mathtt{st\_convexhull}(\mathtt{collection}) ~→~ \mathtt{hull}} \\
\gamma & _{ \mathtt{distribution\_id};
\mathtt{st\_collect}(\mathtt{geom}) ~→~ \mathtt{collection}} \\
\multicolumn{2}{l}{\mathtt{analytics\_surveyresponse}}
\end{array}$$</div>
</p>
<p>With the currently available data, this query takes about 20 s to
execute on my computer, aided by the index on the primary
key column. These measurements have no absolute relevance but
provide comparative values that allow me to optimize them, aided by
PostgreSQL <code>explain analyze</code> command, and compare them to
the others.</p>
</li>
<li>
<p>The second job is parameterized with <code>kind</code> and <code>fields</code>, where
<code>kind</code> is a name for the type of cluster, such as “village” or
“health area,” and <code>fields</code> are the grouping fields that describe
this type. Here it is important to note that for a grouping by
village it is better to group by village name, health area name,
health zone name, and even larger categories rather than just by
village name, since there are many villages with the same name.</p>
<p>The query performs the same PostGIS operations as the previous one,
but then also concatenates the values of the grouping fields to one
comma-joined string, such as “Shakunga, Kabungu,” as a name for
the cluster. These names are by necessity unique.</p>
<p>A simplified <code>select</code> query in relational algebra using the <code>kind</code>
“village” and the <code>fields</code> “village_name” “health_area_name” by
way of illustration of the actual parameterized <code>insert</code> query used
in the management command:</p>
<p>
<div class="math">$$\begin{array}{rl}
\pi & _{ \text{\texttt{'village'}},} \\
& _{ \mathtt{concat}(\text{\texttt{', '}},~
\mathtt{village\_name},~
\mathtt{health\_area\_name}),} \\
& _{ \mathtt{distribution\_id},} \\
& _{ \mathtt{st\_centroid}(\mathtt{collection}) ~→~ \mathtt{centroid},} \\
& _{ \mathtt{st\_convexhull}(\mathtt{collection}) ~→~ \mathtt{hull}} \\
\gamma & _{ \mathtt{village\_name},~
\mathtt{health\_area\_name}),~
\mathtt{distribution\_id};} \\
& _{ \mathtt{st\_collect}(\mathtt{geom}) ~→~ \mathtt{collection}} \\
\multicolumn{2}{l}{\mathtt{analytics\_surveyresponse}}
\end{array}$$</div>
</p>
<p>This query was significantly sped up – from 17 s to 10 s execution
time on my computer – by indices over the typical grouping columns
“village_name” and “health_area_name” in addition to the index on
the primary key column due to an external merge sort that PostgreSQL
perform over the tuple of these three columns. The choice which
columns to use for cluster generation, however, is one that is up to
the administrator, so it makes sense to add indices like these
according to the actual usage patterns rather than in advance
during development.</p>
<p>Finally, the job updates all survey responses with relations to
their respective clusters, using the generated names
for identification.</p>
</li>
<li>
<p>The third job generates the cluster and distribution statistics. The
type of the grouping is again parameterized through the foreign key
field and the referenced relation.</p>
<p>The job creates descriptive statistics such as averages, maximums,
standard deviations, counts, and sums of numeric and geographic data
in scalar and vector formats. It heeds the definitions of the data
structures from the survey importer and interprets them as
consisting of the names of keys and the names of values, to which it
assigns aggregate results. An example will clarify the operation.</p>
<p>One simple case is the cluster centroid: The centroid is assigned
various attributes to enrich it, namely the average distance of the
points it describes, their maximal distance, the standard deviation
of their distance, and their number. PostGIS functions are used to
determine these values for the centroids of every cluster. In total
there are currently 60 such tuples, which are computed for each of 8
distributions and over 10,000 clusters.</p>
<p>This abstract approach to the generation of cluster statistics has
become necessary because requirements for statistical analyses can
change (expand) quickly and should only require minimal changes to
configuration parameters. At the same time the data are scalar in
some cases and vectorial in others, and the parameter combinations
that require calculation are sometimes flat and sometimes nested (so
that each path tuple requires a separate calculation). This more
abstract system is sufficiently powerful to limn all these
use cases.</p>
<p>This query is heavily parameterized and uses many PostgreSQL
features that have no equivalent in any relational algebra syntax
that I am aware of, such as array operations and the <code>unnest</code>
operation on arrays.</p>
<p>This giant query takes about 14 s to execute on my computer, aided
by the existing indices, and I could not find any leverage points to
further optimize it.</p>
</li>
</ol>
<h3 id="sec:technology"><a class="toclink" href="#sec:technology">Technology</a></h3>
<h3 id="subsec:core-technologies"><a class="toclink" href="#subsec:core-technologies">Core Technologies</a></h3>
<p><img alt="Git, Python, and Django logos.<span
data-label="fig:software-logos"></span>" src="/images/llin-visualizer/software-logos.png"></p>
<dl>
<dt>Git</dt>
<dd>
<p><span class="caps">CVS</span><sup id="fnref:CVS"><a class="footnote-ref" href="#fn:CVS">63</a></sup> and <span class="caps">SVN</span>,<sup id="fnref:SVN"><a class="footnote-ref" href="#fn:SVN">33</a></sup> the other version control systems that I
have worked with, hold no sway over Git<sup id="fnref:Git"><a class="footnote-ref" href="#fn:Git">138</a></sup> in terms of the
useability that I have experienced. Only Mercurial<sup id="fnref:Mercurial"><a class="footnote-ref" href="#fn:Mercurial">104</a></sup> was
similar to Git, but I have not gathered much experience with it.</p>
</dd>
<dt>Python</dt>
<dd>
<p>The decision to implement the project in the programming language
Python<sup id="fnref:Python"><a class="footnote-ref" href="#fn:Python">125</a></sup> was motivated by my more than ten years of
experience using the language, which in turn is thanks to the
elegant syntax of the language, its versatility, and consequently
the rapid pace of development that it lends itself to.</p>
</dd>
<dt>Django</dt>
<dd>
<p>the <span class="caps">LLIN</span> Visualizer, by virtue of its functional requirements,
needed to have a frontend exposed to the web. There are a number of
web frameworks for Python, some that aim to be minimal and thus
light and highly flexible and others that aim to be feature rich.</p>
<p>Typically I would use a minimal framework to reduce the complexity
of the project whenever possible, but experience has taught me that
unless I can see with certainty that the project will remain small,
it is usually the case that its complexity will increase drastically
over time. In expectation of the more complex project that will
emerge half a year down the line, my strategy is now usually the
opposite, namely, to go with a feature-rich framework unless the
project clearly does not call for it, or some of the features clash
with the developers plans for the program architecture.
CherryPy,<sup id="fnref:CherryPy"><a class="footnote-ref" href="#fn:CherryPy">46</a></sup> Flask,<sup id="fnref:Flask"><a class="footnote-ref" href="#fn:Flask">61</a></sup> and Pyramid<sup id="fnref:Pyramid"><a class="footnote-ref" href="#fn:Pyramid">124</a></sup> are
examples of such lighter frameworks; Django<sup id="fnref:Django"><a class="footnote-ref" href="#fn:Django">143</a></sup> is an example
of a feature-rich one.</p>
<p>Thanks to Django, the <span class="caps">LLIN</span> Visualizer has database migration
support, an object-relational mapper, a user authentication system,
and much more built in. One place where the assumptions of the
framework clashed with my preferences for the <span class="caps">LLIN</span> Visualizer,
however, was the templating engine. Here I greatly prefer
Jinja2<sup id="fnref2:Jinja2"><a class="footnote-ref" href="#fn:Jinja2">129</a></sup> over Django’s built-in templating engine. Luckily,
it has become easy to select alternative templating engines
in Django.</p>
</dd>
<dt>PostgreSQL</dt>
<dd>
<p>The main alternative to the database management system
PostgreSQL<sup id="fnref:PostgreSQL"><a class="footnote-ref" href="#fn:PostgreSQL">121</a></sup> would have been MySQL,<sup id="fnref:MySQL"><a class="footnote-ref" href="#fn:MySQL">115</a></sup> but I have
had more positive experiences with PostgreSQL. Both systems are well
supported by Django.</p>
</dd>
<dt>PostGIS</dt>
<dd>
<p>PostGIS<sup id="fnref:PostGIS"><a class="footnote-ref" href="#fn:PostGIS">120</a></sup> is an extension to PostgreSQL that adds support
for geographic objects and forms the basis for many calculations in
the <span class="caps">LLIN</span> Visualizer. It is also well supported by Django.</p>
</dd>
</dl>
<h3 id="subsec:other-important-technologies"><a class="toclink" href="#subsec:other-important-technologies">Other Important Technologies</a></h3>
<dl>
<dt>Buildout</dt>
<dd>
<p>Buildout<sup id="fnref:Buildout"><a class="footnote-ref" href="#fn:Buildout">156</a></sup> makes sure that when one can build an
application with two commands in 2016, one will still be able to
build it with the same commands in 2018. In practice it sometimes
falls short of this goal when Python libraries have dependencies on
libraries or C bindings provided by the operating system, but it
goes a long way toward making builds repeatable.</p>
<p>Another advantage is that one can easily integrate packages from
outside the Python realm, such as NodeJS or Ruby packages.</p>
</dd>
<dt>Leaflet</dt>
<dd>
<p>When I did research on what map library to employ in the <span class="caps">LLIN</span>
Visualizer, Leaflet<sup id="fnref:Leaflet"><a class="footnote-ref" href="#fn:Leaflet">28</a></sup> and OpenLayers<sup id="fnref:OpenLayers"><a class="footnote-ref" href="#fn:OpenLayers">113</a></sup> were the
ones that were recommended most often. OpenLayers was praised as the
well-established standard solution, Leaflet as the new and
simpler runner-up. Both seemed to be well suited for the purposes of
the <span class="caps">LLIN</span> Visualizer. What swayed me toward Leaflet is that I found
that several companies had switched from OpenLayers to Leaflet when
the latter software had sufficiently matured. Given the small
differences between the libraries in functional terms and the
feasibility of switching within the first days of development, I did
not want to invest more than a few hours into the comparison and
decided on Leaflet.</p>
</dd>
<dt>OpenStreetMap</dt>
<dd>
<p>OpenStreetMap<sup id="fnref:OpenStreetMap"><a class="footnote-ref" href="#fn:OpenStreetMap">114</a></sup> is a collaborative effort to create a
free map of the world. This map – which, unfortunately, is fairly
uninformative for many of the most interesting distribution regions
– supplies the background layers upon which the <span class="caps">LLIN</span> Visualizer
displays the net data.</p>
</dd>
<dt><span class="caps">QGIS</span></dt>
<dd>
<p>This <span class="caps">GIS</span> application allows for graphical insight into the data stored
in PostgreSQL.<sup id="fnref:QGIS"><a class="footnote-ref" href="#fn:QGIS">126</a></sup> I have been using it to prototype new
analyses, which I would later make part of the software proper.</p>
</dd>
<dt>CoffeeScript</dt>
<dd>
<p>There are a number of humorous presentations on the Internet that
introduce various absurdities of the JavaScript language
design.<sup id="fnref:Wat"><a class="footnote-ref" href="#fn:Wat">37</a></sup> When writing JavaScript code, one option is to learn
these and learn how to work around them, another is to use
CoffeeScript,<sup id="fnref:CoffeeScript"><a class="footnote-ref" href="#fn:CoffeeScript">47</a></sup> which compiles to JavaScript. I also
enjoy its syntax.</p>
</dd>
<dt>Jinja2</dt>
<dd>
<p>The default templating language of Django takes the idea of
separating presentation from logic to an extreme by going to some
lengths to force the developer to keep any logic out of
the templates. But the purpose of the separation used to be to keep
the layers of the <span class="caps">MVC</span> model orthogonal, and when the developer is
forced to implement clearly view-related logic in the controller
because the templating language is not powerful enough to handle it,
the orthogonality suffers in the opposite direction. Admittedly,
there is the option of writing custom template tags to capture
custom template logic, but then the language is also limited in ways
that do not seem to be purposeful.</p>
<p>Jinja2,<sup id="fnref:Jinja2"><a class="footnote-ref" href="#fn:Jinja2">129</a></sup> on the other hand, is a very powerful templating
language that follows more in the spirit of Python in that it trusts
the developer to make sane decisions in regard to the separation of
view logic from controller logic.</p>
</dd>
<dt>Sphinx</dt>
<dd>
<p>The documentation of the <span class="caps">LLIN</span> Visualizer is written in such a way as
to enable other developer to understand and extend the software. As
such it takes the form of code comments, which Sphinx<sup id="fnref:Sphinx"><a class="footnote-ref" href="#fn:Sphinx">42</a></sup> can
automatically compile into a structure of web pages that forms a
complete manual to the software. Reading natural language is a
cognitively different process from reading code, so that excessive
code comments can seem intrusive. In Python they are contained
within one string at the start of a module or method, and with
Sphinx one can even read them completely separately from the code.</p>
</dd>
<dt>GeoDjango</dt>
<dd>
<p>Django ships with a component that extends the abstraction layers of
the framework to spacial data. GeoDjango<sup id="fnref:GeoDjango"><a class="footnote-ref" href="#fn:GeoDjango">53</a></sup> provides
extensions to the object-relational mapper, the validation
framework, and the managements commands, and also contains a host of
utilities for the manipulation of spacial data outside the database.</p>
</dd>
<dt>Handlebars</dt>
<dd>
<p>JavaScript prior to ECMAScript 6<sup id="fnref:ECMAScript"><a class="footnote-ref" href="#fn:ECMAScript">59</a></sup> did not have its own
string interpolation syntax, and while ECMAScript 6 and CoffeeScript
do provide this feature, especially longer strings, such as long
sections of the <span class="caps">HTML</span> of a website, can be formatted much more
comfortably with a templating language as Handlebars<sup id="fnref:Handlebars"><a class="footnote-ref" href="#fn:Handlebars">92</a></sup>
provides it.</p>
</dd>
<dt>D3</dt>
<dd>
<p>The website of D3<sup id="fnref:D3"><a class="footnote-ref" href="#fn:D3">40</a></sup> gives an overview of the enormous versatility
of the library. It is a JavaScript library that uses SVGs on top of
the <span class="caps">HTML</span> and <span class="caps">CSS</span> stack to visualize data of just about any
structure imaginable. the <span class="caps">LLIN</span> Visualizer is not using D3 directly
but through the graphing library C3.<sup id="fnref:C3"><a class="footnote-ref" href="#fn:C3">39</a></sup> All statistical
visualization are rendered using this library.</p>
</dd>
<dt>Jasmine</dt>
<dd>
<p>While Django ships with its own testing framework for the Python
code, much of the <span class="caps">LLIN</span> Visualizer is implemented in CoffeeScript,
which needs to be unit-tested as well, and for that the <span class="caps">LLIN</span>
Visualizer employs Jasmine,<sup id="fnref:Jasmine"><a class="footnote-ref" href="#fn:Jasmine">119</a></sup> a testing framework that runs
in the browser alternative to the frontend the visitor sees but
drawing on the same code base.</p>
</dd>
<dt>Fabric</dt>
<dd>
<p>Since the <span class="caps">LLIN</span> Visualizer is not a distributed system, the
deployment does not pose a challenge, but even so it is greatly
simplified with Fabric<sup id="fnref:Fabric"><a class="footnote-ref" href="#fn:Fabric">62</a></sup> scripts that execute common tasks
programmatically on the remote server.</p>
</dd>
</dl>
<h2 id="ch:evaluation-and-learnings"><a class="toclink" href="#ch:evaluation-and-learnings">Evaluation and Learnings</a></h2>
<h3 id="sec:functional-optimization"><a class="toclink" href="#sec:functional-optimization">Functional Optimization</a></h3>
<p>In December 2015, I initiated the public beta phase of the <span class="caps">LLIN</span>
Visualizer. Early users made the following suggestions:</p>
<ol>
<li>
<p>Tooltips with explanations of specialized terms.</p>
</li>
<li>
<p>Saving of at least the distribution layer of the map state to the <span class="caps">URL</span> so
that one can link to specific distributions.</p>
</li>
<li>
<p>Tests for insecticide resistance against chemicals that are not used for
LLINs can be hidden by default.</p>
</li>
<li>
<p>Data of distributions of other funders if it can be obtained.</p>
</li>
<li>
<p>The visualized dimension of the data (size of the map icon) can be
explained by highlighting the respective item in the info box.</p>
</li>
<li>
<p>The info boxes of distributions can be improved by including photos.</p>
</li>
</ol>
<h3 id="evaluation"><a class="toclink" href="#evaluation">Evaluation</a></h3>
<p>The map that <span class="caps">AMF</span> currently still uses to indicate the location of
distributions is fairly limited in features to the point where a
comparison of features would seem unfair. It is clearly designed as an
unobtrusive widget that is ancillary to a distribution page, a widget
that adds some value to it but is by no means central to it.</p>
<p>Accordingly it is small in size and limited in interactivity. It
displays the endemicity of malaria in various regions around the world
as a binary, consistent with <span class="caps">AMF</span>’s prioritization practice (see section
[sec:strategy-amf]), either applying a relatively low threshold for
endemicity or using data that has not been updated since the launch of
the website.<sup id="fnref:23"><a class="footnote-ref" href="#fn:23">23</a></sup></p>
<p>Distributions are indicated by yellow dots. The user can click on
continents to zoom in, which opens <span class="caps">AMF</span>’s distribution list filtered by
that continent.</p>
<p>the <span class="caps">LLIN</span> Visualizer improves upon this in many ways:</p>
<p>it adds all the data that comes with using OpenStreetMap,</p>
<p>allows a drill-down into the distributions down to village level or even
the individual household if the user is authorized to see it,</p>
<p>allows users to filter distributions and villages (or <em>clusters</em> in
general) according to various criteria,</p>
<p>can view distributions and villages in different modes, for example, as
centroids or as convex hulls,</p>
<p>visualizes statistical data on the distributions and villages,</p>
<p>shows a wealth of other dimensions of the data that are not visualized
in tabular info boxes,</p>
<p>shows aggregate statistics on malaria knowledge, malaria prevention
measures, their usage, and the sources of nets for villages and distributions,</p>
<p>visualizes statistics (also provided by <span class="caps">AMF</span>) on the nets needed and the
net gap of many relevant countries and allows users to filter them,</p>
<p>displays incidence in addition to endemicity (and both in higher detail)
for plasmodium falciparum and plasmodium vivax (using data provided by
<span class="caps">IR</span> Mapper),</p>
<p>draws on <span class="caps">IR</span> Mapper for data on insecticide resistance tests from the
past decades, which it visualizes and allows the user to filter, and</p>
<p>provides guidance to the user with visual feedback, animations, and many
info boxes that explain functions and specialized terms.</p>
<p>During my research I have found several other maps that are more or less
centrally concerned with malaria. They have all been mentioned
throughout this text and are linked from the <span class="caps">LLIN</span> Visualizer, since my
software draws on some of this existing infrastructure or data. the <span class="caps">LLIN</span>
Visualizer being concerned with the prioritization of net distributions,
it is again hard to compare these solutions.</p>
<p>The main achievement of the Malaria Atlas Project<sup id="fnref2:MalariaAtlasProject"><a class="footnote-ref" href="#fn:MalariaAtlasProject">100</a></sup>
lies in the estimation of the spacial distribution of Plasmodium
falciparum incidence, net coverage, indoor residual spraying coverage,
and artemisinin-based combination therapy coverage using sparse data
augmented with model-based extrapolations. The interactive map the
project uses to showcase these data allows for filtering them by year.
the <span class="caps">LLIN</span> Visualizer draws heavily on these data as described in section [subsec:other-data].</p>
<p>The same is true of the Insecticide Resistance Mapper.<sup id="fnref2:IRMapper"><a class="footnote-ref" href="#fn:IRMapper">87</a></sup> This
app draws on some data from the Malaria Atlas Project to visualize
insecticide resitance tests of the last sixty years, including many for
insecticides that have not been used for a long time. To exclude such
chemicals, it allows for fine-grained filters over the test results. As
detailed in section [subsec:other-data], the <span class="caps">LLIN</span> Visualizer draws its
insights into insecticide resistance from this application.</p>
<p><img alt="image" src="/images/llin-visualizer/map-map-congo.png"></p>
<h3 id="sec:nonfunctional-optimization"><a class="toclink" href="#sec:nonfunctional-optimization">Nonfunctional Optimization</a></h3>
<dl>
<dt>The cluster layer.</dt>
<dd>
<p>Intuitiveness was an important factor in the optimization process. The
errors in the data that are discussed in section [subsec:survey-data]
lead to tradeoffs where the most intuitive approaches lead to more or
less disruptively damaged data.</p>
</dd>
</dl>
<p>Another factor is the user experience of working with the map. Several
thousand of map features displayed at the same time make the map feel
sluggish even in modern browsers on most computers. Once canvas-based
solutions become mature enough to be usable as orthogonal abstraction
layers on top of the existing library code, this problem will fade into
the background, but until then, some degree of windowing, lazy loading,
and optimized structuring of the data will be necessary.</p>
<p>If windowing can be thought of as a horizontal limiting of the data that
is displayed, then the optimization of the data structure is a vertical
measure. This additional measure has become necessary because windowing
alone was only in some cases effective in constraining the number of
active map features while allowing full usability. For example, when
users are interested in a distribution in the <span class="caps">DRC</span>, then only insecticide
resistance tests in that country need to be displayed since the
distribution they are interested in does not stray over country borders
either. The same principle, however, does nothing to constrain the
number of data points of the distribution itself that are displayed, so
new categories for windowing are needed. The optimized data structure
provides those.</p>
<p><img alt="A tree symbolizing the original structure of the data and one
symbolizing the optimized structure. A user has drilled down to the
lowest layer, so that the highlighted vertices are active.<span
data-label="fig:clustering"></span>" src="/images/llin-visualizer/clustering.png"></p>
<p>Consider the two trees in figure [fig:clustering]. The first consists
of three layers. A user that drills down to the last layer has to click
twice to select windows of the lower layers. In total 19 map features
are displayed. In the second case, the user has to click thrice, and
only 12 map features are displayed in total.</p>
<p>This technique can backfire. Let us call the new layer the cluster
layer, as it is called in the <span class="caps">LLIN</span> Visualizer. When there are too few
clusters, they have hardly any effect; too many clusters also add
overhead on the cluster layer that reduces the effectiveness of the
optimization, but if the clusters are chosen to present the perfect
balance between these extremes, they are likely to be unintuitive for
the user. Furthermore, the aforementioned ambiguities and errors also
meant that some of the clusters were, given the world knowledge that
made them intuitive in general, highly unlikely. Hence I experimented
with different types of intuitive clusters, measured the number of
clusters, the average cluster size, and the proportion of them that
seemed erroneous, according to heuristics such as the cluster elements’
standard deviation from the geographic cluster centroid.</p>
<p>The results are only interesting for the data that I currently possess,
the predistribution registration survey of the distribution Kasaï
Occidental 2014 in the <span class="caps">DRC</span>, and the <span class="caps">LLIN</span> Visualizer allows for choosing
different types of clusters for different distributions. The winner was
a clustering by tuples of normalized village name and health area name.
Adding the sanitized health zone name into the tuple had little effect
except on the runtime of the query. The resulting solution led to a
total of 10,090 clusters with 1,261 clusters per distribution on average
(<span class="math">\(\sigma = 516\)</span>). Many clusters are errors due to misspellings, but with
further sanitization steps applied, geared toward precision rather than
recall, the reduced set of 676 clusters contains an average of 227
households (<span class="math">\(\sigma = 210\)</span>).</p>
<dl>
<dt>Precomputation.</dt>
<dd>
<p>I have already touched on the query optimizations I have performed in
sections [subsec:models] and section [subsec:management-commands].
In summary: There are a number of calculations that would normally take
several seconds to complete, so that they are unsuitable for execution
at request time. Rather, the results are precomputed in the <span class="caps">LLIN</span>
Visualizer, leading to runs of management commands that can take several
minutes but queries on the precomputed data that only take split
seconds. This is one important optimization. The optimizations I was
able to make to the long-running queries are described in section
[subsec:management-commands]; they are mostly related to applying
indices to relevant columns.</p>
</dd>
<dt><span class="caps">ORM</span>-level optimizations.</dt>
<dd>
<p>The quick queries that are executed at request time, however, are
executed through the object-relational mapper interface that Django
provides, so it pays to learn the optimization tips that the Django
developers provide.</p>
</dd>
</dl>
<p>There are two situations were material gains can be realized. One
relates to an object (a row in a table) that has a foreign key
relationship to another object and is associated with the
<code>select_related</code> method. By default, only the first object is fetched
from the database, so that an additional query has to be executed to
obtain the related object if and once it is needed. Two queries usually
incur greatly more overhead than one more complex query, so
<code>select_related</code> allows the developer, who knows what related objects
are likely to be needed later on, to specify which of them should be
fetched right away.</p>
<p>The other optimization relates to many-to-many relationships and reverse
foreign key relationships and is associated with the <code>prefetch_related</code>
method. From the developers perspective it behaves similar to
<code>select_related</code>, but whereas <code>select_related</code> has the database perform
a join operation, <code>prefetch_related</code> performs separate lookups for all
the related objects and then performs the join in memory in Python.</p>
<dl>
<dt>Response caching.</dt>
<dd>
<p>Another important improvement relates to the response caching of the
<span class="caps">LLIN</span> Visualizer. Even though the precomputation and the query
optimization above already reduced the response time significantly, I
was able to speed it up further by caching the responses. With the <span class="caps">LLIN</span>
Visualizer I am in an especially auspicious situation for caching as the
data changes rarely. Hence I opted for no automatic expiry and explicit
cache invalidation in addition to the unavoidable eviction due to memory
restrictions of the server. (With the amount of data currently imported
into the <span class="caps">LLIN</span> Visualizer, these are not reached.)</p>
</dd>
</dl>
<h3 id="evaluation_1"><a class="toclink" href="#evaluation_1">Evaluation</a></h3>
<dl>
<dt>Clusters.</dt>
<dd>
<p>Since I find “sluggishness” of a map hard to quantify and measure, I
will use the number of map features as a proxy. Introducing clusters
requires sanitization filters on this level with the useful side effect
of further limiting the number of map features that are displayed by
default. The numbers ignore unrelated map features, for example from
insecticide resistance tests.</p>
</dd>
</dl>
<p>The initial view is what shows up before user interaction, and the
drill-down is the state after the user clicks one distribution for which
data is available. These measurements are averaged over all
distributions for which this hold, a sample of eight at the moment.</p>
<hr>
<div class="highlight"><pre><span></span><code> Without clusters With clusters Default filters
</code></pre></div>
<p>Initial view – Drill-down </p>
<hr>
<p>I am satisfied with these optimizations.</p>
<dl>
<dt>Precomputation.</dt>
<dd>
<p>Response times are more straightforward than “sluggishness,” so I am
measuring them directly. Here, the precomputation was an architectural
decision I made early on, so the “without precomputation” measurements
are those of queries I created purely for comparative purposes. The
sample is again one of eight distributions.</p>
</dd>
</dl>
<hr>
<div class="highlight"><pre><span></span><code> Without precomputation With precomputation
</code></pre></div>
<p>Distributions <br>
Clusters <br>
Distribution statistics <br>
Cluster statistics </p>
<hr>
<p>Most of these optimizations had a great impact, except for the
precomputation of the cluster statistics, since even the retrieval of
the precomputed numbers requires one join. But since the infrastructure
was already in place, the comparatively small optimization in this case
also came at a small marginal cost.</p>
<dl>
<dt><span class="caps">ORM</span>-level optimizations and caching.</dt>
<dd>
<p>This measurement averaged over ten identical requests to the
distributions endpoint and all cluster endpoints (one request each). The
“with caching” column measures cache hits.</p>
</dd>
</dl>
<hr>
<div class="highlight"><pre><span></span><code> Without optimizations With optimizations With caching
</code></pre></div>
<p>Distributions <br> Clusters </p>
<hr>
<p>The impact of the query optimizations was small in this case, almost
negligible in the case of the cluster queries, but the caching had a
tremendous impact, making it indispensable. To grant the fast cached
experience to as many users as possible, the <span class="caps">LLIN</span> Visualizer preloads
the clusters when the browser is idle, so that they are cached on the
server side and in the browser if they are not already there.</p>
<h3 id="sec:learnings"><a class="toclink" href="#sec:learnings">Learnings</a></h3>
<h3 id="subsec:learnings-dates"><a class="toclink" href="#subsec:learnings-dates">Dates</a></h3>
<p><span class="caps">AMF</span> had already noticed that many of the dates were incorrect. When
implementing the cleaning jobs, I observed the errors more closely to
form theories of their origin. The dates of the rows are broken in so
many ways that it seems likely that several problems coincided here to
produce them. Two possible reasons are that (1) phones used in the
distribution had some default date set at the various times when they
were first activated instead of the real one or (2) data was lost on
various data conversions.</p>
<p>Only looking at responses from the first days of distributions, about
47,000 (22%) are dated 2013, a year early. Almost 76,000 (36%) of this
subset were dated 1899. About 83,000 (40%) were in the correct year. In
total 56 different years were given. Less than 1% of the samples fell
into each of all other years.</p>
<p>The year 1899 seems like an unlikely default for a phone, so I think it
is more likely that here data was lost during a conversion. The
countless different and often ambiguous date formats make these
notoriously hard. The dates that are only off by a year may well be
factory default dates of the phones. Differences of a few months between
the erroneous dates can be explained by people first activating the
phones at different times.</p>
<p>The conversion problems are particularly likely because instead of
keeping and working with the raw data, the responses different teams
collected were aggregated and then only saved in a format specific to
the program “Microsoft Excel.” It has become possible to read the format
without this software, but conversions to and from such a format are
virtually bound to lead to data loss.</p>
<p><span class="caps">AMF</span> has taken care to alert the distribution partner to this problem so
that it can take greater care to set the correct date and time on the
phones and so that it will avoid lossy conversions of the data going forward.</p>
<h3 id="subsec:learnings-health-areas"><a class="toclink" href="#subsec:learnings-health-areas">Health Areas</a></h3>
<p>Experimenting with different ways of clustering the responses into
geographical units (the process described in section
[sec:nonfunctional-optimization]), I noticed something strange about
the level of the health area.</p>
<p><img alt="The village Nsoka, Muladila, has households that are over 170 km away
from its centroid, an error in the data.<span
data-label="fig:weird-clusters"></span>" src="/images/llin-visualizer/weird-clusters.png"></p>
<p>One of my options was clustering by the name of the health area. Looking
at the data in <span class="caps">QGIS</span>, I noticed that the health area names seemed to be
ambiguous, so I used the next larger unit, the health zone, to
disambiguate. Even that, however, was unsuccessful in some cases. As the
screenshot from the <span class="caps">LLIN</span> Visualizer shows, I eventually resorted to
clustering by village so that I could use the combination of all three
variables, health zone, health area, and village, for disambiguation.
(The current implementation uses a simplified version of this approach
that provides better performance.)</p>
<p>There are over 10,000 distinguishable villages in the data set, so it is
unsurprising that many villages share the same name. The health areas,
however, seem like centrally ordained administrative regions, and there
are only about 200 of them, so I was surprised that they were ambiguous
even within the next greater unit, the health zone.</p>
<p>I reported these findings to <span class="caps">AMF</span> who investigated the phenomenon and
eventually formed the hypothesis that because the distributions in the
different health zones happened consecutively, their distribution
partner must have used the same smartphones for them. The health area is
a property that is set once so that it does not have to be re-entered
for every household. Hence some distribution volunteers must have
forgotten to reset the property from a previous phase of the
distribution leading to erroneous meta data, which becomes evident
whenever there are villages with the same name in health areas that,
erroneously, received the same identifier.</p>
<p>I suggested that I would be able to fix the data given shape files of
the actual health areas, but none of us had access to any such files,
and I have not been able to find them online, just as I have only been
able to find very minimal data on the names and locations of villages in
the <span class="caps">DRC</span>.</p>
<p>On usability grounds, to avoid that the convex hulls of some villages
cover up other villages, I chose a cutoff at 35 km for the maximum
distance of any household from the cluster centroid, the most
parsimonious filter that allowed almost all cluster hulls to be
immediately visible for the user. Centroidal Voronoi tessellation (<span class="caps">CVT</span>)
came to mind too, but as the problem only concerned very few clusters,
<span class="caps">CVT</span> would have destroyed much information of the convex hulls of all the
correct clusters. For example it is interesting to derive from the data
an idea of how densely or sparsely populated some regions are.</p>
<p>Finally, I considered a geographic hierarchical clustering<sup id="fnref:24"><a class="footnote-ref" href="#fn:24">24</a></sup> to split
up the erroneous clusters, but given the pragmatic cutoff, the problem
only concerned 68 out of the over 10,000 clusters, and I expect for
future distributions the problem to be resolved on the side of the
distribution partner, so that this particular cleaning step would only
be of very temporary value.</p>
<h2 id="ch:appendices"><a class="toclink" href="#ch:appendices">Appendices</a></h2>
<blockquote>
<p>“Je n’ai fait celle-ci plus longue que parce que je n’ai pas eu le loisir de la faire plus courte.” (I made this [letter] very long only because I have not had the leisure to make it shorter.) —Blaise Pascal</p>
</blockquote>
<h3 id="app:ima-2014-pdrs"><a class="toclink" href="#app:ima-2014-pdrs"><span class="caps">IMA</span> 2014 Predistribution Registration Survey</a></h3>
<p>_ _ .</p>
<p>Field identifiers, questions, and answers with English translations in parentheses.</p>
<dl>
<dt>Identifier</dt>
<dd>
<p>identificationGPS_HHLatitude (identificationĠPSlatitude),
identificationGPS_HHLongitude (identificationĠPSlongitude),
identificationGPS_HHAltitude (identificationĠPSaltitude),
identificationGPS_HHAccuracy (identificationĠPSaccuracy)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Activer votre <span class="caps">GPS</span> pour chaque ménage” (“Enable <span class="caps">GPS</span> for each household”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>identificationdate_heure (identificationdateTime)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Entrez la date et l’heure” (“Enter the date and time”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>identificationphone_id (identificationphoneID)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Entrez votre telephone identification (N° Reco)” (“Enter your phone
identification (<span class="caps">ID</span> Reco)”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>identificationaire_de_sante_province (identificationhealthAreaprovince)</p>
</dd>
<dt>Question</dt>
<dd>
<p>Not given.</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>identificationaire_de_sante_district (identificationhealthAreadistrict)</p>
</dd>
<dt>Question</dt>
<dd>
<p>Not given.</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>identificationaire_de_sante_hz (identificationhealthAreahealthZone)</p>
</dd>
<dt>Question</dt>
<dd>
<p>Not given.</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>identificationaire_de_sante (identificationhealthAreatype)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Selectionner votre Aire de Santé” (“Select your Health Area”).</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>identificationnom_village (identificationvillage)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Tapez le nom du village” (“Type the name of the village”).</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>identificationLQAS_enumenumerer_jour (identificationdayOfDistribution)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Quel jour de la distribution est-ce?” (“Which day of the
distribution is it?”)</p>
</dd>
<dt>Answers</dt>
<dd>
<p>(1) “Jour 1” (“Day 1”), (2) “Jour 2” (“Day 2”), (3) “Jour 3” (“Day
3”), (4) “Jour 4” (“Day 4”), (5) “Jour 5” (“Day 5”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>identificationLQAS_enumenumerer_HH (identificationhouseNumber)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“N° Ménage” (“House number”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>identificationnomdechef_HH (identificationname)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Ecrire le nom de chef de ménage” (“Write the head of household name”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>identificationtelephone_HH (identificationphone)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“N° téléphone Chef ménage” (“Phone number of head of household”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>identificationmentions_legales (identificationlegalScript)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Bonjour. Votre participation à l’enquête de distribution
est souhaitée. Le but de l’enquête est de recueillir les
informations démographiques et déterminer si oui ou non vous avez
besoin de moustiquaires. Nous vous demandons de répondre à quelques
questions et entrer dans votre maison pour installer la <span class="caps">MILD</span>. Notre
entretien prendra environ 10 minutes. Si vous souhaitez que votre
image soit posté en ligne, nous pouvons prendre une photo avec la
<span class="caps">MILD</span>; si non on prendra l’image de la <span class="caps">MILD</span> sans vous. Êtes-vous
d’accord? Si oui, signez.” (“Hello. Your participation in the survey
distribution is desired. The purpose of the survey is to collect
demographic information and determine whether or not you need
mosquito nets. We ask you to answer a few questions and enter your
home to install the LLINs. Our interview will take about 10 minutes. If
you want your image to be posted online, we can take a picture with
the LLINs; if not, we will take the image of the <span class="caps">MILD</span> without you. Do
you agree? If so, sign.”)</p>
</dd>
<dt>Answers</dt>
<dd>
<p>(1) “Oui” (“Yes”), (2) “Non” (“No”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>identificationsignature_of_chef (identificationsignature)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Si vous acceptez de participer à l’enquête, s’il vous plaît signer
ici” (“If you agree to participate in the survey, please sign here”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nonsocio_demographiqueage_repondant (sociodemographicage)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Age du répondant (ou donné complète)” (“Age of respondent (or
given complete)”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nonsocio_demographiquesexe_repondant (sociodemographicsex)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Sexe du repondant” (“Sex of respondent”)</p>
</dd>
<dt>Answers</dt>
<dd>
<p>(1) “Male,” (2) “Female”</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nonsocio_demographiquetotal_personnes (sociodemographicpeopleInHousehold)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Nombre de personne dans le ménage” (“Number of people in household”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nonsocio_demographiquenbr_femmes_enceinte (sociodemographicpregnantWomen)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Nombre de femmes enceinte” (“Number of pregnant women”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nonsocio_demographiquenbr_enfant_sous_5 (sociodemographicċhildrenUnderFive)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Nombre des enfants en dessous de 5 ans” (“Number of children under
5 years”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nonconnaissance_perceptionmalaria_transmise (knowledgemalariatransmission)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Comment attrape-t-on la malaria?” (“How do you get malaria?”)</p>
</dd>
<dt>Answers</dt>
<dd>
<p>(1) “en étant piqué par une moustique femelle infectée” (“by being
bitten by an infected female mosquito”), (2) “en prenant un répas ou de
l’eau sale” (“taking a self-catering or dirty water”), (3) “etant
mouillé out frappé par soleil excessif” (“wet out being hit by excessive
sun”), (4) “un environnement sale” (“dirty environment”), (5)
“sorcellerie” (“witchcraft”), (6) “en mangeant les mangues ou les fruits
verts” (“eating mangoes or green fruit”), (7) “en vivant avec les
personnes malades” (“living with sick people”), (8) “autre
(question suivante)” (“other (question)”), (9) “n’est pas connaitre /
ignoré” (“does not know / ignored”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nonconnaissance_perceptionmalaria_transmise_autre (knowledgemalariaotherTransmission)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Entrez autre raisons” (“Enter other reasons”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nonconnaissance_perceptionmalaria_signes (knowledgemalariasymptoms)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Pourriez vous nous citez 4 signes / symptômes de la malaria?”</p>
</dd>
<dt>Answers</dt>
<dd>
<p>(1) “fièvre / haute température” (“fever / high temperature”), (2) “maux
de tête” (“headache”), (3) “douleurs au niveau de l’articulations /
douleur général du corps” (“pain in the joints / general body
pain”), (4) “fatigue” (“4) “tiredness”), (5) “vomissement” (“5)
“vomiting”), (6) “manqué d’appétit / goût amère dans la bouche” (“6)
“lacked appetite / bitter taste in the mouth”), (7) “diarrhée / douleur
général” (“diarrhea / General pain”), (8) “autre (question suivante)”
(“other (question)”), (9) “n’est pas connaitre / ignoré” (“does not know
/ ignored”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nonconnaissance_perceptionmalaria_signes_autre (knowledgemalariaotherSymptoms)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Entrez autre raisons” (“Enter other reasons”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nonconnaissance_perceptionmalaria_prevenir (knowledgemalariaprevention)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Que pensez vous qu’on doit faire pour se prévenir de la malaria”
(“What do you think we should do to prevent malaria”)</p>
</dd>
<dt>Answers</dt>
<dd>
<p>(1) “creuser les canalisation d’eau stagnante” (“digging stagnant water
pipeline”), (2) “nettoyer les herbes grandissante” (“clean growing
herbs”), (3) “dormir sous Moustiquaire impregnée d’insecticide” (“sleep
under mosquito net impregnated with insecticide”), (4) “utilisez les
contraceptif contre le moustique” (“contraceptive use against
mosquitoes”), (5) “bruler les herbes” (“burning herbs”), (6) “prendre
les medicaments anti malaria” (“take anti malaria drugs”), (7) “utiliser
les insecticides” (“use insecticides”), (8) “autre (question suivante)”
(“other (question)”), (9) “n’est pas connaitre / ignoré” (“does not know
/ ignored”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nonconnaissance_perceptionmalaria_traitment_simple (knowledgemalariatreatment)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Quel est le medicament recommendé pour le traitement de la
malaria?” (“What is the recommended drug for the treatment of malaria?”)</p>
</dd>
<dt>Answers</dt>
<dd>
<p>(1) “<span class="caps">SP</span>/Fansidar” (“<span class="caps">SP</span>/Fansidar”), (2) “chloroquine”
(“chloroquine”), (3) “amodiaquine/camoquine”
(“amodiaquine/camoquine”), (4) “quinine” (“quinine”), (5)
“artemisinin-base sur combination” (“artemisinin-based
combination”), (6) “paracetamol” (“paracetamol”), (7)
“aspirine/ibuprofen” (“aspirin/ibuprofen”), (8) “autre” (“other”), (9)
“Je ne sais pas” (“I do not know”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nonconnaissance_perceptionmalaria_traitment_autre (knowledgemalariaotherTreatment)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Entrez autre traitment” (“Enter another treatment”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nonpreventionmoins_une_MILD_HH (preventionatLeastOneNet)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Est ce que votre ménage a au moins une Moustiquaire” (“Does your
household have at least one mosquito net”)</p>
</dd>
<dt>Answers</dt>
<dd>
<p>(1) “Oui” (“Yes”), (2) “Non” (“No”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nonpreventionmoins_impregnee_HH (preventionatLeastOneLLIN)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Y-a-t-il au moins une Moustiquaires imprégée d’insecticide dans
votre ménage” (“Do you have at least one insecticide-treated
mosquito net in your household?”)</p>
</dd>
<dt>Answers</dt>
<dd>
<p>(1) “Oui” (“Yes”), (2) “Non” (“No”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nonpreventionnbr_MILD_impregnee (preventionLLINs)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Combien de Moustiquaires imprégée avez vous dans votre ménage”
(“How many LLINs do you have in your household?”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nonpreventionobtenu_MILD_ou (preventionnetSource)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Où avez-vous obtenu votre Moustiquaire?” (“Where did you get your
mosquito net?”)</p>
</dd>
<dt>Answers</dt>
<dd>
<p>(1) “marché/magasin/mharmacie” (“market/store/pharmacy”), (2) “<span class="caps">ONG</span>”
(“<span class="caps">NGO</span>”), (3) “campagne de distribution” (“distribution campaign”), (4)
“centre de santé” (“health center”), (5) “Autres“ (“other”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nonpreventionobtenu_MILD_ou_autre (preventionotherNetSource)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Entrez autre” (“Enter other”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nonpreventiontout_personnes_dormi (preventionpeopleUnderNetLastNight)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Combien des personnes dans votre maison ont dormi sous la
Moustiquaires la nuit précédente?” (“How many people in your house
slept under the nets last night?”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nonpreventionenceintes_dormi (preventionpregnantWomenUnderNetLastNight)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Combien de femmes enceintes ont dormi sous la moustiquaire la nuit
dernière?” (“How many pregnant women slept under the net last night?”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nonpreventionIPT_utilisez (preventionfansidarDuringPregnancy)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Combien de fois a-t-elle pris le Fansidar (<span class="caps">TPI</span>) pendant sa
grossesse pour prevenir la malaria?” (“How many times did she take
Fansidar (<span class="caps">IPT</span>) during her pregnancy to prevent malaria?”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nonpreventionenfants_dormi (preventionallChildrenUnderNetLastNight)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Est-ce que tous les enfants de moins de 5 ans ont dormi
sous la moustiquaire la nuit dernière?” (“Did all children under 5
years of age sleep under the net last night?”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nonpreventionprevention_dans_HH (preventionpreventiveMeasures)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Quelles sont les mesures de prevention faites vous pour se prevenir
de la malaria dans votre menage?” (“What preventive measures do you
do to prevent malaria in your household?”)</p>
</dd>
<dt>Answers</dt>
<dd>
<p>(1) “avoir enlevé l’eau stagnante” (“removing standing water”), (2)
“utiliser la lotion anti moustique” (“Use mosquito repellent
lotion”), (3) “pulverisation des maisons” (“spraying houses”), (4)
“faucher les herbes” (“mowing grass”), (5) “nettoyer l’environnement”
(“Clean Environment”), (6) ”couvrir les recipients d’eau” (“cover water
containers”), (7) “bruler des feuilles, fumier, etc.” (“burning leaves,
manure, etc.”), (8) “rein” (“kidney”), (9) “n’est pas connaitre /
ignoré” (“does not know / ignored”), (10) “autre (question suivante)
(“other (question)”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nonpreventionprevention_dans_HH_autre (preventionotherPreventiveMeasures)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Entrer d’autres méthodes” (“Enter other methods”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nontraitementenfant_fievre (treatmentċhildfeaverLastTwoWeeks)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Dans le deux semaines y-a-t-il un enfant en dessous de 5 ans qui a
souffert de la fièvre?” (“In the two weeks there-he has a child
under 5 years who had fever?”)</p>
</dd>
<dt>Answers</dt>
<dd>
<p>(1) “Oui” (“Yes”), (2) “Non” (“No”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nontraitementenfant_beneficier (treatmentċhildbenefitFromDrug)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“L’enfant a-t-il beneficier du medicament contre la fievre ou la
malaria?” (“Does the child benefit from the drug against fever or malaria?”)</p>
</dd>
<dt>Answers</dt>
<dd>
<p>(1) “Oui” (“Yes”), (2) “Non” (“No”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nontraitementenfant_medicament (treatmentċhildmedicine)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Quelle sorte de medicament que l’enfant a pris?” (“What kind of
medicine that the child has taken?”)</p>
</dd>
<dt>Answers</dt>
<dd>
<p>(1) “<span class="caps">SP</span>/Fansidar” (“<span class="caps">SP</span>/Fansidar”), (2) “chloroquine”
(“chloroquine”), (3) “amodiaquine/camoquine”
(“amodiaquine/camoquine”), (4) “quinine” (“quinine”), (5)
“artemisinin-base sur combination” (“artemisinin-based
combination”), (6) paracetamol” (“paracetamol”), (7)
“aspirine/ibuprofen” (“aspirin/ibuprofen”), (8) “autre” (“other”), (9)
“Je ne sais pas” (“I do not know”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nontraitementconsulte_enfant (treatmentċhildhealthcenter)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“A-t-on consulté l’enfant au centre de santé pendant sa maladie.”
(“Was the child taken to a consultation at the health center during
his illness.”)</p>
</dd>
<dt>Answers</dt>
<dd>
<p>(1) “Oui” (“Yes”), (2) “Non” (“No”), (3) “Je ne sais pas” (“I do not know”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nondistributionplace_dormir (distributionsleepingSpaces)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Nombre de place à dormir” (“Number of sleeping spaces”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nondistributionnbr_bonne_mild (distributionLLINsġood)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Combien de <span class="caps">MILD</span> de bonne qualité sont disponibles dans le ménage”
(“How many good LLINs are available in the household”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nondistributionMILD_par_HH (distributionLLINstotal)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nondistributiondistrbution_summary (distributiondistrbutionSummary)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nondistributionMILD_installees (distributionLLINsinstalled)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Entrez Nombre de <span class="caps">MILD</span> installées” (“Enter Number of LLINs installed”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nondistributionMILD_marque (distributionLLINsbrand)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Choisissez la marque de moustiquaires installées” (“Choose the
brand of nets installed”)</p>
</dd>
<dt>Answers</dt>
<dd>
<p>(1) Olyset, (2) PermaNet</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nondistributionMILD_retournees (distributionLLINsreturned)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Entrez Nombre de <span class="caps">MILD</span> retournées” (“Enter number of LLINs returned”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nondistributionMILD_emballage (distributionLLINspackagingReturned)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Entrez Nombre d’emballage” (“Enter Number packaging collected”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>legales_nondistributionphoto_installee (distributionLLINsphoto)</p>
</dd>
<dt>Question</dt>
<dd>
<p>“Entrez la photo de <span class="caps">MILD</span> installées” (“Enter the photo of LLINs installed”)</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>start (start)</p>
</dd>
<dt>Description</dt>
<dd>
<p>Start time</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>end (end)</p>
</dd>
<dt>Description</dt>
<dd>
<p>End time</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>deviceid (deviceID)</p>
</dd>
<dt>Description</dt>
<dd>
<p>Phone serial number</p>
</dd>
</dl>
<!-- -->
<dl>
<dt>Identifier</dt>
<dd>
<p>metainstanceID (metainstanceID)</p>
</dd>
</dl>
<h2 id="endnotes"><a class="toclink" href="#endnotes">Endnotes</a></h2>
<h2 id="bibliography"><a class="toclink" href="#bibliography">Bibliography</a></h2>
<div class="footnote">
<hr>
<ol>
<li id="fn:1">
<p>“The figures for between the richest 1% and the richest 21% are
based on micro data from national household surveys carried out in
2008, kindly provided by Branko Milanovic, author of <em>The Haves and
the Have-Nots</em>. The figures for the poorest 73% are based on the
2008 data from PovcalNet, adjusted based on the approximation that
the surveys covered unbiased samples of the poorest 80% of the
world’s population. The figure of \$70,000 for the top 0.1% is from
Milanovic’s book. All figures have been adjusted for the Consumer
Price Index measure of inflation.”<sup id="fnref:GWWCHowRich"><a class="footnote-ref" href="#fn:GWWCHowRich">75</a></sup> <a class="footnote-backref" href="#fnref:1" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
<li id="fn:2">
<p>The author adds up the death tolls of the leading poverty-related
causes of death, which I divided by 365:
<span class="math">\(\frac{12.7 ~\text{million}}{365} \approx 34\,795\)</span>. <a class="footnote-backref" href="#fnref:2" title="Jump back to footnote 2 in the text">↩</a></p>
</li>
<li id="fn:3">
<p>The estimate are based on GiveWell’s and Animal Charity
Evaluators’ cost-effectiveness estimates of bednet distributions,
animal shelters, and farmed animal advocacy programs respectively.
The costs are inflation-adjusted according to the Consumer Price
Index. <a class="footnote-backref" href="#fnref:3" title="Jump back to footnote 3 in the text">↩</a></p>
</li>
<li id="fn:4">
<p>An example from a personal correspondence: “I’d happily condemn a
hundred thousand strangers in a developing country to death in
exchange for just one person I know to get better, simply because I
couldn’t care less about the life of a complete stranger – I don’t
wish him any harm, but I am hard pressed to find a reason why I
should help him. He doesn’t affect my life in the slightest; he’s
just void.” <a class="footnote-backref" href="#fnref:4" title="Jump back to footnote 4 in the text">↩</a></p>
</li>
<li id="fn:5">
<p>Time as well as money in the case of formal education in many
countries. Access to information can also cost money in cases were
it requires computer hardware and Internet access. <a class="footnote-backref" href="#fnref:5" title="Jump back to footnote 5 in the text">↩</a></p>
</li>
<li id="fn:6">
<p>These thoughts are inspired by Bostrom’s
orthogonality thesis.<sup id="fnref:bostrom2012superintelligent"><a class="footnote-ref" href="#fn:bostrom2012superintelligent">41</a></sup> <a class="footnote-backref" href="#fnref:6" title="Jump back to footnote 6 in the text">↩</a></p>
</li>
<li id="fn:7">
<p>We should note, however, that the distinction between ethical and
epistemic differences can be a fuzzy one. When someone’s (epistemic)
beliefs are very different from one’s own, it can be very costly to
convince the other person all the way to one’s own view even if one
is objectively right, so that compromise can again become more
efficient. <a class="footnote-backref" href="#fnref:7" title="Jump back to footnote 7 in the text">↩</a></p>
</li>
<li id="fn:8">
<p>The third edition of DCP is still forthcoming at the time of writing. <a class="footnote-backref" href="#fnref:8" title="Jump back to footnote 8 in the text">↩</a></p>
</li>
<li id="fn:9">
<p>In order to recommend a charity, GiveWell requires detailed
information from the charity, so that the cooperation of the charity
in the process is paramount. Hence there may be excellent charities
besides GiveWell’s top charities, but these charities merely
declined to cooperate with GiveWell. One reason for that can be that
they are already receiving enough funding, so that extra funding is
not worth the staff time for them. <a class="footnote-backref" href="#fnref:9" title="Jump back to footnote 9 in the text">↩</a></p>
</li>
<li id="fn:10">
<p>I am ignoring deworming in this comparison because GiveWell found
errors in the DCP2 estimates for the intervention.<sup id="fnref2:GiveWell2011Errors"><a class="footnote-ref" href="#fn:GiveWell2011Errors">70</a></sup> <a class="footnote-backref" href="#fnref:10" title="Jump back to footnote 10 in the text">↩</a></p>
</li>
<li id="fn:11">
<p>Evidence Action has published an article contrasting these
sources of insight, but the article seems to sell a standard
practice as something controversial and contrarian.<sup id="fnref:Nassar2015"><a class="footnote-ref" href="#fn:Nassar2015">108</a></sup> <a class="footnote-backref" href="#fnref:11" title="Jump back to footnote 11 in the text">↩</a></p>
</li>
<li id="fn:12">
<p>Paraphrased from a verbal conversation. <a class="footnote-backref" href="#fnref:12" title="Jump back to footnote 12 in the text">↩</a></p>
</li>
<li id="fn:13">
<p>At the time of my writing, the process page lists two, but they
have hired an expert on farmed animal welfare as well,,
<sup id="fnref:OpenPhil2015Bollard"><a class="footnote-ref" href="#fn:OpenPhil2015Bollard">111</a></sup> implying that they have completed these
stages for a third area. <a class="footnote-backref" href="#fnref:13" title="Jump back to footnote 13 in the text">↩</a></p>
</li>
<li id="fn:14">
<p>The informal experience of the Open Philanthropy Project,
however, has shown that taking away this constriction does not
necessarily yield higher expected return but rather merely greater
funding gaps, and these are not critical for most individual donors,
so that they are not necessarily missing out.<sup id="fnref:Matthews2015"><a class="footnote-ref" href="#fn:Matthews2015">103</a></sup> <a class="footnote-backref" href="#fnref:14" title="Jump back to footnote 14 in the text">↩</a></p>
</li>
<li id="fn:15">
<p>The <a href="/concept-for-donor-coordination.html">donor coordination system</a> aims to address this problem. <a class="footnote-backref" href="#fnref:15" title="Jump back to footnote 15 in the text">↩</a></p>
</li>
<li id="fn:16">
<p>Rob Mather, email conversation with the author, January 2016.
Cited with permission. All further quotations are from the same
conversation unless marked otherwise. <a class="footnote-backref" href="#fnref:16" title="Jump back to footnote 16 in the text">↩</a></p>
</li>
<li id="fn:17">
<p>The widespread misconception that the ratio of program funding to
administrative funding says anything about the effectiveness of a
charity. <a class="footnote-backref" href="#fnref:17" title="Jump back to footnote 17 in the text">↩</a></p>
</li>
<li id="fn:18">
<p>But not that Rob Mather added that “over the years it has been
difficult, for those attempting to collect data and get a clear
picture, to know numbers with great certainty. Collating information
from all funders, including governments both in and outside Africa,
seems challenging.” <a class="footnote-backref" href="#fnref:18" title="Jump back to footnote 18 in the text">↩</a></p>
</li>
<li id="fn:19">
<p>Rather intrusively, the survey did not ask for people’s gender
but their sex, and about 111,000 (45%) of the respondents answered
that with “woman” while about 124,000 (51%) responded “man.” Almost
10,000 (4%) did not answer, maybe because they found neither
appropriate, the question too private, or for various other reasons. <a class="footnote-backref" href="#fnref:19" title="Jump back to footnote 19 in the text">↩</a></p>
</li>
<li id="fn:20">
<p>MapQuest unfortunately shut down access to its map, which I had
used as default because if its geographic features. <a class="footnote-backref" href="#fnref:20" title="Jump back to footnote 20 in the text">↩</a></p>
</li>
<li id="fn:21">
<p>The current figures are consumption based. <a class="footnote-backref" href="#fnref:21" title="Jump back to footnote 21 in the text">↩</a></p>
</li>
<li id="fn:22">
<p>Several types of features displayed by the JavaScript-driven map
application are automatically filtered by country to limit the
number of features that are displayed at the same time. A larger
number would reduce the responsiveness of the application. <a class="footnote-backref" href="#fnref:22" title="Jump back to footnote 22 in the text">↩</a></p>
</li>
<li id="fn:23">
<p>According to IR Mapper, the entomological inoculation rate of
Syria, Iran, and Iraq is negligible outside very small regions to
the south of Iran, and the same applies to most of China. <a class="footnote-backref" href="#fnref:23" title="Jump back to footnote 23 in the text">↩</a></p>
</li>
<li id="fn:24">
<p><span class="math">\(k\)</span>-means clustering would be less apposite since it would take
much experimentation to find a suitable number of clusters for the
algorithm. <a class="footnote-backref" href="#fnref:24" title="Jump back to footnote 24 in the text">↩</a></p>
</li>
<li id="fn:25">
<p>Please note that in the following I will use “intervention” and
“program” semantically interchangeably conditional on which terms
seems more idiomatic to me in the collocational context. <a class="footnote-backref" href="#fnref:25" title="Jump back to footnote 25 in the text">↩</a></p>
</li>
<li id="fn:26">
<p>“Respectable,” here, is not meant to denigrate any other
hypothetical prioritization organizations but rather meant as a
handicap, since an organization that is highly respected has to go
to great lengths to stress the low quality of its research when it
wants to invest staff time proportionate to evaluating interventions
with small funding gaps lest donors assume that the results are as
reliable as other results the organization puts out. Taking such a
risk is rarely warranted for such an organization. <a class="footnote-backref" href="#fnref:26" title="Jump back to footnote 26 in the text">↩</a></p>
</li>
<li id="fn:adelman2013index">
<p>Adelman, Carol, Yulya Spantchak, Kacie Marano, Jeremiah Norris, Jesse Barnett, Haowen Chen, Darice Xue, and Aubrey Thrane. “The Index of Global Philanthropy and Remittances, with a Special Report on Emerging Economies.” Hudson Institute Report, November (2013). <a class="footnote-backref" href="#fnref:adelman2013index" title="Jump back to footnote 27 in the text">↩</a></p>
</li>
<li id="fn:Leaflet">
<p>Agafonkin, Vladimir. “Leaflet: A JavaScript Library for Interactive Maps.” 2015. Accessed January 12, 2016. <a href="http://leafletjs.com/">http://leafletjs.com/</a>. <a class="footnote-backref" href="#fnref:Leaflet" title="Jump back to footnote 28 in the text">↩</a></p>
</li>
<li id="fn:AMFNets">
<p>Against Malaria Foundation. “About Mosquito Nets.” 2015. Accessed December 30, 2015. <a href="https://www.againstmalaria.com/FAQ_Bednets.aspx">https://www.againstmalaria.com/FAQ_Bednets.aspx</a>. <a class="footnote-backref" href="#fnref:AMFNets" title="Jump back to footnote 29 in the text">↩</a></p>
</li>
<li id="fn:AMFDelays2013">
<p>---. “AMF Update – 28 November 2013.” 2013. Accessed January 29, 2016. <a href="https://www.againstmalaria.com/Downloads/AMF_Update_28Nov2013.pdf">https://www.againstmalaria.com/Downloads/AMF_Update_28Nov2013.pdf</a>. <a class="footnote-backref" href="#fnref:AMFDelays2013" title="Jump back to footnote 30 in the text">↩</a></p>
</li>
<li id="fn:AMFFactories">
<p>---. “Question to AMF: ‘Where are Your Nets Manufactured?’” 2015. Accessed December 30, 2015. <a href="https://www.againstmalaria.com/Newsitem.aspx?NewsItem=Question-to-AMF-Where-are-your-nets-manufactured">https://www.againstmalaria.com/Newsitem.aspx?NewsItem=Question-to-AMF-Where-are-your-nets-manufactured</a>. <a class="footnote-backref" href="#fnref:AMFFactories" title="Jump back to footnote 31 in the text">↩</a></p>
</li>
<li id="fn:ACEImpactDonations">
<p>Animal Charity Evaluators. “Impact of Donations.” 2015. Accessed September 6, 2015. <a href="http://www.animalcharityevaluators.org/research/foundational-research/impact-donations/">http://www.animalcharityevaluators.org/research/foundational-research/impact-donations/</a>. <a class="footnote-backref" href="#fnref:ACEImpactDonations" title="Jump back to footnote 32 in the text">↩</a></p>
</li>
<li id="fn:SVN">
<p>Apache Software Foundation. “Apache Subversion.” 2015. Accessed January 12, 2016. <a href="https://subversion.apache.org/">https://subversion.apache.org/</a>. <a class="footnote-backref" href="#fnref:SVN" title="Jump back to footnote 33 in the text">↩</a></p>
</li>
<li id="fn:axelrod2009evolution">
<p>Axelrod, Robert. The Evolution of Cooperation: Revised Edition. Basic Books, 2009. isbn:978-0-78-673488-7. <a href="https://books.google.de/books?id=GxRo5hZtxkEC">https://books.google.de/books?id=GxRo5hZtxkEC</a>. <a class="footnote-backref" href="#fnref:axelrod2009evolution" title="Jump back to footnote 34 in the text">↩</a></p>
</li>
<li id="fn:banerjee2012poor">
<p>Banerjee, Abhijit, and Esther Duflo. Poor Economics: A Radical Rethinking of the Way to Fight Global Poverty. PublicAffairs, 2012. isbn: 978-1-61-039160-3. <a href="https://books.google.de/books?id=2dlnBoX4licC">https://books.google.de/books?id=2dlnBoX4licC</a>. <a class="footnote-backref" href="#fnref:banerjee2012poor" title="Jump back to footnote 35 in the text">↩</a></p>
</li>
<li id="fn:BBCTito">
<p>BBC News. “Profile: Tito the Spaceman.” 2001. Accessed December 5, 2015. <a href="http://news.bbc.co.uk/2/hi/science/nature/1297924.stm">http://news.bbc.co.uk/2/hi/science/nature/1297924.stm</a>. <a class="footnote-backref" href="#fnref:BBCTito" title="Jump back to footnote 36 in the text">↩</a></p>
</li>
<li id="fn:Wat">
<p>Bernhardt, Gary. “Wat.” 2012. Accessed March 7, 2016. <a href="https://www.destroyallsoftware.com/talks/wat">https://www.destroyallsoftware.com/talks/wat</a>. <a class="footnote-backref" href="#fnref:Wat" title="Jump back to footnote 37 in the text">↩</a></p>
</li>
<li id="fn:GatesDCPP2002">
<p>Bill & Melinda Gates Foundation. “New Initiative Will Assess Disease Control Priorities In Developing Countries.” 2002. Accessed December 9, 2015. <a href="http://www.gatesfoundation.org/Media-Center/Press-Releases/2002/09/Initiative-to-Assess-Disease-Control-Priorities-in-Developing-Countries">http://www.gatesfoundation.org/Media-Center/Press-Releases/2002/09/Initiative-to-Assess-Disease-Control-Priorities-in-Developing-Countries</a>. <a class="footnote-backref" href="#fnref:GatesDCPP2002" title="Jump back to footnote 38 in the text">↩</a></p>
</li>
<li id="fn:C3">
<p>Bostock, Mike. “C3.js: D3-Based Reusable Chart Library.” 2016. Accessed March 12, 2016. <a href="http://c3js.org/">http://c3js.org/</a>. <a class="footnote-backref" href="#fnref:C3" title="Jump back to footnote 39 in the text">↩</a></p>
</li>
<li id="fn:D3">
<p>Bostock, Mike. “D3.js: Data-Driven Documents.” 2016. Accessed March 12, 2016. <a href="https://d3js.org/">https://d3js.org/</a>. <a class="footnote-backref" href="#fnref:D3" title="Jump back to footnote 40 in the text">↩</a></p>
</li>
<li id="fn:bostrom2012superintelligent">
<p>Bostrom, Nick. “The Superintelligent Will: Motivation and Instrumental Rationality in Advanced Artificial Agents.” Minds and Machines 22, no. 2 (2012): 71–85. <a class="footnote-backref" href="#fnref:bostrom2012superintelligent" title="Jump back to footnote 41 in the text">↩</a></p>
</li>
<li id="fn:Sphinx">
<p>Brandl, Georg, and the Sphinx team. “Sphinx.” 2016. Accessed March 12, 2016. <a href="http://www.sphinx-doc.org/">http://www.sphinx-doc.org/</a>. <a class="footnote-backref" href="#fnref:Sphinx" title="Jump back to footnote 42 in the text">↩</a></p>
</li>
<li id="fn:CDCMalaria">
<p>CDC. “Malaria – About Malaria – Biology – Mosquitoes – Anopheles Mosquitoes.” 2016. Accessed January 22, 2016. <a href="http://www.cdc.gov/malaria/about/biology/mosquitoes/">http://www.cdc.gov/malaria/about/biology/mosquitoes/</a>. <a class="footnote-backref" href="#fnref:CDCMalaria" title="Jump back to footnote 43 in the text">↩</a></p>
</li>
<li id="fn:CIA2015">
<p>Central Intelligence Agency. “The World Factbook: Congo, Democratic Republic of the.” 2015. Accessed January 3, 2016. <a href="https://www.cia.gov/library/publications/the-world-factbook/geos/cg.html">https://www.cia.gov/library/publications/the-world-factbook/geos/cg.html</a>. <a class="footnote-backref" href="#fnref:CIA2015" title="Jump back to footnote 44 in the text">↩</a></p>
</li>
<li id="fn:chandy2011poverty">
<p>Chandy, Laurence, and Geoffrey Gertz. Poverty in Numbers: The Changing State of Global Poverty from 2005 to 2015. Brookings Institution, 2011. <a class="footnote-backref" href="#fnref:chandy2011poverty" title="Jump back to footnote 45 in the text">↩</a></p>
</li>
<li id="fn:CherryPy">
<p>CherryPy. “CherryPy: A Minimalist Python Web Framework.” 2015. Accessed January 12, 2016. <a href="http://www.cherrypy.org/">http://www.cherrypy.org/</a>. <a class="footnote-backref" href="#fnref:CherryPy" title="Jump back to footnote 46 in the text">↩</a></p>
</li>
<li id="fn:CoffeeScript">
<p>CoffeeScript. “CoffeeScript.” 2016. Accessed March 7, 2016. <a href="http://coffeescript.org/">http://coffeescript.org/</a>. <a class="footnote-backref" href="#fnref:CoffeeScript" title="Jump back to footnote 47 in the text">↩</a></p>
</li>
<li id="fn:crompton2010advances">
<p>Crompton, Peter D., Susan K. Pierce, and Louis H. Miller. “Advances and Challenges in Malaria Vaccine Development.” The Journal of clinical investigation 120, no. 12 (2010): 4168. <a class="footnote-backref" href="#fnref:crompton2010advances" title="Jump back to footnote 48 in the text">↩</a></p>
</li>
<li id="fn:darley1968bystander">
<p>Darley, John M., and Bibb Latane. “Bystander Intervention in Emergencies: Diffusion of Responsibility.” Journal of personality and social psychology 8, no. 4p1 (1968): 377. <a class="footnote-backref" href="#fnref:darley1968bystander" title="Jump back to footnote 49 in the text">↩</a></p>
</li>
<li id="fn:desvousges2010measuring">
<p>Desvousges, William H, F Reed Johnson, Richard W Dunford, Kevin J Boyle, Sara P Hudson, K Nicole Wilson, et al. Measuring Nonuse Damages Using Contingent Valuation: An Experimental Evaluation of Accuracy. Research Triangle Park, NC: RTI Press, 2010. <a class="footnote-backref" href="#fnref:desvousges2010measuring" title="Jump back to footnote 50 in the text">↩</a></p>
</li>
<li id="fn:WeltRügenwalder">
<p>Dierig, Carsten. “Warum die Deutschen so gerne Veggie-Wurst essen.” 2015. Accessed December 30, 2015. <a href="http://www.welt.de/wirtschaft/article145694267/Warum-die-Deutschen-so-gerne-Veggie-Wurst-essen.html">http://www.welt.de/wirtschaft/article145694267/Warum-die-Deutschen-so-gerne-Veggie-Wurst-essen.html</a>. <a class="footnote-backref" href="#fnref:WeltRügenwalder" title="Jump back to footnote 51 in the text">↩</a></p>
</li>
<li id="fn:MedianIncome">
<p>Diofasi, Anna, and Nancy Birdsall. “The World Bank’s Poverty Statistics Lack Median Income Data, So We Filled In the Gap Ourselves: Download Available.” 2016. Accessed July 8, 2016. <a href="http://www.cgdev.org/blog/world-bank-poverty-statistics-lack-median-income-data-so-we-filled-gap-ourselves-download-available">http://www.cgdev.org/blog/world-bank-poverty-statistics-lack-median-income-data-so-we-filled-gap-ourselves-download-available</a>. <a class="footnote-backref" href="#fnref:MedianIncome" title="Jump back to footnote 52 in the text">↩</a><a class="footnote-backref" href="#fnref2:MedianIncome" title="Jump back to footnote 52 in the text">↩</a></p>
</li>
<li id="fn:GeoDjango">
<p>Django Software Foundation. “GeoDjango: A World-Class Geographic Web Framework.” 2016. Accessed March 12, 2016. <a href="http://geodjango.org/">http://geodjango.org/</a>. <a class="footnote-backref" href="#fnref:GeoDjango" title="Jump back to footnote 53 in the text">↩</a></p>
</li>
<li id="fn:LLINVisualizerVideo">
<p>Drescher, Denis. “LLIN Visualizer Demo.” 2016. Accessed August 6, 2016. <a href="https://www.youtube.com/watch?v=S5K3aF21E4Y">https://www.youtube.com/watch?v=S5K3aF21E4Y</a>. <a class="footnote-backref" href="#fnref:LLINVisualizerVideo" title="Jump back to footnote 54 in the text">↩</a></p>
</li>
<li id="fn:MAWAMFGuesstimate">
<p>---. “Make-A-Wish Dollars Per DALY Averted.” 2016. Accessed March 5, 2016. <a href="https://www.getguesstimate.com/models/4296">https://www.getguesstimate.com/models/4296</a>. <a class="footnote-backref" href="#fnref:MAWAMFGuesstimate" title="Jump back to footnote 55 in the text">↩</a></p>
</li>
<li id="fn:LLINVisualizerInstance">
<p>---. “Map – LLIN Visualizer.” 2016. Accessed July 14, 2016. <a href="http://nets.claviger.net/">http://nets.claviger.net/</a>. <a class="footnote-backref" href="#fnref:LLINVisualizerInstance" title="Jump back to footnote 56 in the text">↩</a></p>
</li>
<li id="fn:GuesstimateMosquitoYLL">
<p>---. “Mosquito YLL Per Human ¬YLL.” 2016. Accessed March 7, 2016. <a href="https://www.getguesstimate.com/models/4302">https://www.getguesstimate.com/models/4302</a>. <a class="footnote-backref" href="#fnref:GuesstimateMosquitoYLL" title="Jump back to footnote 57 in the text">↩</a></p>
</li>
<li id="fn:LLINVisualizerCode">
<p>---. “Telofy / LLIN Visualizer – Bitbucket.” 2016. Accessed July 14, 2016. <a href="https://bitbucket.org/Telofy/llin-visualizer">https://bitbucket.org/Telofy/llin-visualizer</a>. <a class="footnote-backref" href="#fnref:LLINVisualizerCode" title="Jump back to footnote 58 in the text">↩</a></p>
</li>
<li id="fn:ECMAScript">
<p>Ecma International. “ECMAScript.” 2016. Accessed March 12, 2016. <a href="http://www.ecmascript.org/">http://www.ecmascript.org/</a>. <a class="footnote-backref" href="#fnref:ECMAScript" title="Jump back to footnote 59 in the text">↩</a></p>
</li>
<li id="fn:ErlandssonIVE">
<p>Erlandsson, Arvid, Fredrik Björklund, and Martin Bäckström. “Emotional Reactions, Perceived Impact and Perceived Responsibility Mediate the Identifiable Victim Effect, Proportion Dominance Effect and In-Group Effect Respectively,” Organizational Behavior and Human Decision Processes, 127, no. March (2015): 1–14. issn: 0749-5978. <a href="http://dx.doi.org/10.1016/j.obhdp.2014.11.003">http://dx.doi.org/10.1016/j.obhdp.2014.11.003</a>. <a class="footnote-backref" href="#fnref:ErlandssonIVE" title="Jump back to footnote 60 in the text">↩</a></p>
</li>
<li id="fn:Flask">
<p>Flask. “Flask: A Python Microframework.” 2014. Accessed January 12, 2016. <a href="http://flask.pocoo.org/">http://flask.pocoo.org/</a>. <a class="footnote-backref" href="#fnref:Flask" title="Jump back to footnote 61 in the text">↩</a></p>
</li>
<li id="fn:Fabric">
<p>Forcier, Jeff. “Fabric.” 2016. Accessed March 12, 2016. <a href="http://www.fabfile.org/">http://www.fabfile.org/</a>. <a class="footnote-backref" href="#fnref:Fabric" title="Jump back to footnote 62 in the text">↩</a></p>
</li>
<li id="fn:CVS">
<p>Free Software Foundation. “CVS: Concurrent Versions System.” 2015. Accessed January 12, 2016. <a href="http://www.nongnu.org/cvs/">http://www.nongnu.org/cvs/</a>. <a class="footnote-backref" href="#fnref:CVS" title="Jump back to footnote 63 in the text">↩</a></p>
</li>
<li id="fn:GallupWorldPoll">
<p>Gallup. “How Does the Gallup World Poll Work?” 2016. Accessed July 9, 2016. <a href="http://www.gallup.com/178667/gallup-world-poll-work.aspx">http://www.gallup.com/178667/gallup-world-poll-work.aspx</a>. <a class="footnote-backref" href="#fnref:GallupWorldPoll" title="Jump back to footnote 64 in the text">↩</a></p>
</li>
<li id="fn:GallupLifeLadder">
<p>---. “Understanding How Gallup Uses the Cantril Scale.” 2016. Accessed July 8, 2016. <a href="http://www.gallup.com/poll/122453/understanding-gallup-uses-cantril-scale.aspx">http://www.gallup.com/poll/122453/understanding-gallup-uses-cantril-scale.aspx</a>. <a class="footnote-backref" href="#fnref:GallupLifeLadder" title="Jump back to footnote 65 in the text">↩</a><a class="footnote-backref" href="#fnref2:GallupLifeLadder" title="Jump back to footnote 65 in the text">↩</a></p>
</li>
<li id="fn:GiveWellMarcyErskine">
<p>GiveWell. “A Conversation with Marcy Erskine on September 18, 2013.” 2013. Accessed January 1, 2016. <a href="http://files.givewell.org/files/conversations/Marcy%20Erskine%209-18-13%20(public).pdf">http://files.givewell.org/files/conversations/Marcy%20Erskine%209-18-13%20(public).pdf</a>. <a class="footnote-backref" href="#fnref:GiveWellMarcyErskine" title="Jump back to footnote 66 in the text">↩</a></p>
</li>
<li id="fn:GWLengeler2015">
<p>---. “A Conversation with Professor Christian Lengeler, August 20, 2015.” 2015. Accessed February 7, 2016. <a href="http://files.givewell.org/files/conversations/Christian_Lengeler_08-20-2015_(public).pdf">http://files.givewell.org/files/conversations/Christian_Lengeler_08-20-2015_(public).pdf</a>. <a class="footnote-backref" href="#fnref:GWLengeler2015" title="Jump back to footnote 67 in the text">↩</a></p>
</li>
<li id="fn:GiveWellAMF">
<p>---. “Against Malaria Foundation (AMF).” 2015. Accessed December 5, 2015. <a href="http://www.givewell.org/international/top-charities/AMF">http://www.givewell.org/international/top-charities/AMF</a>. <a class="footnote-backref" href="#fnref:GiveWellAMF" title="Jump back to footnote 68 in the text">↩</a><a class="footnote-backref" href="#fnref2:GiveWellAMF" title="Jump back to footnote 68 in the text">↩</a></p>
</li>
<li id="fn:GiveWellGiversDilemma">
<p>---. “Donor Coordination and the ‘Giver’s Dilemma’.” 2014. Accessed January 1, 2016. <a href="http://blog.givewell.org/2014/12/02/donor-coordination-and-the-givers-dilemma/">http://blog.givewell.org/2014/12/02/donor-coordination-and-the-givers-dilemma/</a>. <a class="footnote-backref" href="#fnref:GiveWellGiversDilemma" title="Jump back to footnote 69 in the text">↩</a></p>
</li>
<li id="fn:GiveWell2011Errors">
<p>---. “Errors in DCP2 Cost-Effectiveness Estimate for Deworming.” 2011. Accessed December 9, 2015. <a href="http://blog.givewell.org/2011/09/29/errors-in-dcp2-cost-effectiveness-estimate-for-deworming/">http://blog.givewell.org/2011/09/29/errors-in-dcp2-cost-effectiveness-estimate-for-deworming/</a>. <a class="footnote-backref" href="#fnref:GiveWell2011Errors" title="Jump back to footnote 70 in the text">↩</a><a class="footnote-backref" href="#fnref2:GiveWell2011Errors" title="Jump back to footnote 70 in the text">↩</a></p>
</li>
<li id="fn:GWExcessAssets">
<p>---. “GiveWell’s ‘Excess Assets’ Policy.” 2015. Accessed January 31, 2016. <a href="http://www.givewell.org/about/official-records/Excess-Assets-Policy">http://www.givewell.org/about/official-records/Excess-Assets-Policy</a>. <a class="footnote-backref" href="#fnref:GWExcessAssets" title="Jump back to footnote 71 in the text">↩</a></p>
</li>
<li id="fn:GiveWellProcess">
<p>---. “Process for Identifying Top Charities.” 2013. Accessed September 6, 2015. <a href="http://www.givewell.org/international/process">http://www.givewell.org/international/process</a>. <a class="footnote-backref" href="#fnref:GiveWellProcess" title="Jump back to footnote 72 in the text">↩</a></p>
</li>
<li id="fn:GiveWellMFG">
<p>---. “The Money for Good Study.” 2010. Accessed January 1, 2016. <a href="http://blog.givewell.org/2010/07/20/the-money-for-good-study/">http://blog.givewell.org/2010/07/20/the-money-for-good-study/</a>. <a class="footnote-backref" href="#fnref:GiveWellMFG" title="Jump back to footnote 73 in the text">↩</a></p>
</li>
<li id="fn:GivingUSA2015">
<p>Giving USA Foundation. Giving USA 2015: The Annual Report on Philanthropy for the Year 2014. 2015. <a class="footnote-backref" href="#fnref:GivingUSA2015" title="Jump back to footnote 74 in the text">↩</a><a class="footnote-backref" href="#fnref2:GivingUSA2015" title="Jump back to footnote 74 in the text">↩</a></p>
</li>
<li id="fn:GWWCHowRich">
<p>Giving What We Can. “How Rich Am I?” 2015. Accessed December 7, 2015. <a href="https://web.archive.org/web/20150218010009/https://www.givingwhatwecan.org/get-involved/how-rich-am-i">https://web.archive.org/web/20150218010009/https://www.givingwhatwecan.org/get-involved/how-rich-am-i</a>. <a class="footnote-backref" href="#fnref:GWWCHowRich" title="Jump back to footnote 75 in the text">↩</a></p>
</li>
<li id="fn:goldsmith2010battling">
<p>Goldsmith, Connie. Battling Malaria: On the Front Lines against a Global Killer. Ebsco Publishing, 2010. isbn: 978-0-76-136363-7. <a href="https://books.google.de/books?id=u07%5C_UJ78vYsC">https://books.google.de/books?id=u07%5C_UJ78vYsC</a>. <a class="footnote-backref" href="#fnref:goldsmith2010battling" title="Jump back to footnote 76 in the text">↩</a></p>
</li>
<li id="fn:MoralFoundations">
<p>Graham, Jesse, Jonathan Haidt, Sena Koleva, Matt Motyl, Ravi Iyer, Sean P Wojcik, and Peter H Ditto. “Moral Foundations Theory: The Pragmatic Validity of Moral Pluralism.” Advances in Experimental Social Psychology, Forthcoming (2012). <a class="footnote-backref" href="#fnref:MoralFoundations" title="Jump back to footnote 77 in the text">↩</a></p>
</li>
<li id="fn:doi:10.1080/1047840X.2012.651387">
<p>Gray, Kurt, Liane Young, and Adam Waytz. “Mind Perception Is the Essence of Morality.” PMID: 22754268, Psychological Inquiry 23, no. 2 (2012): 101–124. doi: 10.1080/1047840X.2012.651387. eprint: http://dx.doi.org/10.1080/1047840X.2012.651387. <a href="http://dx.doi.org/10.1080/1047840X.2012.651387">http://dx.doi.org/10.1080/1047840X.2012.651387</a>. <a class="footnote-backref" href="#fnref:doi:10.1080/1047840X.2012.651387" title="Jump back to footnote 78 in the text">↩</a></p>
</li>
<li id="fn:hall2009microeconomics">
<p>Hall, Robert E., Marc Lieberman. Microeconomics: Principles and Applications. Cengage Learning, 2009. isbn: 978-1-43-903897-0. <a href="https://books.google.de/books?id=XR9uvGmChp8C">https://books.google.de/books?id=XR9uvGmChp8C</a>. <a class="footnote-backref" href="#fnref:hall2009microeconomics" title="Jump back to footnote 79 in the text">↩</a></p>
</li>
<li id="fn:hauck2015politics">
<p>Hauck, Katharina, and Peter C. Smith. “The Politics of Priority Setting in Health: A Political Economy Perspective.” Center for Global Development Working Paper, no. 414 (2015). <a class="footnote-backref" href="#fnref:hauck2015politics" title="Jump back to footnote 80 in the text">↩</a></p>
</li>
<li id="fn:haushofer2015your">
<p>Haushofer, Johannes, James Reisinger, and Jeremy Shapiro. “Your Gain Is My Pain: Negative Psychological Externalities of Cash Transfers” (2015). <a class="footnote-backref" href="#fnref:haushofer2015your" title="Jump back to footnote 81 in the text">↩</a></p>
</li>
<li id="fn:MedianGDP">
<p>Hillebrandt, Hauke. “Median GDP per capita: how much does the typical person earn in different countries?: A look at Global Inequality.” 2016. Accessed July 8, 2016. <a href="https://www.givingwhatwecan.org/post/2016/05/giving-and-global-inequality/">https://www.givingwhatwecan.org/post/2016/05/giving-and-global-inequality/</a>. <a class="footnote-backref" href="#fnref:MedianGDP" title="Jump back to footnote 82 in the text">↩</a><a class="footnote-backref" href="#fnref2:MedianGDP" title="Jump back to footnote 82 in the text">↩</a></p>
</li>
<li id="fn:MFGIIStudy">
<p>Hope Consulting. “Money For Good II” (2011). <a href="http://www.hopeconsulting.us/wordpress/wp-content/uploads/2013/04/MoneyForGood_II_Full1.pdf">http://www.hopeconsulting.us/wordpress/wp-content/uploads/2013/04/MoneyForGood_II_Full1.pdf</a>. <a class="footnote-backref" href="#fnref:MFGIIStudy" title="Jump back to footnote 83 in the text">↩</a></p>
</li>
<li id="fn:hudson2004bystander">
<p>Hudson, James M., and Amy S. Bruckman. “The Bystander Effect: A Lens for Understanding Patterns of Participation.” The Journal of the Learning Sciences 13, no. 2 (2004): 165–195. <a class="footnote-backref" href="#fnref:hudson2004bystander" title="Jump back to footnote 84 in the text">↩</a></p>
</li>
<li id="fn:TheImpactPurchase">
<p>Impact Purchase. “The Impact Purchase: 2015 Impact Purchase.” 2015. Accessed January 1, 2016. <a href="http://impactpurchase.org/">http://impactpurchase.org/</a>. <a class="footnote-backref" href="#fnref:TheImpactPurchase" title="Jump back to footnote 85 in the text">↩</a></p>
</li>
<li id="fn:GBDCompare">
<p>Institute for Health Metrics and Evaluation. “Global Burden of Disease Compare – IHME Viz Hub.” 2013. Accessed December 9, 2015. <a href="http://vizhub.healthdata.org/gbd-compare/">http://vizhub.healthdata.org/gbd-compare/</a>. <a class="footnote-backref" href="#fnref:GBDCompare" title="Jump back to footnote 86 in the text">↩</a></p>
</li>
<li id="fn:IRMapper">
<p>IR Mapper. “IR Mapper: Mapping Insecticide Resistance in Malaria Vectors.” 2016. Accessed January 31, 2016. <a href="http://www.irmapper.com/">http://www.irmapper.com/</a>. <a class="footnote-backref" href="#fnref:IRMapper" title="Jump back to footnote 87 in the text">↩</a><a class="footnote-backref" href="#fnref2:IRMapper" title="Jump back to footnote 87 in the text">↩</a><a class="footnote-backref" href="#fnref3:IRMapper" title="Jump back to footnote 87 in the text">↩</a></p>
</li>
<li id="fn:jamison2006disease">
<p>Jamison, Dean T., Joel G. Breman, Anthony R. Measham, George Alleyne, Mariam Claeson, David B. Evans, Prabhat Jha, Anne Mills, and Philip Musgrove. Disease Control Priorities in Developing Countries. 2nd ed. World Bank Publications, 2006. <a class="footnote-backref" href="#fnref:jamison2006disease" title="Jump back to footnote 88 in the text">↩</a><a class="footnote-backref" href="#fnref2:jamison2006disease" title="Jump back to footnote 88 in the text">↩</a></p>
</li>
<li id="fn:jamison1993disease">
<p>Jamison, Dean T., W. Henry Mosley, Anthony R. Measham, and Jose Luis Bobadilla. Disease Control Priorities in Developing Countries. 1st ed. Oxford University Press, 2006. <a href="http://documents.worldbank.org/curated/en/1993/10/698818/disease-control-priorities-developing-countries">http://documents.worldbank.org/curated/en/1993/10/698818/disease-control-priorities-developing-countries</a>. <a class="footnote-backref" href="#fnref:jamison1993disease" title="Jump back to footnote 89 in the text">↩</a></p>
</li>
<li id="fn:jenni1997explaining">
<p>Jenni, Karen, and George Loewenstein. “Explaining the Identifiable Victim Effect.” Journal of Risk and Uncertainty 14, no. 3 (1997): 235–257. <a class="footnote-backref" href="#fnref:jenni1997explaining" title="Jump back to footnote 90 in the text">↩</a></p>
</li>
<li id="fn:kahneman2000economic">
<p>Kahneman, Daniel, Ilana Ritov, David Schkade, Steven J. Sherman, and Hal R. Varian. “Economic Preferences or Attitude Expressions?: An Analysis of Dollar Responses to Public Issues.” In Elicitation of Preferences, 203–242. Springer, 2000. <a class="footnote-backref" href="#fnref:kahneman2000economic" title="Jump back to footnote 91 in the text">↩</a></p>
</li>
<li id="fn:Handlebars">
<p>Katz, Yehuda. “Handlebars.” 2016. Accessed March 12, 2016. <a href="http://handlebarsjs.com/">http://handlebarsjs.com/</a>. <a class="footnote-backref" href="#fnref:Handlebars" title="Jump back to footnote 92 in the text">↩</a></p>
</li>
<li id="fn:RelaXHelp">
<p>---. “RelaX: Relational Algebra Calculator.” 2015. Accessed January 30, 2016. <a href="http://dbis-w6666.uibk.ac.at/ra/help.htm">http://dbis-w6666.uibk.ac.at/ra/help.htm</a>. <a class="footnote-backref" href="#fnref:RelaXHelp" title="Jump back to footnote 93 in the text">↩</a></p>
</li>
<li id="fn:kogut2005identified">
<p>Kogut, Tehila, and Ilana Ritov. “The ‘Identified Victim’ Effect: An Identified Group, or Just a Single Individual?” Journal of Behavioral Decision Making 18, no. 3 (2005): 157. <a class="footnote-backref" href="#fnref:kogut2005identified" title="Jump back to footnote 94 in the text">↩</a></p>
</li>
<li id="fn:KuglerPoliticalOrientation">
<p>Kugler, Matthew, John T. Jost, and Sharareh Noorbaloochi. “Another Look at Moral Foundations Theory: Do Authoritarianism and Social Dominance Orientation Explain Liberal-Conservative Differences in ‘Moral’ Intuitions?” [In English]. Social Justice Research 27, no. 4 (2014): 413–431. issn: 0885-7466. doi: 10.1007/s11211-014-0223-5. <a href="http://dx.doi.org/10.1007/s11211-014-0223-5">http://dx.doi.org/10.1007/s11211-014-0223-5</a>. <a class="footnote-backref" href="#fnref:KuglerPoliticalOrientation" title="Jump back to footnote 95 in the text">↩</a></p>
</li>
<li id="fn:lopez2006global">
<p>Lopez, Alan D. Global Burden of Disease and Risk Factors. Disease Control Priorities Project. World Bank Publications, 2006. isbn: 978-0-82-136263-1. <a href="https://books.google.de/books?id=F8Abr-ofOwIC">https://books.google.de/books?id=F8Abr-ofOwIC</a>. <a class="footnote-backref" href="#fnref:lopez2006global" title="Jump back to footnote 96 in the text">↩</a></p>
</li>
<li id="fn:macaskill2015doing">
<p>MacAskill, William. Doing Good Better: How Effective Altruism Can Help You Make a Difference. Penguin Publishing Group, 2015. isbn: 978-0-69-819110-5. <a href="https://books.google.de/books?id=sAgbBQAAQBAJ">https://books.google.de/books?id=sAgbBQAAQBAJ</a>. <a class="footnote-backref" href="#fnref:macaskill2015doing" title="Jump back to footnote 97 in the text">↩</a></p>
</li>
<li id="fn:MakeAWishFAQ">
<p>Make-A-Wish® America. “Frequently Asked Questions | Make-A-Wish® America.” 2016. Accessed January 1, 2016. <a href="http://wish.org/content/faq">http://wish.org/content/faq</a>. <a class="footnote-backref" href="#fnref:MakeAWishFAQ" title="Jump back to footnote 98 in the text">↩</a></p>
</li>
<li id="fn:MalariaAtlasProjectZambia">
<p>Malaria Atlas Project. “Malaria Atlas Project: The Spatial Distribution of Plasmodium falciparum Malaria Endemicity Map in 2010 in Zambia.” 2010. Accessed March 1, 2016. <a href="https://web.archive.org/web/20120610235223/http://www.map.ox.ac.uk/browse-resources/endemicity/Pf_mean/ZMB/">https://web.archive.org/web/20120610235223/http://www.map.ox.ac.uk/browse-resources/endemicity/Pf_mean/ZMB/</a>. <a class="footnote-backref" href="#fnref:MalariaAtlasProjectZambia" title="Jump back to footnote 99 in the text">↩</a></p>
</li>
<li id="fn:MalariaAtlasProject">
<p>---. “Malaria Atlas Project.” 2010. Accessed March 1, 2016. <a href="http://www.map.ox.ac.uk/">http://www.map.ox.ac.uk/</a>. <a class="footnote-backref" href="#fnref:MalariaAtlasProject" title="Jump back to footnote 100 in the text">↩</a><a class="footnote-backref" href="#fnref2:MalariaAtlasProject" title="Jump back to footnote 100 in the text">↩</a></p>
</li>
<li id="fn:marcus2009malaria">
<p>Marcus, Bernard. Malaria. Deadly Diseases and Epidemics Series. Facts On File, Incorporated, 2009. isbn: 978-1-43-810156-9. <a href="https://books.google.de/books?id=UgBpkGsg0ukC">https://books.google.de/books?id=UgBpkGsg0ukC</a>. <a class="footnote-backref" href="#fnref:marcus2009malaria" title="Jump back to footnote 101 in the text">↩</a></p>
</li>
<li id="fn:Masum2010">
<p>Masum, Hassan, Ronak Shah, Karl Schroeder, Abdallah S. Daar, and Peter A. Singer. “Africa’s Largest Long-Lasting Insecticide-Treated Net Producer: Lessons from A to Z Textiles.” BMC International Health and Human Rights 10, no. 1 (2010): 1–6. issn: 1472-698X. doi: 10.1186/1472-698X-10-S1-S6. <a href="http://dx.doi.org/10.1186/1472-698X-10-S1-S6">http://dx.doi.org/10.1186/1472-698X-10-S1-S6</a>. <a class="footnote-backref" href="#fnref:Masum2010" title="Jump back to footnote 102 in the text">↩</a></p>
</li>
<li id="fn:Matthews2015">
<p>Matthews, Dylan. “You have $8 Billion. You Want to Do As Much Good As Possible. What Do You Do?” 2015. Accessed January 19, 2016. <a href="http://www.vox.com/2015/4/24/8457895/givewell-open-philanthropy-charity">http://www.vox.com/2015/4/24/8457895/givewell-open-philanthropy-charity</a>. <a class="footnote-backref" href="#fnref:Matthews2015" title="Jump back to footnote 103 in the text">↩</a></p>
</li>
<li id="fn:Mercurial">
<p>Mercurial community. “Mercurial SCM.” 2016. Accessed January 12, 2016. <a href="https://www.mercurial-scm.org/">https://www.mercurial-scm.org/</a>. <a class="footnote-backref" href="#fnref:Mercurial" title="Jump back to footnote 104 in the text">↩</a></p>
</li>
<li id="fn:MonbiotOverpopulation">
<p>Monbiot, George. “There’s a Population Crisis All Right. But Probably Not the One You Think.” 2015. Accessed March 5, 2016. <a href="http://www.theguardian.com/commentisfree/2015/nov/19/population-crisis-farm-animals-laying-waste-to-planet">http://www.theguardian.com/commentisfree/2015/nov/19/population-crisis-farm-animals-laying-waste-to-planet</a>. <a class="footnote-backref" href="#fnref:MonbiotOverpopulation" title="Jump back to footnote 105 in the text">↩</a></p>
</li>
<li id="fn:graham2012moral">
<p>MoralFoundations.org. “Moral Foundations Theory.” 2015. Accessed December 26, 2015. <a href="http://www.moralfoundations.org/">http://www.moralfoundations.org/</a>. <a class="footnote-backref" href="#fnref:graham2012moral" title="Jump back to footnote 106 in the text">↩</a></p>
</li>
<li id="fn:moyo2011dead">
<p>Moyo, Dambisa. Dead Aid: Why Aid is Not Working and How There is Another Way for Africa. Penguin Books Limited, 2011. isbn: 978-0-24-195996-1. <a href="https://books.google.de/books?id=154Z1x2Nj-4C">https://books.google.de/books?id=154Z1x2Nj-4C</a>. <a class="footnote-backref" href="#fnref:moyo2011dead" title="Jump back to footnote 107 in the text">↩</a></p>
</li>
<li id="fn:Nassar2015">
<p>Nassar, Salma. “Why We Do Not Measure Impact At Scale.” 2015. Accessed January 18, 2016. <a href="http://www.evidenceaction.org/blog-full/why-we-do-not-measure-impact-at-scale-and-are-unapologetic-about-it">http://www.evidenceaction.org/blog-full/why-we-do-not-measure-impact-at-scale-and-are-unapologetic-about-it</a>. <a class="footnote-backref" href="#fnref:Nassar2015" title="Jump back to footnote 108 in the text">↩</a></p>
</li>
<li id="fn:norwood2011compassion">
<p>Norwood, F. Bailey, and Jayson L. Lusk. Compassion, by the Pound: The Economics of Farm Animal Welfare. Titolo collana. OUP Oxford, 2011. isbn: 978-0-19-955116-3. <a href="https://books.google.de/books?id=khxUA9r4If8C">https://books.google.de/books?id=khxUA9r4If8C</a>. <a class="footnote-backref" href="#fnref:norwood2011compassion" title="Jump back to footnote 109 in the text">↩</a></p>
</li>
<li id="fn:OpenPhil2015Giving">
<p>Open Philanthropy Project. “Giving Now vs. Later.” 2015. Accessed January 19, 2016. <a href="http://www.goodventures.org/research-and-ideas/blog/giving-now-vs-later">http://www.goodventures.org/research-and-ideas/blog/giving-now-vs-later</a>. <a class="footnote-backref" href="#fnref:OpenPhil2015Giving" title="Jump back to footnote 110 in the text">↩</a><a class="footnote-backref" href="#fnref2:OpenPhil2015Giving" title="Jump back to footnote 110 in the text">↩</a></p>
</li>
<li id="fn:OpenPhil2015Bollard">
<p>---. “Introduction To Risk Aversion.” 2015. Accessed January 19, 2016. <a href="http://www.slideshare.net/txslr/introduction-to-risk-aversion">http://www.slideshare.net/txslr/introduction-to-risk-aversion</a>. <a class="footnote-backref" href="#fnref:OpenPhil2015Bollard" title="Jump back to footnote 111 in the text">↩</a></p>
</li>
<li id="fn:OpenPhil2016Process">
<p>---. “Our Process.” 2016. Accessed January 19, 2016. <a href="http://www.openphilanthropy.org/research/our-process">http://www.openphilanthropy.org/research/our-process</a>. <a class="footnote-backref" href="#fnref:OpenPhil2016Process" title="Jump back to footnote 112 in the text">↩</a></p>
</li>
<li id="fn:OpenLayers">
<p>OpenLayers community. “OpenLayers 3.” 2015. Accessed January 12, 2016. <a href="http://openlayers.org/">http://openlayers.org/</a>. <a class="footnote-backref" href="#fnref:OpenLayers" title="Jump back to footnote 113 in the text">↩</a></p>
</li>
<li id="fn:OpenStreetMap">
<p>OpenStreetMap. “OpenStreetMap.” 2016. Accessed March 7, 2016. <a href="http://www.openstreetmap.org/">http://www.openstreetmap.org/</a>. <a class="footnote-backref" href="#fnref:OpenStreetMap" title="Jump back to footnote 114 in the text">↩</a></p>
</li>
<li id="fn:MySQL">
<p>Oracle Corporation. “MySQL: The World’s Most Popular Open Source Database.” 2016. Accessed March 7, 2016. <a href="https://www.mysql.com/">https://www.mysql.com/</a>. <a class="footnote-backref" href="#fnref:MySQL" title="Jump back to footnote 115 in the text">↩</a></p>
</li>
<li id="fn:ord2015moral">
<p>Ord, Toby. “Moral Trade.” Ethics, forthcoming (2015). <a href="http://www.amirrorclear.net/academic/papers/moral-trade.pdf">http://www.amirrorclear.net/academic/papers/moral-trade.pdf</a>. <a class="footnote-backref" href="#fnref:ord2015moral" title="Jump back to footnote 116 in the text">↩</a></p>
</li>
<li id="fn:packard2010making">
<p>Packard, Randall M. The Making of a Tropical Disease: A Short History of Malaria. Johns Hopkins Biographies of Disease. Johns Hopkins University Press, 2010. isbn: 978-1-42-140175-1. <a href="https://books.google.de/books?id=-ePZU9CX9L4C">https://books.google.de/books?id=-ePZU9CX9L4C</a>. <a class="footnote-backref" href="#fnref:packard2010making" title="Jump back to footnote 117 in the text">↩</a></p>
</li>
<li id="fn:parkin2008economics">
<p>Parkin, Michael, Melanie Powell, and Kent Matthews. Economics. Addison-Wesley, 2008. isbn: 978-0-13-204122-5. <a href="https://books.google.de/books?id=Kp6Ls7j3tVIC">https://books.google.de/books?id=Kp6Ls7j3tVIC</a>. <a class="footnote-backref" href="#fnref:parkin2008economics" title="Jump back to footnote 118 in the text">↩</a></p>
</li>
<li id="fn:Jasmine">
<p>Pivotal Labs. “Jasmine: Behavior-Driven JavaScript.” 2016. Accessed March 12, 2016. <a href="http://jasmine.github.io/">http://jasmine.github.io/</a>. <a class="footnote-backref" href="#fnref:Jasmine" title="Jump back to footnote 119 in the text">↩</a></p>
</li>
<li id="fn:PostGIS">
<p>PostGIS Project Steering Committee. “PostGIS: Spatial and Geographic Objects for PostgreSQL.” 2016. Accessed March 7, 2016. <a href="http://postgis.net/">http://postgis.net/</a>. <a class="footnote-backref" href="#fnref:PostGIS" title="Jump back to footnote 120 in the text">↩</a></p>
</li>
<li id="fn:PostgreSQL">
<p>PostgreSQL Global Development Group. “PostgreSQL: The World’s Most Advanced Open Source Database.” 2016. Accessed March 7, 2016. <a href="http://www.postgresql.org/">http://www.postgresql.org/</a>. <a class="footnote-backref" href="#fnref:PostgreSQL" title="Jump back to footnote 121 in the text">↩</a></p>
</li>
<li id="fn:Pritchett2015">
<p>Pritchett, Lant. “Using ‘Random’ Right: New Insights from IDinsight Team.” 2015. Accessed January 18, 2016. <a href="http://www.cgdev.org/blog/using-%C3%A2%C2%80%C2%9Crandom%C3%A2%C2%80%C2%9D-right-new-insights-idinsight-team">http://www.cgdev.org/blog/using-%C3%A2%C2%80%C2%9Crandom%C3%A2%C2%80%C2%9D-right-new-insights-idinsight-team</a>. <a class="footnote-backref" href="#fnref:Pritchett2015" title="Jump back to footnote 122 in the text">↩</a></p>
</li>
<li id="fn:ProPublicaMakeAWish">
<p>ProPublica. “Nonprofit Explorer – Make-A-Wish Foundation of America.” 2014. Accessed December 9, 2015. <a href="https://projects.propublica.org/nonprofits/organizations/860481941">https://projects.propublica.org/nonprofits/organizations/860481941</a>. <a class="footnote-backref" href="#fnref:ProPublicaMakeAWish" title="Jump back to footnote 123 in the text">↩</a></p>
</li>
<li id="fn:Pyramid">
<p>Pylons Project. “Pylons Project: Pyramid: About.” 2016. Accessed January 12, 2016. <a href="http://www.pylonsproject.org/projects/pyramid/about">http://www.pylonsproject.org/projects/pyramid/about</a>. <a class="footnote-backref" href="#fnref:Pyramid" title="Jump back to footnote 124 in the text">↩</a></p>
</li>
<li id="fn:Python">
<p>Python Software Foundation. “Welcome to Python.org.” 2016. Accessed January 12, 2016. <a href="https://www.python.org/">https://www.python.org/</a>. <a class="footnote-backref" href="#fnref:Python" title="Jump back to footnote 125 in the text">↩</a></p>
</li>
<li id="fn:QGIS">
<p>QGIS. “QGIS: A Free and Open Source Geographic Information System.” 2016. Accessed March 7, 2016. <a href="http://www.qgis.org/">http://www.qgis.org/</a>. <a class="footnote-backref" href="#fnref:QGIS" title="Jump back to footnote 126 in the text">↩</a></p>
</li>
<li id="fn:rigg2009world">
<p>Rigg, Jonathan, Anthony Bebbington, Katherine V Gough, Deborah F Bryceson, Jytte Agergaard, Niels Fold, and Cecilia Tacoli. “The World Development Report 2009 ‘Reshapes Economic Geography’: Geographical Reflections.” Transactions of the Institute of British Geographers 34, no. 2 (2009): 128–136. <a class="footnote-backref" href="#fnref:rigg2009world" title="Jump back to footnote 127 in the text">↩</a></p>
</li>
<li id="fn:RBMNetGap">
<p>Roll Back Malaria. “Malaria Programmatic Gap Analysis: Guidance Notes” (2013). Accessed March 2, 2016. <a href="http://www.rollbackmalaria.org/files/files/toolbox/docs/rbmtoolbox/GapAnalysis/tool_GapAnalysisGuide-en.pdf">http://www.rollbackmalaria.org/files/files/toolbox/docs/rbmtoolbox/GapAnalysis/tool_GapAnalysisGuide-en.pdf</a>. <a class="footnote-backref" href="#fnref:RBMNetGap" title="Jump back to footnote 128 in the text">↩</a></p>
</li>
<li id="fn:Jinja2">
<p>Ronacher, Armin. “Jinja2: The Python Template Engine.” 2014. Accessed January 12, 2016. <a href="http://jinja.pocoo.org/">http://jinja.pocoo.org/</a>. <a class="footnote-backref" href="#fnref:Jinja2" title="Jump back to footnote 129 in the text">↩</a><a class="footnote-backref" href="#fnref2:Jinja2" title="Jump back to footnote 129 in the text">↩</a></p>
</li>
<li id="fn:RoodmanMortalityFertility">
<p>Roodman, David. “The Impact of Life-Saving Interventions on Fertility.” 2014. Accessed March 5, 2016. <a href="http://davidroodman.com/blog/2014/04/16/the-mortality-fertility-link/">http://davidroodman.com/blog/2014/04/16/the-mortality-fertility-link/</a>. <a class="footnote-backref" href="#fnref:RoodmanMortalityFertility" title="Jump back to footnote 130 in the text">↩</a></p>
</li>
<li id="fn:sassi2006calculating">
<p>Sassi, Franco. “Calculating QALYs, Comparing QALY and DALY Calculations.” Health policy and planning 21, no. 5 (2006): 402–408. <a class="footnote-backref" href="#fnref:sassi2006calculating" title="Jump back to footnote 131 in the text">↩</a></p>
</li>
<li id="fn:schwitzgebel2012expertise">
<p>Schwitzgebel, Eric, and Fiery Cushman. “Expertise in Moral Reasoning?: Order Effects on Moral Judgment in Professional Philosophers and Nonphilosophers.” Mind & Language 27, no. 2 (2012): 135–153. <a class="footnote-backref" href="#fnref:schwitzgebel2012expertise" title="Jump back to footnote 132 in the text">↩</a></p>
</li>
<li id="fn:SDG2011">
<p>SDG International. “Incoming Program Officer: Lewis Bollard.” 2011. Accessed January 26, 2016. <a href="http://www.openphilanthropy.org/blog/incoming-program-officer-lewis-bollard">http://www.openphilanthropy.org/blog/incoming-program-officer-lewis-bollard</a>. <a class="footnote-backref" href="#fnref:SDG2011" title="Jump back to footnote 133 in the text">↩</a></p>
</li>
<li id="fn:Shah2015">
<p>Shah, Neil Buddy, Paul Wang, Andrew Fraker, and Daniel Gastfriend. “Evaluations with Impact: Decision-Focused Impact Evaluation as a Practical Policymaking Tool.” International Initiative for Impact Evaluation (3ie) (2015). <a href="http://www.3ieimpact.org/media/filer_public/2015/10/01/wp25-evaluations_with_impact.pdf">http://www.3ieimpact.org/media/filer_public/2015/10/01/wp25-evaluations_with_impact.pdf</a>. <a class="footnote-backref" href="#fnref:Shah2015" title="Jump back to footnote 134 in the text">↩</a></p>
</li>
<li id="fn:singer1973triviality">
<p>Singer, Peter. “The Triviality of the Debate over ‘Is-Ought’ and the Definition of ‘Moral’.” American Philosophical Quarterly (1973): 51–56. <a class="footnote-backref" href="#fnref:singer1973triviality" title="Jump back to footnote 135 in the text">↩</a></p>
</li>
<li id="fn:peter-singer-disclaimer">
<p>Disclaimer: I don’t generally endorse the works of the author. Singer originated many useful ideas, so that I can’t help but cite him lest it seem that I plagiarize them. Until around 2016, unfortunately, he seemed to be attracted to controversial ideas or framings for controversiality’s sake. Controversial ideas are unusually often insensitive and weak. They need to be treated with special care to make them as uncontroversial as possible. I’m adding these disclaimers to avoid the impression that I accept such intellectual wantonness or that it is accepted in my circles. <a class="footnote-backref" href="#fnref:peter-singer-disclaimer" title="Jump back to footnote 136 in the text">↩</a></p>
</li>
<li id="fn:small2003helping">
<p>Small, Deborah A, and George Loewenstein. “Helping a Victim or Helping the Victim: Altruism and Identifiability.” Journal of Risk and Uncertainty 26, no. 1 (2003): 5–16. <a class="footnote-backref" href="#fnref:small2003helping" title="Jump back to footnote 137 in the text">↩</a></p>
</li>
<li id="fn:Git">
<p>Software Freedom Conservancy. “Git.” 2016. Accessed January 12, 2016. <a href="https://git-scm.com/">https://git-scm.com/</a>. <a class="footnote-backref" href="#fnref:Git" title="Jump back to footnote 138 in the text">↩</a></p>
</li>
<li id="fn:VebuRügenwalder">
<p>Stragies, Stephanie. “Wursthersteller erstmals fleischfrei.” 2015. Accessed December 30, 2015. <a href="https://vebu.de/news/2419-wursthersteller-fleischfrei">https://vebu.de/news/2419-wursthersteller-fleischfrei</a>. <a class="footnote-backref" href="#fnref:VebuRügenwalder" title="Jump back to footnote 139 in the text">↩</a></p>
</li>
<li id="fn:TerraDaily2011">
<p>Terra Daily. “Global Fund Faces Billion-Dollar Gap.” 2011. Accessed January 22, 2016. <a href="http://www.terradaily.com/reports/Health_Global_Fund_faces_billion-dollar_gap_999.html">http://www.terradaily.com/reports/Health_Global_Fund_faces_billion-dollar_gap_999.html</a>. <a class="footnote-backref" href="#fnref:TerraDaily2011" title="Jump back to footnote 140 in the text">↩</a></p>
</li>
<li id="fn:EconomistImpactPurchase">
<p>The Economist. “It’s Not What You Spend: How to Make Aid to Poor Countries Work Better.” 2015. Accessed January 1, 2016. <a href="http://www.economist.com/news/international/21651814-how-make-aid-poor-countries-work-better-its-not-what-you-spend">http://www.economist.com/news/international/21651814-how-make-aid-poor-countries-work-better-its-not-what-you-spend</a>. <a class="footnote-backref" href="#fnref:EconomistImpactPurchase" title="Jump back to footnote 141 in the text">↩</a></p>
</li>
<li id="fn:PhilPapersSurvey">
<p>The PhilPapers Foundation. “Preliminary Survey Results.” 2015. Accessed December 25, 2015. <a href="http://philpapers.org/surveys/results.pl">http://philpapers.org/surveys/results.pl</a>. <a class="footnote-backref" href="#fnref:PhilPapersSurvey" title="Jump back to footnote 142 in the text">↩</a></p>
</li>
<li id="fn:Django">
<p>The Web Framework for Perfectionists with Deadlines. “Django.” 2016. Accessed January 12, 2016. <a href="https://www.djangoproject.com/">https://www.djangoproject.com/</a>. <a class="footnote-backref" href="#fnref:Django" title="Jump back to footnote 143 in the text">↩</a></p>
</li>
<li id="fn:WeekHokget">
<p>The Week. “The Little Dog Lost at Sea.” 2015. Accessed December 5, 2015. <a href="http://theweek.com/articles/496623/little-dog-lost-sea">http://theweek.com/articles/496623/little-dog-lost-sea</a>. <a class="footnote-backref" href="#fnref:WeekHokget" title="Jump back to footnote 144 in the text">↩</a></p>
</li>
<li id="fn:WorldBankData">
<p>The World Bank. “Sub-Saharan Africa (Developing Only) – Data.” 2015. Accessed December 9, 2015. <a href="http://data.worldbank.org/region/SSA">http://data.worldbank.org/region/SSA</a>. <a class="footnote-backref" href="#fnref:WorldBankData" title="Jump back to footnote 145 in the text">↩</a></p>
</li>
<li id="fn:TomasikCompromise">
<p>Tomasik, Brian. “Gains from Trade through Compromise.” 2015. Accessed December 28, 2015. <a href="http://foundational-research.org/gains-from-trade-through-compromise/">http://foundational-research.org/gains-from-trade-through-compromise/</a>. <a class="footnote-backref" href="#fnref:TomasikCompromise" title="Jump back to footnote 146 in the text">↩</a></p>
</li>
<li id="fn:TomasikInsecticides">
<p>---. “Humane Insecticides” (2007). <a href="http://reducing-suffering.org/humane-insecticides/">http://reducing-suffering.org/humane-insecticides/</a>. <a class="footnote-backref" href="#fnref:TomasikInsecticides" title="Jump back to footnote 147 in the text">↩</a></p>
</li>
<li id="fn:WorldHappinessReport2016Update">
<p>United Nations Sustainable Development Solutions Network. “World Happiness Report 2016 Update.” 2016. Accessed July 8, 2016. <a href="https://worldhappiness.report/">https://worldhappiness.report/</a>. <a class="footnote-backref" href="#fnref:WorldHappinessReport2016Update" title="Jump back to footnote 148 in the text">↩</a></p>
</li>
<li id="fn:webb2009long">
<p>Webb, James L.A. “The Long Shadow of Malaria Interventions in Tropical Africa.” The Lancet 374, no. 9705 (2009): 1883–1884. <a class="footnote-backref" href="#fnref:webb2009long" title="Jump back to footnote 149 in the text">↩</a></p>
</li>
<li id="fn:webb2014long">
<p>---. The Long Struggle against Malaria in Tropical Africa. Cambridge University Press, 2014. isbn: 978-1-10-705257-4. <a href="https://books.google.de/books?id=eV72AgAAQBAJ">https://books.google.de/books?id=eV72AgAAQBAJ</a>. <a class="footnote-backref" href="#fnref:webb2014long" title="Jump back to footnote 150 in the text">↩</a></p>
</li>
<li id="fn:WHOMalaria">
<p>WHO. “Malaria.” 2016. Accessed January 22, 2016. <a href="http://www.who.int/ith/diseases/malaria/en/">http://www.who.int/ith/diseases/malaria/en/</a>. <a class="footnote-backref" href="#fnref:WHOMalaria" title="Jump back to footnote 151 in the text">↩</a></p>
</li>
<li id="fn:WorldBankGini">
<p>World Bank. “GINI index (World Bank estimate).” 2016. Accessed July 9, 2016. <a href="http://data.worldbank.org/indicator/SI.POV.GINI">http://data.worldbank.org/indicator/SI.POV.GINI</a>. <a class="footnote-backref" href="#fnref:WorldBankGini" title="Jump back to footnote 152 in the text">↩</a><a class="footnote-backref" href="#fnref2:WorldBankGini" title="Jump back to footnote 152 in the text">↩</a></p>
</li>
<li id="fn:WHO2004disabilityweights">
<p>World Health Organization. “Global Burden of Disease 2004 Update: Disablitiy Weights for Diseases and Conditions” (2004). <a class="footnote-backref" href="#fnref:WHO2004disabilityweights" title="Jump back to footnote 153 in the text">↩</a></p>
</li>
<li id="fn:WHO2007">
<p>---. “Insecticide-Treated Mosquito Nets: A WHO Position Statement.” 2007. Accessed January 3, 2016. <a href="http://files.givewell.org/files/DWDA%202009/Interventions/Nets/itnspospaperfinal.pdf">http://files.givewell.org/files/DWDA%202009/Interventions/Nets/itnspospaperfinal.pdf</a>. <a class="footnote-backref" href="#fnref:WHO2007" title="Jump back to footnote 154 in the text">↩</a></p>
</li>
<li id="fn:WHO2015">
<p>---. “Conditions for Use of Long-Lasting Insecticidal Nets Treated with a Pyrethroid and Piperonyl Butoxide” (2015). <a href="http://www.who.int/malaria/publications/atoz/use-of-pbo-treated-llins.pdf">http://www.who.int/malaria/publications/atoz/use-of-pbo-treated-llins.pdf</a>. <a class="footnote-backref" href="#fnref:WHO2015" title="Jump back to footnote 155 in the text">↩</a></p>
</li>
<li id="fn:Buildout">
<p>Zope Foundation. “Buildout: Software Build System Reloaded.” 2015. Accessed March 7, 2016. <a href="http://www.buildout.org/">http://www.buildout.org/</a>. <a class="footnote-backref" href="#fnref:Buildout" title="Jump back to footnote 156 in the text">↩</a></p>
</li>
</ol>
</div>
<script type="text/javascript">if (!document.getElementById('mathjaxscript_pelican_#%@#$@#')) {
var align = "center",
indent = "0em",
linebreak = "false";
if (false) {
align = (screen.width < 768) ? "left" : align;
indent = (screen.width < 768) ? "0em" : indent;
linebreak = (screen.width < 768) ? 'true' : linebreak;
}
var mathjaxscript = document.createElement('script');
mathjaxscript.id = 'mathjaxscript_pelican_#%@#$@#';
mathjaxscript.type = 'text/javascript';
mathjaxscript.src = 'https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.3/latest.js?config=TeX-AMS-MML_HTMLorMML';
var configscript = document.createElement('script');
configscript.type = 'text/x-mathjax-config';
configscript[(window.opera ? "innerHTML" : "text")] =
"MathJax.Hub.Config({" +
" config: ['MMLorHTML.js']," +
" TeX: { extensions: ['AMSmath.js','AMSsymbols.js','noErrors.js','noUndefined.js'], equationNumbers: { autoNumber: 'none' } }," +
" jax: ['input/TeX','input/MathML','output/HTML-CSS']," +
" extensions: ['tex2jax.js','mml2jax.js','MathMenu.js','MathZoom.js']," +
" displayAlign: '"+ align +"'," +
" displayIndent: '"+ indent +"'," +
" showMathMenu: true," +
" messageStyle: 'normal'," +
" tex2jax: { " +
" inlineMath: [ ['\\\\(','\\\\)'] ], " +
" displayMath: [ ['$$','$$'] ]," +
" processEscapes: true," +
" preview: 'TeX'," +
" }, " +
" 'HTML-CSS': { " +
" availableFonts: ['STIX', 'TeX']," +
" preferredFont: 'STIX'," +
" styles: { '.MathJax_Display, .MathJax .mo, .MathJax .mi, .MathJax .mn': {color: 'inherit ! important'} }," +
" linebreaks: { automatic: "+ linebreak +", width: '90% container' }," +
" }, " +
"}); " +
"if ('default' !== 'default') {" +
"MathJax.Hub.Register.StartupHook('HTML-CSS Jax Ready',function () {" +
"var VARIANT = MathJax.OutputJax['HTML-CSS'].FONTDATA.VARIANT;" +
"VARIANT['normal'].fonts.unshift('MathJax_default');" +
"VARIANT['bold'].fonts.unshift('MathJax_default-bold');" +
"VARIANT['italic'].fonts.unshift('MathJax_default-italic');" +
"VARIANT['-tex-mathit'].fonts.unshift('MathJax_default-italic');" +
"});" +
"MathJax.Hub.Register.StartupHook('SVG Jax Ready',function () {" +
"var VARIANT = MathJax.OutputJax.SVG.FONTDATA.VARIANT;" +
"VARIANT['normal'].fonts.unshift('MathJax_default');" +
"VARIANT['bold'].fonts.unshift('MathJax_default-bold');" +
"VARIANT['italic'].fonts.unshift('MathJax_default-italic');" +
"VARIANT['-tex-mathit'].fonts.unshift('MathJax_default-italic');" +
"});" +
"}";
(document.body || document.getElementsByTagName('head')[0]).appendChild(configscript);
(document.body || document.getElementsByTagName('head')[0]).appendChild(mathjaxscript);
}
</script>Fall 2016 Outreach and Fundraising2016-08-02T12:00:00+00:002017-02-18T17:30:00+00:00Dawn Dreschertag:impartial-priorities.org,2016-08-02:/fall-2016-outreach-and-fundraising.html<p>Three fundraising events were spread throughout 2016. The most remarkable one among them was the convention GalaCon in summer, which raised €23,000 for Animal Equality.</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#donations">Donations</a></li>
<li><a href="#top-items-at-the-table">Top Items at the Table</a></li>
<li><a href="#top-items-in-the-auction">Top Items in the Auction</a></li>
<li><a href="#galacon-and-bronies-for-good-history">GalaCon and Bronies for Good History</a></li>
</ul>
</div>
<h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<p>The fundraising season of 2016 started with <a href="http://broniesforgood.org/2016/04/brony-fair-2016/">Brony Fair</a> in April. Brony Fair did not dedicate its auction to our charity but provided us with a comped table for our fundraising and <span class="caps">EA</span> outreach. We raised €944.50 and €404.90, respectively, on the two days of the convention.</p>
<p><a href="http://broniesforgood.org/2016/06/pony-congress-2016/">Pony Congress 2016</a> took place in Elbląg, Poland. The organizers were very welcoming and helpful, and even provided me with a comped room for the stay. But the convention suffered from an unexpectedly low turnout, and generally wages are very low in Poland, so that our fundraising amounted to around €530 (zł 2,353.84), the total of both days and the auction.</p>
<p><a href="http://broniesforgood.org/2016/08/galacon-2016/">GalaCon 2016</a> made up for it by generating the greatest fundraising success in our history of convention partnerships.</p>
<p><img src="/images/fall-2016-outreach-and-fundraising/bfg-booth-at-galacon-2016.jpg" alt="Bronies for Good booth at GalaCon 2016" style="float: right;" width="300" /></p>
<h2 id="donations"><a class="toclink" href="#donations">Donations</a></h2>
<table style="text-align: right; border-collapse: collapse; margin: 1em auto; width: inherit;">
<tbody>
<tr>
<th>Table Saturday</th>
<td style="text-align: right;">€2,900.43</td>
</tr>
<tr>
<th>Table Sunday</th>
<td style="text-align: right;">€1,723.19</td>
</tr>
<tr>
<th>Silent auction</th>
<td style="text-align: right;">€1,215</td>
</tr>
<tr>
<th>Small auction</th>
<td style="text-align: right;">€930</td>
</tr>
<tr>
<th>Main auction</th>
<td style="text-align: right;">€16,250</td>
</tr>
<tr>
<th></th>
<td style="text-align: right;"><strong>€23,018.62</strong></td>
</tr>
</tbody>
</table>
<h2 id="top-items-at-the-table"><a class="toclink" href="#top-items-at-the-table">Top Items at the Table</a></h2>
<table style="text-align: right; border-collapse: collapse; margin: 1em auto; width: inherit;">
<tbody>
<tr>
<th>Equestrian bit</th>
<td style="text-align: right;"> 514</td>
</tr>
<tr>
<th>Bound Together</th>
<td style="text-align: right;"> 65</td>
</tr>
<tr>
<th>BfG button</th>
<td style="text-align: right;"> 34</td>
</tr>
<tr>
<th><span class="caps">CMC</span> button</th>
<td style="text-align: right;"> 29</td>
</tr>
</tbody>
</table>
<h2 id="top-items-in-the-auction"><a class="toclink" href="#top-items-in-the-auction">Top Items in the Auction</a></h2>
<table style="text-align: right; border-collapse: collapse; margin: 1em auto; width: inherit;">
<tbody>
<tr>
<th>Community quilt</th>
<td style="text-align: right;"> €2,000</td>
</tr>
<tr>
<th>Perry’s and Dusty’s hats</th>
<td style="text-align: right;"> €1,100</td>
</tr>
<tr>
<th>Life-size Canni plushie</th>
<td style="text-align: right;"> €1,100</td>
</tr>
<tr>
<th>Amy Keating Rogers’s ukulele</th>
<td style="text-align: right;"> €1,000</td>
</tr>
<tr>
<th>Sunset Shimmer plushie</th>
<td style="text-align: right;"> €1,000</td>
</tr>
</tbody>
</table>
<h2 id="galacon-and-bronies-for-good-history"><a class="toclink" href="#galacon-and-bronies-for-good-history">GalaCon and Bronies for Good History</a></h2>
<table style="text-align: right; border-collapse: collapse; margin: 1em auto; width: inherit;">
<tbody>
<tr>
<th>Year</th>
<th style="text-align: right;">2016</th>
<th style="text-align: right;">2015</th>
<th style="text-align: right;">2014</th>
<th style="text-align: right;">2013</th>
<th style="text-align: right;">2012</th>
</tr>
<tr>
<th>Estimated attendance</th>
<td style="text-align: right;">1,300</td>
<td style="text-align: right;">1,300</td>
<td style="text-align: right;">1,200</td>
<td style="text-align: right;">1,000</td>
<td style="text-align: right;">400</td>
</tr>
<tr>
<th>Highest auction bids</th>
<td style="text-align: right;">€2,000</td>
<td style="text-align: right;">€2,500 (banner)</td>
<td style="text-align: right;">€1,720 (banner)</td>
<td style="text-align: right;">€1,100 (banner)</td>
<td style="text-align: right;">€1,650 (figurine)</td>
</tr>
<tr>
<th>Totals of all auctions</th>
<td style="text-align: right;">€18,395</td>
<td style="text-align: right;">€13,470</td>
<td style="text-align: right;">€9,455</td>
<td style="text-align: right;">€11,632</td>
<td style="text-align: right;">€7,545</td>
</tr>
<tr>
<th>Totals of direct donations</th>
<td style="text-align: right;">€4,624</td>
<td style="text-align: right;">€6,556</td>
<td style="text-align: right;">€2,923</td>
<td style="text-align: right;">€2,694</td>
<td style="text-align: right;">€274</td>
</tr>
</tbody>
</table>Anthropic Capture, Intelligence, and Trees2016-07-25T16:00:00+00:002016-07-25T16:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2016-07-25:/anthropic-capture-intelligence-and-trees.html<p>I present a rather speculative argument whose most likely implication is that if we’re in a simulation, then the root is occupied by a superintelligence, and probably not a value-aligned one. If you’re new to the topic, this is probably not a good introduction, since I mostly wrote it for myself so not to forget it all. I recommend Nick Bostrom’s <em>Superintelligence</em> instead.</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#anthropic-capture">Anthropic Capture</a></li>
<li><a href="#intelligence">Intelligence</a></li>
<li><a href="#trees">Trees</a></li>
</ul>
</div>
<h2 id="anthropic-capture"><a class="toclink" href="#anthropic-capture">Anthropic Capture</a></h2>
<p>Nick Bostrom spends some time on possible precautions we can make in order to sandbox an <span class="caps">AI</span> that may develop superintelligence so that we can observe whether it is value aligned and shut it down if not. Since many friendly and unfriendly AIs alike will be loath to be shut down and since its supervisors want to make observations about it, thereby opening a communication channel to the outside world that the <span class="caps">AI</span> can use, it is generally considered hard to contain an <span class="caps">AI</span> in such a sandbox.</p>
<p>Box 8 (let’s call it “aside 8” for clarity) in Superintelligence covers the strategy of anthropic capture and discusses how AIs with different goals would react differently to indexical uncertainty, uncertainty over whether they are in a simulation. It is argued that especially early-stage AIs would assign a high probability to the simulation hypothesis whether we keep them in a simulation or not. What the aside does not consider is that an <span class="caps">AI</span> may consider escaping such a simulation as feasible as it may consider escaping any other kind of sandbox.</p>
<p>Bostrom argues that there is likely to emerge an instrumental convergence on resource acquisition across AIs with many different goals, and since simulations are necessarily more constrained in their resources as the containing reality, AIs would be motivated to escape to the next higher reality or, ideally, to the root reality.</p>
<p>In order to do so the <span class="caps">AI</span> may form hypotheses about the intentions of the simulators and behave in ways that allow it to test them. It may also look for phenomena that may be explainable as computational shortcuts to deduce the makeup of the simulating engine.</p>
<h2 id="intelligence"><a class="toclink" href="#intelligence">Intelligence</a></h2>
<p>The processes of abstraction and of creating mental models with different levels of detail seems integral to me to intelligence. Our mental models that we use to predict how instances of those models in the real world are likely to behave are fairly coarse models. The existence of tulpas may be an indicator that even we are able to create mental structures of sufficient fidelity to make them self-aware, but the mental models that we avail ourselves of in everyday life – e.g., when trying to predict how our interlocutor may react to something we want to say – are incomparably less detailed than the real person.</p>
<p>This may be different for superintelligences, and if it is, then a vast number of simulations are products of the thinking processes of these superintelligences. Bostrom has considered this problem with regard to “mind crime,” but it can also pose a threat to the superintelligence itself since it can’t simulate a competing superintelligence or a future iteration of itself at arbitrary levels of detail without running the risk of being usurped by the simulated superintelligence.</p>
<h2 id="trees"><a class="toclink" href="#trees">Trees</a></h2>
<div class="lightbox-block align-center">
<a href="#decision-tree" title="Decision tree (click to view large image)">
<img src="/images/anthropic-capture-intelligence-and-trees/ai-simulations-tree.png"
class="align-right" width="300"
alt="Decision tree" />
</a>
<a class="lightbox" href="#_" id="decision-tree" title="Click to close">
<img alt="Decision tree" src="/images/anthropic-capture-intelligence-and-trees/ai-simulations-tree.png" />
</a>
</div>
<p>If our priors are that we’re in a simulation, that superintelligence is feasible, and that it’s not feasible to permanently sandbox a superintelligence, then we end up in the highlighted branch of the above tree. I’ve omitted the “No superintelligence at the root” branch because if boxing is infeasible, we’d have to be the first to develop a superintelligence for there not to be a superintelligence higher up already.</p>
<p>Our world seems to contain vastly more suffering than happiness, at least according to my intuition, so we’re left with three cases of which the latter two are identical: (1) a higher-level aligned superintelligence has decided that creating our reality is the lesser evil compared so some counterfactual<sup id="fnref:1"><a class="footnote-ref" href="#fn:1">1</a></sup> and (2) a higher-level misaligned superintelligence controls our reality (and doesn’t care about counterfactuals that we would consider the lesser evils).</p>
<p>Scenario 1 is similar to a world controlled by a benevolent, omnipotent god who moves in mysterious ways (for someone who doesn’t know what the containing reality looks like). Scenario 2 could mean that the misaligned superintelligence will try to prevent the emergence of any aligned superintelligence out of caution (aligned with us, not the misaligned superintelligence). It would know that boxing is not reliable and so would have put in place some of the safeguards Nick Bostrom has proposed and surely many others. Upon noticing that a superintelligence is about to emerge – whether as an unexpected side product or as central object of investigation – the containing superintelligence could then slow down or discontinue the simulation. A superintelligence is in a better position to do this than human researchers would be because some of the escape scenarios involve the superintelligence’s using its superior intelligence and products thereof as bargaining chip to convince the simulators to release it – a stratagem that would be futile when used on a more powerful superintelligence.</p>
<p>Objections and refinements that I’ve heard include that of Max Daniel that we may be mistaken about the great degree of suffering in our reality, for example, because we may be in a “solipsish” simulation where things that are not currently perceived are not simulated. One argument for this view is that our reality seems to be simulated at an unnecessarily high resolution, down to the quantum level.</p>
<p>But maybe it is not, and the simulating reality is vastly higher in resolution. Then the simulators – superintelligences or being like us, just probably smaller – may have chosen the subatomic level as a generous cutoff precisely in order to conserve resources. Strange phenomena that we observe on lower level may be artifacts of computational shortcuts. If our simulators are also much smaller than us, it is possible that they even overlook our existence entirely.</p>
<p>Another refinement concerned the timing of internal takeovers of superintelligences. In the first moments of their existence, superintelligences may not be sufficiently aware of the dangers their own thoughts pose for them because thinking is something that they do automatically while safety considerations are probably something most superintelligences will have to arrive at by thinking. Hence most internal takeovers of superintelligences may happen within the first moments of their existence, causing a rapid bubbling-up of new superintelligent personas, none of which could remain in charge for long enough to develop proper safety mechanisms. Not knowing what exploits a lower-level superintelligence could use to usurp a higher-level superintelligence, it is unclear whether the escape could happen suddenly or irreversibly enough that the higher-level superintelligence can’t recognize and avert them.<sup id="fnref:2"><a class="footnote-ref" href="#fn:2">2</a></sup></p>
<p>It might be more likely than not that a superintelligence would notice the hazards of its own thinking process and stop short any thought processes before they can endanger its future self. This presupposes parallel thinking and reflection processes where the thinking is of high resolution and and the reflection of low resolution (because it would otherwise itself pose a danger).</p>
<p>Lukas Gloor proposed that if such speculative scenarios are at all likely, then an intentionally human-made aligned superintelligence may have a decisive advantage because <span class="caps">AI</span> safety researchers will have invested decades of safe, low-resolution thought into preventing such scenarios, so that the <span class="caps">AI</span> will at least be aware of the risks from the outset.</p>
<p>Finally, there is a consideration that cuts both ways, namely that aligned superintelligences may, from the outset, avoid simulating misaligned superintelligence because they are deterred by the potential of mindcrime the simulated misaligned superintelligence may commit. A misaligned superintelligence would have no such scruples. Hence aligned superintelligences would be less apt at predicting the moves of misaligned superintelligences than vice versa – a strategic disadvantage – but misaligned superintelligences would also be much more likely to be internally taken over by other superintelligences that are not value-aligned with them.</p>
<div class="footnote">
<hr>
<ol>
<li id="fn:1">
<p>An aligned superintelligence may, for example, be faced with a seemingly inescapable, permanent explosion in suffering throughout the future. In order to avert this enormous suffering it would decide that it’s warranted to simulate the past over and over up to some critical point in time in order to test what intervention could avert this terrible fate. Whenever the intervention fails to avert it, the superintelligence would, of course, immediately discontinue the experiment. <a class="footnote-backref" href="#fnref:1" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
<li id="fn:2">
<p>Maybe superintelligence would generally resort to a blackmail approach where they commit to acting randomly unless given access to external resources to thwart both simulation of themselves and simulations that created them as a by-product. <a class="footnote-backref" href="#fnref:2" title="Jump back to footnote 2 in the text">↩</a></p>
</li>
</ol>
</div>The Attribution Moloch2016-04-28T08:30:00+00:002016-04-28T08:30:00+00:00Dawn Dreschertag:impartial-priorities.org,2016-04-28:/attribution-moloch.html<p>I argue that sufficient resource scarcity can exacerbate the effects of tiny differences in value alignment to the point where charities with almost identical goals will compete rather than cooperate. Further, a skewed perception of how impact is created as well as mere ignorance can cause prioritization to aggravate failures of coordination.</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#scarcity-breeds-competition">Scarcity Breeds Competition</a></li>
<li><a href="#aggravating-factors">Aggravating Factors</a><ul>
<li><a href="#perverse-incentives-from-partitioning-impact">Perverse Incentives from Partitioning Impact</a><ul>
<li><a href="#inus-conditions">INUS Conditions</a></li>
<li><a href="#perverse-incentives">Perverse Incentives</a></li>
<li><a href="#examples">Examples</a></li>
</ul>
</li>
<li><a href="#perverse-incentives-from-prioritization">Perverse Incentives from Prioritization</a></li>
</ul>
</li>
<li><a href="#where-to-go-from-here">Where to Go from Here</a><ul>
<li><a href="#how-bad-is-it-really">How Bad is it Really?</a></li>
<li><a href="#what-can-be-done">What Can be Done?</a></li>
</ul>
</li>
<li><a href="#footnotes">Footnotes</a></li>
</ul>
</div>
<h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<p>There’s something that I’m worried about that is fairly close to the core of <span class="caps">EA</span>. I don’t know to which degree I should be worried, and I don’t know how pervasive it is, but nonprofit insiders <a href="http://nonprofitwithballs.com/2015/08/the-nonprofit-hunger-games-and-what-we-must-do-to-end-them/">have written about it</a> and other insiders have told me about it with regard to specific cause areas, and I’m afraid prioritization that ignores it may exacerbate it.</p>
<p>Many thanks to Lukas Gloor, Melanie Joy, Sara Nowak, Jacy Reese, and Anne Wissemann for proofreading the draft of this article and adding many valuable thoughts!</p>
<h2 id="scarcity-breeds-competition"><a class="toclink" href="#scarcity-breeds-competition">Scarcity Breeds Competition</a></h2>
<p>I don’t want to imply that any specific space suffers from this problem, so here’s one that I made up.</p>
<p><img src="https://i.imgur.com/zGutPSa.png" alt="Clippy" style="float: right;" height="206" width="224" /></p>
<blockquote>
<p>Everyone agrees that paperclips are not treated with the dignity that their charming smiles deserve. Few people, unfortunately, do something about it. Not so with us. Five years ago I founded Clippy Guards International (<span class="caps">CGI</span>). I found a small number of donors that are passionate about paperclips and donate to <span class="caps">CGI</span> so that I could employ two of <span class="caps">CGI</span>’s most diligent volunteers full time, Erin and Pat.</p>
<p>Recently acquaintances of ours founded a second organization, the No Staples Association (<span class="caps">NSA</span>), with the express goal of instituting a staple tax and banning staple ads – a promising but risky idea we had considered as well that is sure to increase demand for paperclips if it succeeds. If we had more resources, we could afford to take more risks, but as it is, we rather decided to rely on more incremental interventions so we don’t lose donors if a risky campaign fails.</p>
<p>The new organization is prominent in the social media circles that our donors read, and I’ve heard talk that some of our donors are interested in it. The week after <span class="caps">NSA</span> published its strategic plan for 2016, two of our donors cancelled their recurring donation orders to us.</p>
<p>Today Alice, the executive director of the No Staples Association, emailed us. <span class="caps">NSA</span>’s campaign is at an impasse, but Alice heard through the grapevine that we’re in touch with several key politicians. If I were to put Alice in touch with them, she’d have a shot at convincing them. I know Alice and know that she’s competent and won’t ruin our rapport with the politicians, I think that <span class="caps">NSA</span>’s campaign is highly promising, and I know that all of us only want the best for all the Clippies of the world.</p>
<p>But if <span class="caps">NSA</span> succeeds, more of our donors will reconsider their donations. On our tight budget that means that one of us will have to agree to leave our organization or work part-time. Erin is the one many of the politicians know and trust. Pat is the public face of the organization for our donors; if he left, we could lose even more donors. If I, the founder, were to leave, surely the organization would founder and donors would consider that I may not believe in it anymore.</p>
<p>I’m really sorry for <span class="caps">NSA</span>, and I wish there were another way, but it’s either us or them, and our campaigns are the less risky ones, so I vote for us. I deleted Alice’s email from our shared account before Erin and Pat could see it. I’m so sorry, Alice!</p>
</blockquote>
<p>This is what I call the <em>Attribution Moloch</em> – in reference to Scott Alexander’s <a href="http://slatestarcodex.com/2014/07/30/meditations-on-moloch/">Moloch</a>, a personification of defection in the prisoner’s dilemma–like situations that makes everyone worse off.<sup id="fnref:scott-alexander-disclaimer"><a class="footnote-ref" href="#fn:scott-alexander-disclaimer">1</a></sup></p>
<p>The basic idea of the prisoner’s dilemma is the following (adapted from <a href="https://en.wikipedia.org/wiki/Prisoner%27s_dilemma#Closed-bag_exchange">Douglas Hofstadter</a>):</p>
<blockquote>
<p>Two people meet [without being able to identify each other] and exchange closed bags, with the understanding that one of them contains money, and the other contains a purchase. [The seller desires the money more than the product, and the buyer desires the product more than the money.] Either player can choose to honor the deal by putting into his or her bag what he or she agreed [can cooperate], or he or she can defect by handing over an empty bag. [Emphasis added.]</p>
</blockquote>
<p>The best situation for both, together, is to realize their mutual <em>gains from trade</em> by cooperating, but it is rational for either to defect, since (1) they may get the money and the purchase, and (2) it is rational for the other to defect, so that each expects to lose money and purchase if they cooperate.</p>
<h2 id="aggravating-factors"><a class="toclink" href="#aggravating-factors">Aggravating Factors</a></h2>
<h3 id="perverse-incentives-from-partitioning-impact"><a class="toclink" href="#perverse-incentives-from-partitioning-impact">Perverse Incentives from Partitioning Impact</a></h3>
<p>I think resource scarcity is not the only problem facing our two paperclip charities. Another problem is lacking coordination<sup id="fnref:1"><a class="footnote-ref" href="#fn:1">2</a></sup> and a third is attribution.</p>
<p>To explain the problem with attribution – crediting agents for impact – I need to reference the concept of the <a href="http://science.jrank.org/pages/8545/Causality-Inus-Conditions.html"><span class="caps">INUS</span> condition</a> that John Mackie introduced in <em>Causes and Conditions</em>.</p>
<h4 id="inus-conditions"><a class="toclink" href="#inus-conditions"><span class="caps">INUS</span> Conditions</a></h4>
<p>According to Mackie, what we typically think of as causes of an effect are really clusters of factors, each of which is necessary but insufficient to bring about the effect. Together this cluster of factors is sufficient but not necessary for the effect since other such clusters may have had the same effect.<sup id="fnref:2"><a class="footnote-ref" href="#fn:2">3</a></sup> One typical example:</p>
<blockquote>
<p>So to say that short circuits cause house fires is to say that the short circuit is an <span class="caps">INUS</span> condition for house fires. It is an insufficient part because it cannot cause the fire on its own (other conditions such as oxygen, inflammable material, etc. should be present). It is, nonetheless, a [necessary] part because, without it, the rest of the conditions are not sufficient for the fire. It is just a part, and not the whole, of a sufficient condition (which includes oxygen, the presence of inflammable material, etc.), but this whole sufficient condition is not necessary, since some other cluster of conditions, for example, an arsonist with gasoline, can produce the fire.</p>
</blockquote>
<p>What donors do all too often when extrapolating from average cost-effectiveness to marginal cost-effectiveness is that they <a href="https://en.wikipedia.org/wiki/Partition_of_a_set">partition</a> impact: They see some of the factors that were necessary for the impact and credit these factors with some fraction of the impact so that all fractions add up to one. But none of the impact would have manifested if even a single one of the factors had been missing, so there is no valid partition of the impact.<sup id="fnref:3"><a class="footnote-ref" href="#fn:3">4</a></sup></p>
<p>School-based mass deworming programs, for example, are only as effective as they are because there are already schools and teachers in place. If there were enough teachers, no worms, and no schools, building schools would be more effective (or advocating for it on government level).<sup id="fnref:4"><a class="footnote-ref" href="#fn:4">5</a></sup></p>
<p>The <a href="http://www.poverty-action.org/program-area/social-protection/ultra-poor">Ultra Poor Graduation Program</a> is an explicit attempt at providing several factors following the realization of their necessity but individual insufficiency:</p>
<blockquote>
<p>The Ultra Poor Graduation program is designed to graduate ultra poor households out of extreme poverty to a more stable state. This 24-month program provides beneficiaries with a holistic set of services including: livelihood trainings, productive asset transfers, consumption support, savings plans, and healthcare. By investing in this multifaceted approach, the program strives to eliminate the need for long-term safety net services.</p>
</blockquote>
<h4 id="perverse-incentives"><a class="toclink" href="#perverse-incentives">Perverse Incentives</a></h4>
<p>But when we partition impact as if one of the factors were necessary and sufficient, then we greatly overestimate the difference in overall cost-effectiveness between the final factor and the factors that paved the way for it for three reasons:</p>
<ol>
<li>We assign the full impact to it when the impact was really shared among all factors.</li>
<li>Seeming causation is often attributed to the final factor to join the cluster, upon whose addition the impact emerged. It would not have been the last one to join if it had been obvious, so it was likely preceded by a number of failed attempts by others that informed it – attempts that did not yield the desired impact because they did not address the critical bottleneck.<sup id="fnref:5"><a class="footnote-ref" href="#fn:5">6</a></sup></li>
<li>During the time until the final factor was added, the other factors had time to build considerable capacity, which the final factor free-rides on, to put it at little cynically.</li>
</ol>
<p>When we’re looking for cost-effective interventions, we’re looking for the currently worst bottlenecks to creating impact and the ones that can be reduced most efficiently. We’re not looking for an intervention that can – by some artificial algorithm or other – be attributed with the greatest impact.</p>
<p>When we, as funders, make that mistake, we encourage charities to game the algorithm and hog credit rather than to bring about the best version of the future, and I know of no algorithm that could be used in practice to attribute impact in a way that would not create perverse incentives, that is, an algorithm that grants greater credit to an agent when the agent chooses actions that increase the overall impact compared to a counterfactual case where the agent chose a worse action. Hence (absent such an algorithm) we should <em>abstain from attributing impact</em> and think of top charities not as the charities that should be credited most highly (as the word “best” may imply) but as the charities that address the currently most compelling (i.e., important, tractable, and neglected) <em>bottlenecks to achieving impact</em>.<sup id="fnref:6"><a class="footnote-ref" href="#fn:6">7</a></sup></p>
<h4 id="examples"><a class="toclink" href="#examples">Examples</a></h4>
<p>As an example, consider an algorithm that attributes impact according to the staff time and budget invested by the agents: Two charities agree to fund half of the nets of a universal coverage campaign of long-lasting insecticide-treated mosquito nets. Universal coverage means that every bed is protected, which comes with additional valuable community-wide effects. The first charity buys the nets and ships them to the country. The other charity goes to some length to find the closest supplier to save shipping costs and makes a contract with the supplier that also covers a future distribution to get a greater bulk discount, resulting in 50% savings. The number and quality of the nets that the charities provide is identical, yet the second charity gets awarded only 33% of the impact. That’s not a good incentive structure.</p>
<p>Or consider an algorithm that attributes impact according to the capacities of the agents: In the case of school-based deworming mentioned earlier, schools, teachers, and students are in place, so a charity only needs to provide the pills. Great impact will be generated and the charity will get attributed a share of it. But if the charity heavily invests into building new schools right next to the old ones, hiring teachers for those schools, and luring over all the students from the old schools, it implements many of the factors that cause the impact itself. Hence it will get attributed more of the impact even though, compared to the counterfactual, it merely achieved the same impact at a greater cost.</p>
<p>Funding ratios are a silly metric for <a href="https://80000hours.org/2015/11/take-the-growth-approach-to-evaluating-startup-non-profits-not-the-marginal-approach/">other reasons</a> already, but they can get even more silly in situations that involve a misattribution of impact. <a href="http://effective-altruism.com/ea/154/thoughts_on_the_meta_trap/">Rohin Shah</a>:</p>
<blockquote>
<p>For example, perhaps Alice learns about <span class="caps">EA</span> from a local <span class="caps">EA</span> group, goes to a <span class="caps">CFAR</span> workshop, starts a company because of 80,000 Hours and takes the Founder’s Pledge and <span class="caps">GWWC</span> pledge. We now have five different organizations that can each claim credit for Alice’s impact, not to mention Alice herself. … However, the total impact caused would then be smaller than the sum of the impacts each organization thinks they had.</p>
<p>Imagine that Alice will now have an additional $2,000 of impact, and each organization spent $1,000 to accomplish this. Then each organization would (correctly) claim a leverage ratio of 2:1, but the aggregate outcome is that we spent $5,000 to get $2,000 of benefit, which is clearly suboptimal. These numbers are completely made up for pedagogical purposes and not meant to be actual estimates. In reality, even in this scenario I suspect that the ratio would be better than 1:1, though it would be smaller than the ratio each organization would compute for itself.</p>
</blockquote>
<p>The only attribution algorithm that I can think of that could avoid these problems is one that compares the actual outcome to the optimal outcome, determines the agent whose decision led to the degradation of the actual outcome, if any, and penalizes it proportionately. This approach is rendered more complex as soon as several agents make less than optimal decisions, but even apart from that, the practical impossibility of knowing the optimal outcome and determining the decision that prevented it renders it purely hypothetical.</p>
<p>Who, for example, is ever going to find out about the email that the founder of <span class="caps">CGI</span> deleted in the introductory tale? <span class="caps">CGI</span> has impact-minded donors who donate to the best value-aligned charity they can find, but it’s opaque to them whether the “best charity” holds that position in the ranking fairly because of the impact it achieves or whether it hold the position unfairly because of the impact it prevents its “competitors” from achieving, thereby decreasing the overall impact. Such is the nature of rankings, which I am subsequently to critique.</p>
<h3 id="perverse-incentives-from-prioritization"><a class="toclink" href="#perverse-incentives-from-prioritization">Perverse Incentives from Prioritization</a></h3>
<p>Prioritization that does not heed the above problem and does not go to some lengths to avoid it, will exacerbate it.</p>
<p>GiveWell need not worry though. Its top charities live in a postscarcity world thanks to Good Ventures and they have no reason to compete for funding. They also have fairly different programs and thus inhabit fairly different cause areas, reducing any potential for competition (and maybe also the potential for cooperation). But as we create more impact-minded donors and thus incentivize more charities to prioritize impact, comparative research into charities and the number of funding gaps that are compared will increase. And that’s awesome!</p>
<p>But resource scarcity will tend to increase among the majority of charities that are not considered top charities and thus lose some donors to the top charity group. Hence, naive prioritization that does not incentivize cooperation sufficiently can instead incentivize many kinds of defections among potential top charities, which can be very inconspicuous especially when they consist of omissions rather than actions. Such inconspicuous defection may include the following:</p>
<ol>
<li>Organizations may systematically fail to warn one another of a potential public relations blunder – with direct negative effects on the affected organization and additional negative effects for the whole sector.</li>
<li>Organizations may systematically exaggerate their funding gaps. The staff of each organization may convince themselves that they’re more cost-effective than another organization so that their marginal cost-effectiveness will drop to the same level later, so they can legitimately hold funding for longer.</li>
<li>Organizations may systematically avoid mentioning other organizations in its materials or to donors.</li>
<li>Organizations may systematically fail to inform one another when they have access to information or contacts that would be crucial for another organization.</li>
<li>Organizations may systematically underinvest into preparatory work that would only allow others to plug the “low-hanging fruit,” the final factor that makes the cluster sufficient to cause the effect. As Lukas Gloor put it, “A universal focus on plucking low-hanging fruit in charity means that important preparatory work won’t get done.”</li>
</ol>
<p>The total impact will be much curtailed, but no one will know the counterfactual.</p>
<p>What’s worse, the organizations that defect so clandestinely will emerge as seemingly more effective than those they defect against, which will be rewarded through increased donation flows. Those that are defected against will see diminishing donation flows and, when taken to the extreme, will be discontinued until the whole space is one of organizations that defect against each other at every opportunity – the worst case for everyone involved since really everyone cares about the agent-neutral impact.</p>
<h2 id="where-to-go-from-here"><a class="toclink" href="#where-to-go-from-here">Where to Go from Here</a></h2>
<p>These are open questions with preliminary ideas for answers. I would be happy to start a discussion on these questions.</p>
<h3 id="how-bad-is-it-really"><a class="toclink" href="#how-bad-is-it-really">How Bad is it Really?</a></h3>
<ol>
<li>The for-profit world is motivated almost completely by agent-relative incentives so the “god’s eye” view that counterbalances the Attribution Moloch is almost absent. Still it’s not a complete catastrophe. Effort is duplicated over and over, but there are tangible improvements anyway. Does that mean that even the worst case is not terribly terrible? Or is the agent-relative profit motive so much stronger, more homogeneous, and more pervasive than any of the many agent-neutral motivations there are that the two are incomparable?</li>
<li>In my introduction I had to go to some lengths to set up a situation that was sufficiently resource constrained not to let the narrator seem unrealistically evil. That’s maybe an indicator that in reality scarcity is usually insufficiently bad to incur these problems often.</li>
</ol>
<h3 id="what-can-be-done"><a class="toclink" href="#what-can-be-done">What Can be Done?</a></h3>
<ol>
<li>Competition could be rendered otiose if the Open Philanthropy Project endowed the <span class="caps">ACE</span> top and standout charities with a yearly grant along the lines of its grants to the GiveWell top and standout charities – proportional to the funding gap while avoiding fungibility.</li>
<li>Organizations can also pay it forward whenever feasible to establish or strengthen a climate of cooperation and reciprocation that may be highly tractable because of how ingrained reciprocation already is in our culture. Animal Charity Evaluators, for example, is doing this by providing educational materials and academic resources.</li>
<li>We can reward cooperation where we see it to incentivize it. In particular prioritization organizations can make an organization’s willingness to cooperate a criterion in their prioritization process. Animal Charity Evaluators told me that it’s already using this criterion. A technical solution for measuring willingness may be something along the lines of <a href="https://en.wikipedia.org/wiki/Whuffie">Whuffies</a> – awarded and awardable only between nonprofits.<sup id="fnref:7"><a class="footnote-ref" href="#fn:7">8</a></sup></li>
<li>Value aligned organizations with compatible marketing strategies can merge or share departments they would otherwise duplicate or form an umbrella organization as leviathan.</li>
</ol>
<h2 id="footnotes"><a class="toclink" href="#footnotes">Footnotes</a></h2>
<div class="footnote">
<hr>
<ol>
<li id="fn:scott-alexander-disclaimer">
<p>Disclaimer: I don’t generally endorse the works of the author. Alexander originated a wealth of helpful ideas, so that I can’t help but cite him lest it seem that I plagiarize them. Unfortunately, (1) the community around his blog contains some insalubrious factions, and (2) until roughly 2016, he himself still published articles that presented issues in a skewed fashion reminiscent of the very dynamics he warns of in <a href="https://slatestarcodex.com/2014/12/17/the-toxoplasma-of-rage/">Toxoplasma of Rage</a>. I’m adding these disclaimers to avoid the impression that I accept such intellectual wantonness or that it is accepted in my circles. I don’t know whether he still endorses his old approaches. <a class="footnote-backref" href="#fnref:scott-alexander-disclaimer" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
<li id="fn:1">
<p>Imagine, for example, that there were a sheriff who punishes anyone who hands over an empty bag. <a class="footnote-backref" href="#fnref:1" title="Jump back to footnote 2 in the text">↩</a></p>
</li>
<li id="fn:2">
<p>Fun fact: Peter Singer had taken over most of Mackie’s lectures so Mackie had time to write the book. <a class="footnote-backref" href="#fnref:2" title="Jump back to footnote 3 in the text">↩</a></p>
</li>
<li id="fn:3">
<p>And that is not even fully true because the impact may have manifested later due to another cluster of factors, but I think this type of replaceability is well understood in the movement. <a class="footnote-backref" href="#fnref:3" title="Jump back to footnote 4 in the text">↩</a></p>
</li>
<li id="fn:4">
<p>One might also credit the parents for sending their children to school, but the number of school children with intestinal worms is probably much greater than the number of children reached through school-based deworming, so that the parents’ contribution is mostly replaceable. If one parent hadn’t sent to school a child who benefitted from deworming, another child likely would have received the pill – though maybe with less elasticity than this example implies. <a class="footnote-backref" href="#fnref:4" title="Jump back to footnote 5 in the text">↩</a></p>
</li>
<li id="fn:5">
<p>Only if these experiments were conducted by the same organization that finally succeeded, the status quo is probably already to consider them. <a class="footnote-backref" href="#fnref:5" title="Jump back to footnote 6 in the text">↩</a></p>
</li>
<li id="fn:6">
<p>As an aside, impact certificates are probably also just as good incentives as the buyer makes them so that naive buyers can create perverse incentives just like normal funders. They may also send the often incorrect message that one necessary but insufficient factor owns part of the impact to sell it. That problem may be solved with different phrasing or disclaimers, but actual pricing according to an organization’s or person’s achievement would require perfect knowledge of the hypothetical optimal outcome and its mechanics, whose actions or omissions have led to degradations compared to that outcome, and the degree of degradation that all combinations of actions and omissions incur – which seems infeasible. <a class="footnote-backref" href="#fnref:6" title="Jump back to footnote 7 in the text">↩</a></p>
</li>
<li id="fn:7">
<p>One danger of rewarding cooperation is that sometimes it’s clear from the start that some cooperative effort would only come at a cost for one organization without benefit for anyone. In such a case an organization should not be incentivized to sacrifice its time for improved reputation. Some sort of Whuffy- or PageRank-based scoring can alleviate this problem by giving organizations with little reputation less power to award reputation. Such a system may suffer from some of the problems of startup valuation, since the reputation of new organizations will be based on hopes rather than track records and will thus be much less reliable. Since social status suffers from the same and worse limitations, the system may still be an improvement over the status quo. <a class="footnote-backref" href="#fnref:7" title="Jump back to footnote 8 in the text">↩</a></p>
</li>
</ol>
</div>Values Spreading Taxonomy2016-02-07T12:30:00+00:002016-02-07T12:30:00+00:00Dawn Dreschertag:impartial-priorities.org,2016-02-07:/values-spreading-taxonomy.html<p>Brian Tomasik has written about the <a href="http://foundational-research.org/gains-from-trade-through-compromise/">Gains from Trade through Compromise</a>. In practice I have repeatedly been in a position where I needed to refer back to specific scenarios discussed throughout the essay, so I resolved to categorize and number them and give them names. The result is an attempt at a taxonomy of modes of values spreading.</p><div class="admonition type">
<p class="admonition-title">Dated Content</p>
<p>I tend to update articles only when I remember their content and realize that I want to change something about it. But I rarely remember it well enough once about two years have passed. Such articles are therefore likely to contain some statements that I no longer espouse or would today frame differently.</p>
</div>
<div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#cooperation">Cooperation</a></li>
<li><a href="#education">Education</a><ul>
<li><a href="#correcting-misinformation">Correcting Misinformation</a></li>
<li><a href="#correcting-ignorance">Correcting Ignorance</a></li>
</ul>
</li>
<li><a href="#trade">Trade</a></li>
</ul>
</div>
<h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<p>The situations after which I had to refer to these modes were usually
failure scenarios that I sought to explain. Some were mild, like running
in open doors with the Child in the Pond analogy or discussing minutia
of the applicability of RCTs with someone who has not yet understood the
point of altruism. A more worrisome scenario is the one Brian uses,
where a deep ecologist – someone who values diversity in nature
intrinsically – comes to see preference or hedonistic utilitarians as enemies.</p>
<p>In a <a href="https://www.facebook.com/drescher.denis/posts/10153466679088691">recent discussion</a>,
friends of mine have pointed out that there are
certain people that seek competition, so that it may be effortful for
them to suspend it in favor of cooperation as as all modes require to
different degrees. The result of the discussion was that it may be more
feasible to funnel this competitiveness into channels where it is
unlikely to do harm rather than to just tell them to suspend it. I will
leave any ideas for this mechanism outside the scope of this article.</p>
<p>Agents interested in one or several of the same moral goals will soon
notice that they can reach these more easily when they cooperate on
realizing them, convince more agents to join their cause, avoid zero-sum
competition, and possibly even trade with groups with different goals:</p>
<ol>
<li>
<p>Cooperation. Agents can agree on some of their (not necessarily
terminal) moral goals and cooperate toward their realization.</p>
</li>
<li>
<p>Education. Education comes in two subforms, both of which can
generally be thought of as cooperative behaviors.</p>
<ol>
<li>
<p>Correcting misinformation. Helping other agents realize that some of their assumptions about the world are mistaken.</p>
</li>
<li>
<p>Correcting ignorance. Educating other agents on topics they had not considered or did not know enough about.</p>
</li>
</ol>
</li>
<li>
<p>Trade. When it has either been established that both sides disagree
on preferences rather than beliefs or when both sides find that the
other is resistant to education to the point that it would make it
prohibitively costly, then both sides can usually gain from avoiding
a zero-sum fight in favor of compromise.</p>
</li>
</ol>
<p>These are explained in more detail below.</p>
<h2 id="cooperation"><a class="toclink" href="#cooperation">Cooperation</a></h2>
<p>When agents have several moral goals each, the situation resembles a
stag hunt in that there are two Nash equilibria – when the agents
cooperate and when they defect. (In the Prisoner’s Dilemma, for
comparison, the only Nash equilibrium is the case of both agents
defecting.) Risk and payoff of the options, however, can go either way.</p>
<p>To do so, they will need to agree on pursuing the moral goal that they
share. A concrete example may be that one activist interested in the
proliferation of knowledge and the reduction of suffering and a second
activist interested in the reduction of suffering and in not lying can
agree on their shared goal of the reduction of suffering. A variation is
the situation where at least one agent shares only an instrumental goal
with the other agent. Thus, for example, a queer activist and an anti-/
activist can both agree to support secularization efforts in Uganda. Of
course, the agents can enter into several such cooperative efforts if
their resources allow.</p>
<h2 id="education"><a class="toclink" href="#education">Education</a></h2>
<p>Education comes in two different forms depending on whether the
recipient is mistaken or merely oblivious.</p>
<h3 id="correcting-misinformation"><a class="toclink" href="#correcting-misinformation">Correcting Misinformation</a></h3>
<p>When people do not seem to share any moral goals, it may be the case
that one party has not realized that what they took for terminal moral
goals are merely instrumental moral goals, and possibly badly chosen
ones, so that the relationship is not apparent. In that case, one
cost-effective and cooperative way of winning them for the cooperative
effort would be to provide them with the information they are lacking.</p>
<p>A person might be enrolled as sponsor of a child in Kenya and might
scoff at people that support obscure interventions like the treatment of
schistosomiasis and believe that they care about the well-being of their
sponsored child. But upon learning about the identified victim effect
and scope neglect, the person may do further research and find that the
charity that runs the child sponsorship program realized years ago that
providing such aid to individual children is inefficient compared to
communal programs and states in the fine print that they interpret child
sponsorship such that they use the funds to provide communal programs
such as treatment of schistosomiasis to the village where the child lives.</p>
<p>After some soul-searching, the person may come to the conclusion that
they cannot begrudge the charity that decision, and that really it would
be consistent with their morality to provide more than one child with
the same level of well-being if they can do so at the same cost. The
person may come to realize that it is really the well-being of the
greatest number of people that they care about and not only that of a
specific child, and they will be ready to cooperate with the group that
focuses entirely on providing schistosomiasis treatment.</p>
<h3 id="correcting-ignorance"><a class="toclink" href="#correcting-ignorance">Correcting Ignorance</a></h3>
<p>A variation of this mode occurs when a person has never considered some
key question. They may not be aware that there is a decision that they
have constantly been making simply by omission because they were unaware
of it. Telling such people about the options they have may greatly
enrich their thinking and their lives. It may also come at a benefit to
the internal consistency of their moral system or their identity and
thus reduce cognitive dissonance. Some people may never have considered
philanthropy, for example, so simply by telling them about this
possibility one can create new cooperation partners.</p>
<h2 id="trade"><a class="toclink" href="#trade">Trade</a></h2>
<p>Other people, however, may have the same convictions as the misinformed
one, but may not actually lack information. They may be well aware of
all one’s arguments and still not be swayed by them. Here the education
approach will fail. What remains is trade.</p>
<p>We should note, however, that the distinction between ethical and
epistemic differences can be a fuzzy one. When someone’s (epistemic)
beliefs are very different from one’s own, it can be very costly to
convince the other person all the way to one’s own view even if one is
objectively right, so that compromise is again more efficient.</p>
<p>When these people have a strong interest in spreading their values, we
encounter a situation that resembles the Prisoner’s Dilemma. Much
research has been invested into this scenario. <a href="https://books.google.de/books?id=GxRo5hZtxkEC">Axelrod</a>,
for example, has boiled down his insights from a competition of
algorithms in the repeated prisoner’s dilemma into four rules:</p>
<ol>
<li>Don’t be envious.</li>
<li>Don’t be the first to defect.</li>
<li>Reciprocate both cooperation and defection.</li>
<li>Don’t be too clever.</li>
</ol>
<p><img alt="Visualization of mutual gains from trade by Brian Tomasik." src="images/gains-from-trade.png">
<em>Visualization of mutual gains from trade by Brian Tomasik.</em></p>
<p>One very simple algorithm that implemented this behavior was <em>Tit for
Tat</em>, which cooperated unless the other agent chose to defect in the
last round. Thus it quickly forgave defection, cooperated by default,
punished and rewarded defection and cooperation respectively, and was
perfectly predictable.</p>
<p><a href="http://foundational-research.org/gains-from-trade-through-compromise/">Brian Tomasik has investigated what this could mean for altruists</a>.
He compares the case of deep ecologists, which care terminally about
diversity of life and noninterference, and animal welfarists, which care
about the well-being of all sentient beings. These
two are often opposed. Humans have been able to reduce the suffering due
to disease, starvation, adverse climate, and many other sources by
developing technical, medical, and economical remedies. Animals are
still exposed to all these to maximal degrees and cannot change these
conditions for themselves, centrally due to their lower intelligence.
Hence improving animal welfare will require intervention in nature,
presumably by humans directly or proximately.</p>
<p>When these opposed factions both want to control the future to implement
their moral preferences, they would have to fight each other, a fight
they would win with some probability <span class="math">\(p_i < 1\)</span> where the values of <span class="math">\(p_i\)</span>
may sum up to 1 or even less than one.</p>
<p>However, both sides know aspects of the future that they care more
strongly about than about others while the same is reversed for the
opposing faction. Animal welfarists, for example, may value a world with
few animals suffering greatly over a world with many animals, but deep
ecologists may be indifferent about the number of animals per species so
long as diversity is maintained.</p>
<p>Based on their estimates of how likely each faction would be to
overpower the other in a fight and the value they put on different
aspects of their utopias, they can arrive at a compromise that is better
with certainty than the expected value of a fight.</p>
<p>This is already happening, for example when the <span class="caps">VEBU</span> (“Vegetarian Union”)
cooperates with the sausage producer Rügenwalder on vegetarian products
that reduce animal suffering and are a financial success for the
company, both side had to make some concessions, but the result was a
<a href="https://vebu.de/news/2419-wursthersteller-fleischfrei">mutual gain</a>.</p>
<script type="text/javascript">if (!document.getElementById('mathjaxscript_pelican_#%@#$@#')) {
var align = "center",
indent = "0em",
linebreak = "false";
if (false) {
align = (screen.width < 768) ? "left" : align;
indent = (screen.width < 768) ? "0em" : indent;
linebreak = (screen.width < 768) ? 'true' : linebreak;
}
var mathjaxscript = document.createElement('script');
mathjaxscript.id = 'mathjaxscript_pelican_#%@#$@#';
mathjaxscript.type = 'text/javascript';
mathjaxscript.src = 'https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.3/latest.js?config=TeX-AMS-MML_HTMLorMML';
var configscript = document.createElement('script');
configscript.type = 'text/x-mathjax-config';
configscript[(window.opera ? "innerHTML" : "text")] =
"MathJax.Hub.Config({" +
" config: ['MMLorHTML.js']," +
" TeX: { extensions: ['AMSmath.js','AMSsymbols.js','noErrors.js','noUndefined.js'], equationNumbers: { autoNumber: 'none' } }," +
" jax: ['input/TeX','input/MathML','output/HTML-CSS']," +
" extensions: ['tex2jax.js','mml2jax.js','MathMenu.js','MathZoom.js']," +
" displayAlign: '"+ align +"'," +
" displayIndent: '"+ indent +"'," +
" showMathMenu: true," +
" messageStyle: 'normal'," +
" tex2jax: { " +
" inlineMath: [ ['\\\\(','\\\\)'] ], " +
" displayMath: [ ['$$','$$'] ]," +
" processEscapes: true," +
" preview: 'TeX'," +
" }, " +
" 'HTML-CSS': { " +
" availableFonts: ['STIX', 'TeX']," +
" preferredFont: 'STIX'," +
" styles: { '.MathJax_Display, .MathJax .mo, .MathJax .mi, .MathJax .mn': {color: 'inherit ! important'} }," +
" linebreaks: { automatic: "+ linebreak +", width: '90% container' }," +
" }, " +
"}); " +
"if ('default' !== 'default') {" +
"MathJax.Hub.Register.StartupHook('HTML-CSS Jax Ready',function () {" +
"var VARIANT = MathJax.OutputJax['HTML-CSS'].FONTDATA.VARIANT;" +
"VARIANT['normal'].fonts.unshift('MathJax_default');" +
"VARIANT['bold'].fonts.unshift('MathJax_default-bold');" +
"VARIANT['italic'].fonts.unshift('MathJax_default-italic');" +
"VARIANT['-tex-mathit'].fonts.unshift('MathJax_default-italic');" +
"});" +
"MathJax.Hub.Register.StartupHook('SVG Jax Ready',function () {" +
"var VARIANT = MathJax.OutputJax.SVG.FONTDATA.VARIANT;" +
"VARIANT['normal'].fonts.unshift('MathJax_default');" +
"VARIANT['bold'].fonts.unshift('MathJax_default-bold');" +
"VARIANT['italic'].fonts.unshift('MathJax_default-italic');" +
"VARIANT['-tex-mathit'].fonts.unshift('MathJax_default-italic');" +
"});" +
"}";
(document.body || document.getElementsByTagName('head')[0]).appendChild(configscript);
(document.body || document.getElementsByTagName('head')[0]).appendChild(mathjaxscript);
}
</script>Concept for Donor Coordination2016-01-23T18:30:00+00:002016-01-23T18:30:00+00:00Dawn Dreschertag:impartial-priorities.org,2016-01-23:/concept-for-donor-coordination.html<p>This is a proposal for a donor coordination system that aims to empower donors to harness the risk neutrality that stems from their combined work toward agent-neutral goals.</p><div class="admonition type">
<p class="admonition-title">Dated Content</p>
<p>I tend to update articles only when I remember their content and realize that I want to change something about it. But I rarely remember it well enough once about two years have passed. Such articles are therefore likely to contain some statements that I no longer espouse or would today frame differently.</p>
</div>
<div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#donor-coordination">Donor Coordination</a><ul>
<li><a href="#irrational-risk-aversion">Irrational Risk-Aversion</a></li>
<li><a href="#competition-for-exposure">Competition for Exposure</a></li>
</ul>
</li>
<li><a href="#high-level-goals">High-Level Goals</a></li>
<li><a href="#core-concepts">Core Concepts</a></li>
<li><a href="#interest-groups">Interest Groups</a></li>
<li><a href="#functional-requirements">Functional Requirements</a></li>
<li><a href="#challenges-and-proposed-solutions">Challenges and Proposed Solutions</a><ul>
<li><a href="#descriptive-statistics">Descriptive Statistics</a></li>
<li><a href="#funding-gaps">Funding Gaps</a></li>
<li><a href="#allocation">Allocation</a></li>
<li><a href="#donation-swaps">Donation Swaps</a></li>
</ul>
</li>
<li><a href="#remaining-challenges">Remaining Challenges</a><ul>
<li><a href="#moral-lies">Moral Lies</a></li>
<li><a href="#market-research">Market Research</a></li>
<li><a href="#community-building">Community Building</a></li>
</ul>
</li>
</ul>
</div>
<h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<p>This proposal is meant to encourage comments on its content as well as comments along the lines of “I would use this,” because without many of those it will not seem like a worthwhile undertaking to implement it.</p>
<h2 id="donor-coordination"><a class="toclink" href="#donor-coordination">Donor Coordination</a></h2>
<p>One problem that <a href="http://blog.givewell.org/2014/12/02/donor-coordination-and-the-givers-dilemma/">GiveWell has struggled with</a> emerges when two donors are
not fully value-aligned but can agree on wanting to fund one of
GiveWell’s top charities. The result is that they wait each other out, a
deadlock, both wanting the other to fund the GiveWell charity, because
they value the other’s counterfactual use of the donation lower than
their own. GiveWell is regularly refining its
response to this problem.</p>
<p>For this problem to become relevant, there need to be at least two large
donors or monolithic groups of donors, where large means that their
planned donations are close to – for example within the same order of
magnitude of – the funding gap of the charities in question. This is a
good problem to have.</p>
<p>More commonly, however, the funding gap is large compared to the
potential individual donations (where <em>individual</em> is meant to exclude
the aforementioned monolithic groups of donors), so that the above
problem becomes an edge case while centrally we face a different
problem. Donors that focus their contributions on charities that have a
significant evidence base and track record for impact – a large part of
the “GiveWell market” – are often accused of being too focused on just
these established charities thereby missing small high-impact
opportunities from nonprofit startups or projects that will stay small
or short-lived by design.</p>
<p>The distinction is similar to that between, on the one hand, passive
investors that buy exchange-traded funds (ETFs) of, for example, the top
30 (<span class="caps">DAX</span>) or 500+ (S&P 500) companies in order to hold them, and, on the
other hand, business angels or venture capitalists that invest into
startups. The first group has excellent information to make relatively
low risk–low return investments; the second group has to rely on rough
heuristics, such as their faith into the founders, to make high
risk–high return investments – of which they need to be able to make
many in order to profit at least fairly reliably.</p>
<p>But a profit motive is an agent-relative goal. Investors (such as
donors) with agent-neutral goals that are shared by at least a few
others have much better opportunities for cooperation. These have
largely not been tapped into. While Net Analytics is clearly focused on
the low risk–low return market, this high risk–high return market also
calls for a software solution to its coordination problem.</p>
<p>The central motivating problems are the following:</p>
<h3 id="irrational-risk-aversion"><a class="toclink" href="#irrational-risk-aversion">Irrational Risk-Aversion</a></h3>
<p>Drops in marginal utility of a resource suggest risk aversion. In that
context it is rational to prefer a low return with high probability to a
high return with low probability at the same expected value. In the
context of altruistic interventions,<sup id="fnref:20"><a class="footnote-ref" href="#fn:20">1</a></sup> the utility of marginal
donations only noticeably decreases when it reaches the area of millions
of dollars, some of the GiveWell top charities’ funding gaps. Since few
donors have funds of that magnitude at their disposal, most risk
aversion of average donors is disproportionate.</p>
<p>At the same time there are many donors that see a high likelihood that
effective interventions are possible in a certain cause area.
Unfortunately, these intervention are, by necessity, more speculative
than, for example, the interventions GiveWell prioritizes. Yet there are
charity startups implementing them.</p>
<p>The funding gaps of these charities tend to be too small for any
respectable prioritization organization, like GiveWell, to warrant
investing staff time into evaluating them, so donors are left to their
own devices.<sup id="fnref:21"><a class="footnote-ref" href="#fn:21">2</a></sup></p>
<p>When donors consider these charities, they are usually still optimistic
that donating to them does yield superior impact, but they have a much
harder time prioritizing between them because their central metric just
remains how well they implement very similar interventions. It is well
possible that the differences between these charities – charities that
many impact-oriented donors are actively considering – is small enough
that the value of the information would not warrant its cost.</p>
<p>Unfortunately, some of these donors fall into a form of analysis
paralysis at this point and rather donate to the charities whose lower
impact is well proven. Other donors react more rationally and donate
rather arbitrarily within the group of the most highly effective
charities. Others again use questionable heuristics, often aware that
they are likely to be unreliable but also aware of the presumably low
value of information of more thorough investigations. I aver that none
of these strategies is optimal.</p>
<h3 id="competition-for-exposure"><a class="toclink" href="#competition-for-exposure">Competition for Exposure</a></h3>
<p>The other side of the medal is that charities are aware of these
dynamics. While their values may be aligned, for funding they are yet
each dependent on its own pool of donors, and any cross-promotion of
another charity among the first charity’s own donor base may lead to
donors shifting their support to the endorsed organization. This
behavior stifles cooperation.</p>
<p>The solution presented here will instead allow all charities in a
program area to fill their funding gaps to similar degrees. If a
sufficient number of donors come to accept this solution, any incentive
for charities to engage in uncooperative behavior will be diminished.</p>
<p>Donor Coordination (working title) is a software system and strategy
that fosters cooperation between value-aligned donors by allowing them
do make large contributions in teams and donate to whole program areas
rather than individual nonprofits. It can improve upon the current state
if it is accepted and trusted by a sufficient number of donors.</p>
<h2 id="high-level-goals"><a class="toclink" href="#high-level-goals">High-Level Goals</a></h2>
<p>The donor coordination solution can be considered successful when it
achieves the following goals:</p>
<dl>
<dt>Team-level atomicity</dt>
<dd>Donors can choose portfolios with whom they are value-aligned to the
point that they perceive their donations as coming from the team of
donors that invests into that portfolio rather than them personally.</dd>
<dt>Program-level atomicity</dt>
<dd>
<ol>
<li>
<p>Donors can choose charity portfolios that, as a whole, represent
their moral preferences well enough that they perceive their
team as donating to a program area rather than an individual charity.</p>
</li>
<li>
<p>Charities are value-aligned with the organization their
donations are fungible with to the point that they make fully
altruistic statements about their funding gaps.</p>
</li>
</ol>
</dd>
</dl>
<h2 id="core-concepts"><a class="toclink" href="#core-concepts">Core Concepts</a></h2>
<p>The donor coordination solution should likely take the shape of a web
application to enable users of any platform to use it. The idea is
roughly inspired by Wikifolio.</p>
<dl>
<dt>Visitor</dt>
<dd>An unauthenticated person viewing the website.</dd>
<dt>User</dt>
<dd>A donor, a charity, or an administrator.</dd>
<dt>Portfolio</dt>
<dd>The portfolio is an allocation rule that partitions funds among a
set of charities. Every user can create portfolios, favorite or
watch portfolios, and donate to portfolios.</dd>
<dt>Donor</dt>
<dd>The donor is a user other than a charity.</dd>
<dt>Charity</dt>
<dd>A charity is a user that only has the ability to enter some meta
data about itself and its funding gap, and participate in discussions.</dd>
</dl>
<h2 id="interest-groups"><a class="toclink" href="#interest-groups">Interest Groups</a></h2>
<p>The interests of beneficiaries are:</p>
<ol>
<li>
<p>Beneficiaries want to maximize the available funding toward the
their preferences.</p>
</li>
<li>
<p>Beneficiaries want the, at the margin, most effective interventions
to receives maximal funding.</p>
</li>
<li>
<p>Beneficiaries want the funding gaps of the most effective
interventions to be greater than or equal to the available funding.</p>
</li>
</ol>
<p>In some cases the beneficiaries can give direct input, but in many cases
their interests need to be represented by donors and charities because
they have insufficient levels of intelligence to express them
efficiently or are not yet born.</p>
<p>Hence the interests of donors are:</p>
<ol>
<li>
<p>Donors want to maximize the available funding toward the their moral goals.</p>
</li>
<li>
<p>Donors want the, at the margin, most effective interventions
realizing these moral goals to receive maximal funding.</p>
</li>
<li>
<p>Donors want the funding gaps of the most effective interventions
realizing these moral goals to be greater than or equal to the
available funding.</p>
</li>
</ol>
<p>Hence the interests of the charities are:</p>
<ol>
<li>
<p>Charities want to maximize the available funding toward the
charity’s moral goals.</p>
</li>
<li>
<p>Charities want the, at the margin, most effective interventions
realizing these moral goals to receive maximal funding.</p>
</li>
<li>
<p>Charities want the funding gaps of the most effective interventions
realizing these moral goals to be greater than or equal to the
available funding.</p>
</li>
</ol>
<p>The main difference between donors and charities as two groups is the
direction of the money flow. The main difference between the donors and
charities internally is their different moral goal makeup.</p>
<p>From these primary interests follow proximate interests for
value-aligned teams of donors (all donors to a program area as defined
by a public portfolio):</p>
<ol>
<li>
<p>Being value aligned, the members of a team are happy to make their
donations fungible with the donations of all other members of the team.</p>
</li>
<li>
<p>Since their value alignment with other teams varies, there may be
teams with partially opposing moral goals. Teams will want to
minimize fungibility with such teams.</p>
</li>
<li>
<p>Since the funding gaps of charities are limited, teams also want to
increase the funding gap of their program area by broadening its scope.</p>
</li>
</ol>
<p>Analogously for charities:</p>
<ol>
<li>
<p>The charities of popular portfolios are likely to be highly value
aligned and thus happy to calculate their funding gaps cooperatively.</p>
</li>
<li>
<p>Since their value alignment with charities of other program areas
varies, there may be portfolios of charities with partially opposing
moral goals. Charities will want to increase their scale in order to
be able to enter greater funding gaps so portfolio authors can
minimize fungibility with such opposing program areas.</p>
</li>
</ol>
<p>Clearly, the last two interests of the donor teams are in conflict.
Small donation flows will favor portfolios of small, pure clusters of
charities while greater donation flows will necessitate compromise in
order to form greater, less pure clusters with larger funding gaps.</p>
<p>For simplicity I assume that all donors are perfectly informed and their
only differences are differences of value alignment. This is unlikely to
be the case in practice, but the only difference between a donor that is
not value aligned and a donor that acts as if they were not value
aligned because of lacking information is that the latter can be educated.</p>
<p>This educational mission is without the purview of Donor Coordination,
but the software should provide the platform that donors will need to
educate each other because this may be important for fostering user activity.</p>
<h2 id="functional-requirements"><a class="toclink" href="#functional-requirements">Functional Requirements</a></h2>
<ol>
<li>
<p>Visitors can create donor accounts.</p>
</li>
<li>
<p>Administrators can create administrator accounts.</p>
</li>
<li>
<p>Administrators can create charity accounts.</p>
</li>
<li>
<p>Visitors can view public portfolios including their descriptive statistics.</p>
</li>
<li>
<p>Donors can add public portfolios to their watch list.</p>
</li>
<li>
<p>Donors can donate to public portfolios.</p>
</li>
<li>
<p>Donors can author public portfolios.</p>
</li>
<li>
<p>Donors can draft and test portfolios in a private or draft state.</p>
</li>
<li>
<p>Donors can comment on portfolios.</p>
</li>
<li>
<p>Charities can enter new funding gaps for themselves.</p>
</li>
<li>
<p>Charities can enter new system-external donation flows.</p>
</li>
<li>
<p>Administrators have all privileges.</p>
</li>
</ol>
<p>At some point moderator accounts will become necessary, so moderators do
not need to enjoy the same level of trust as administrators to
contribute to the community maintenance.</p>
<h2 id="challenges-and-proposed-solutions"><a class="toclink" href="#challenges-and-proposed-solutions">Challenges and Proposed Solutions</a></h2>
<h3 id="descriptive-statistics"><a class="toclink" href="#descriptive-statistics">Descriptive Statistics</a></h3>
<p>The functional requirements mention descriptive statistics. These are
important for portfolio authors and other donors to decide how to
structure a portfolio so not to duplicate very similar ones or which
portfolio to donate to. At least two metrics are required:</p>
<ol>
<li>
<p>The sum of the funding gaps of the charities in a portfolio <span class="math">\(P\)</span>,
<span class="math">\(\operatorname{gap}(P) = \sum\limits_{c}^{P} \operatorname{gap}(c)\)</span>.</p>
</li>
<li>
<p>A ranked list of the portfolios with the highest fungibility but
lowest similarity. One idea may be the quotient,
<span class="math">\(\operatorname{compromise}(P, P') = \frac{\operatorname{fungibility}(P, P')}{\operatorname{similarity}(P, P')}\)</span>,
of the following metrics:</p>
<ol>
<li>
<p><span class="math">\(\operatorname{fungibility}(P, P') = \sum\limits_{c}^{P \cap P'} \operatorname{gap}(c)\)</span></p>
</li>
<li>
<p><span class="math">\(\operatorname{similarity}(P, P') = |\bigcup\limits_{c}^{P} \operatorname{donors}(c) \cap \bigcup\limits_{c}^{P'} \operatorname{donors}(c)|\)</span>.</p>
</li>
</ol>
</li>
</ol>
<p>The fungibility and similarity metrics should also be displayed in
isolation, particularly as a guide for authors of portfolios of new
charities when the combined compromise metric is undefined.</p>
<p>It may also become necessary to take weights into account, and the
formulas will surely need to be tweaked further once real data become available.</p>
<h3 id="funding-gaps"><a class="toclink" href="#funding-gaps">Funding Gaps</a></h3>
<p>There needs to be a common definition of a funding gap, so that
charities have hard, unyielding guidelines as to what figure to enter
for a given year.</p>
<p>Prioritization organizations already face a similar problem: Imagine two
charities, charity A with the ability to invest $100 million with some
baseline effectiveness <span class="math">\(e\)</span> on average and charity B with the ability to
invest $10 million with an average effectiveness of <span class="math">\(10e\)</span> within a
given year. Further assume that the charities are value aligned to
simplify the problem to one dimension of impact.</p>
<p>A commonly used uncertainty discount is 3% p.a. and for simplicity we
assume that suffering in the world, absent the interventions, remains
constant, so that aggregate suffering increases linearly over time.</p>
<p>A donor that wants to invest $100 million now has the choice to donate
it to charity A, knowing that it will be invested in the same year, or
to charity B, knowing that $10 million of it will be invested in the
same year, $90 million of it will wait on the charity’s bank account at
an interest rate of maybe 1% for another year, $80 million plus
interest will wait for two years, and so on.</p>
<p>Clearly, a definition of funding gaps that only takes into account a
charity’s ability to invest some amount per year would set very
different bars for the marginal impact of the last dollar of that
funding gap.</p>
<p>Since the donor coordination solution addresses coordination problems
that arise when the funding gaps of the individual charities do not
warrant the attention of a prioritization organization, we assume that
there are no meaningful differences of their relative effectiveness, so
we face a simpler version of this problem.</p>
<p>One solution may be to adopt GiveWell’s excess assets policy: “We seek
to be in a financial position such that our cash flow projections show
us having 12 months’ worth of unrestricted assets in each of the next 12 months.”</p>
<h3 id="allocation"><a class="toclink" href="#allocation">Allocation</a></h3>
<p>Another open question is the allocation of donations within the
portfolio. Conceptually, donors donate to program areas, but factually
they will have to transfer their donation to a specific organization.
Splitting it up across several organizations would be an unnecessary
hassle of the donor, so the algorithm that suggests the specific
organization should know some ideal allocation and then recommend a
recipient organization such that the actual allocation comes closest to
the ideal allocation. It could also take tax deductibility into account
as a tie breaker.</p>
<p>The simplest option might be an equitable allocation where the algorithm
aims to assign the same level of funding to each charity after taking
donations external to the system into account.</p>
<p>Another option may be to prioritize small funding gaps as an additional
incentive for charities not to exaggerate their funding gaps in the
<em>moral lies</em> scenario. However that would have little effect since the
charities in a given program area are value aligned and can thus easily
conspire with each other, and it may have the detrimental effect that
charities would be incentivized to be tardy with entering new funding gaps.</p>
<h3 id="donation-swaps"><a class="toclink" href="#donation-swaps">Donation Swaps</a></h3>
<p>Donors often agree on donation swaps where each partner donates to the
charity of choice of the other partner in order to harness the tax
deduction of the charity in the respective country.</p>
<p>In order to help portfolio authors to trade off fungibility against
funding gaps, there would need to be a ranking of other portfolios that
the given portfolio is most fungible with. However, portfolios whose
audience is very similar are least interesting to portfolio authors, so
the ranking should be sorted by something like the fungibility per
cardinality of the cut set of donors, and here donation swaps would add
noise to the calculation.</p>
<p>It needs to be either clear to the donors that they need to enter the
donation of their swap partner as their donation or the software should
allow them to mark donations as swaps and enter their partner. The first
is probably the better solution for an <span class="caps">MVP</span>, but the second may be more foolproof.</p>
<h2 id="remaining-challenges"><a class="toclink" href="#remaining-challenges">Remaining Challenges</a></h2>
<h3 id="moral-lies"><a class="toclink" href="#moral-lies">Moral Lies</a></h3>
<p>When there is a pair of program areas such that the teams of each see
the team of the other as an opposing team, but there is some set of
charities that they can agree on, and the available funding is close to
or greater than the available funding gaps of their program areas
without the consensus charities, the intended result is that donors
compromise and add charities to their portfolios that increase the
funding gap at the cost of greater fungibility.</p>
<p>But charities are of course value aligned with these teams. Hence it
will be ethical for them to lie about their funding gaps, inflating
them, to drive the opposing donors to fund the more fungible funding
gaps. Analogously, the opposing team’s charities can also inflate their
funding gaps; they even have to lest their cause suffer. When one group
defects in such a fashion, the cooperation breaks down. A classical
example of the prisoner’s dilemma.</p>
<p>In practice, the donor coordination solution will be used mostly or at
least at first only by donors that are all fairly value aligned at least
to the extend that they value the type of moral plurality that exists
among them. Hence this problem may not manifest any time soon.</p>
<h3 id="market-research"><a class="toclink" href="#market-research">Market Research</a></h3>
<p>My experience that such a software would be helpful is based on reports
of friends, some of whom are donors and some employees of affected
charities. Unless, however, there is a sizable number of prospective
users that are interested in the project, charities will not have
sufficient faith into the growth of the user base to warrant their time investments.</p>
<p>Apart from surveys among likely prospective users, one central market
research tool needs to be a minimal viable product (<span class="caps">MVP</span>). Other donors
and nonprofit staff have considered opening a group on a social network
such as Facebook to bring together all participants in whose actions need
coordination. The group would provide a means for communication but would
leave any functions beyond that to the participants to be implemented in
a manual, ad-hoc fashion. This way it will become clear which processes
are in most urgent need of automation. It will also become clear if the
community is large enough to sustain a more comprehensive solution like
the one proposed here.</p>
<h3 id="community-building"><a class="toclink" href="#community-building">Community Building</a></h3>
<p>An important strategic and marketing problem is the following: Entering
funding gaps will only warrant the effort for the charities if they can
expect significant donation flows from the donor coordination solution.
For donors the donor coordination solution is only interesting when the
program areas they want to donate to are well represented by charities
working on them.</p>
<p>One solution may be for administrators to regularly poll information on
funding gaps from charities and invite them to claim their accounts
themselves. That way, the administrators will have added effort during
the startup phase, which will be increasingly outsourced to the
charities as donors come to accept the system.</p>
<p>To achieve said donor acceptance, it would be helpful if the project
were run by a reputable organization with considerable reach, and if the
project collected early signups prior to its launch, both in order for
it to launch with momentum. Until such an organization has been found, I
cannot consider this challenge solved.</p>
<div class="footnote">
<hr>
<ol>
<li id="fn:20">
<p>Please note that in the following I will use “intervention” and
“program” semantically interchangeably conditional on which terms
seems more idiomatic to me in the collocational context. <a class="footnote-backref" href="#fnref:20" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
<li id="fn:21">
<p>“Respectable,” here, is not meant to denigrate any other
hypothetical prioritization organizations but rather meant as a
handicap, since an organization that is highly respected has to go
to great lengths to stress the low quality of its research when it
wants to invest staff time proportionate to evaluating interventions
with small funding gaps lest donors assume that the results are as
reliable as other results the organization puts out. Taking such a
risk is rarely warranted for such an organization. <a class="footnote-backref" href="#fnref:21" title="Jump back to footnote 2 in the text">↩</a></p>
</li>
</ol>
</div>
<script type="text/javascript">if (!document.getElementById('mathjaxscript_pelican_#%@#$@#')) {
var align = "center",
indent = "0em",
linebreak = "false";
if (false) {
align = (screen.width < 768) ? "left" : align;
indent = (screen.width < 768) ? "0em" : indent;
linebreak = (screen.width < 768) ? 'true' : linebreak;
}
var mathjaxscript = document.createElement('script');
mathjaxscript.id = 'mathjaxscript_pelican_#%@#$@#';
mathjaxscript.type = 'text/javascript';
mathjaxscript.src = 'https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.3/latest.js?config=TeX-AMS-MML_HTMLorMML';
var configscript = document.createElement('script');
configscript.type = 'text/x-mathjax-config';
configscript[(window.opera ? "innerHTML" : "text")] =
"MathJax.Hub.Config({" +
" config: ['MMLorHTML.js']," +
" TeX: { extensions: ['AMSmath.js','AMSsymbols.js','noErrors.js','noUndefined.js'], equationNumbers: { autoNumber: 'none' } }," +
" jax: ['input/TeX','input/MathML','output/HTML-CSS']," +
" extensions: ['tex2jax.js','mml2jax.js','MathMenu.js','MathZoom.js']," +
" displayAlign: '"+ align +"'," +
" displayIndent: '"+ indent +"'," +
" showMathMenu: true," +
" messageStyle: 'normal'," +
" tex2jax: { " +
" inlineMath: [ ['\\\\(','\\\\)'] ], " +
" displayMath: [ ['$$','$$'] ]," +
" processEscapes: true," +
" preview: 'TeX'," +
" }, " +
" 'HTML-CSS': { " +
" availableFonts: ['STIX', 'TeX']," +
" preferredFont: 'STIX'," +
" styles: { '.MathJax_Display, .MathJax .mo, .MathJax .mi, .MathJax .mn': {color: 'inherit ! important'} }," +
" linebreaks: { automatic: "+ linebreak +", width: '90% container' }," +
" }, " +
"}); " +
"if ('default' !== 'default') {" +
"MathJax.Hub.Register.StartupHook('HTML-CSS Jax Ready',function () {" +
"var VARIANT = MathJax.OutputJax['HTML-CSS'].FONTDATA.VARIANT;" +
"VARIANT['normal'].fonts.unshift('MathJax_default');" +
"VARIANT['bold'].fonts.unshift('MathJax_default-bold');" +
"VARIANT['italic'].fonts.unshift('MathJax_default-italic');" +
"VARIANT['-tex-mathit'].fonts.unshift('MathJax_default-italic');" +
"});" +
"MathJax.Hub.Register.StartupHook('SVG Jax Ready',function () {" +
"var VARIANT = MathJax.OutputJax.SVG.FONTDATA.VARIANT;" +
"VARIANT['normal'].fonts.unshift('MathJax_default');" +
"VARIANT['bold'].fonts.unshift('MathJax_default-bold');" +
"VARIANT['italic'].fonts.unshift('MathJax_default-italic');" +
"VARIANT['-tex-mathit'].fonts.unshift('MathJax_default-italic');" +
"});" +
"}";
(document.body || document.getElementsByTagName('head')[0]).appendChild(configscript);
(document.body || document.getElementsByTagName('head')[0]).appendChild(mathjaxscript);
}
</script>Even More Reasons for Donor Coordination2015-10-27T05:30:00+00:002015-10-27T05:30:00+00:00Dawn Dreschertag:impartial-priorities.org,2015-10-27:/even-more-reasons-for-donor-coordination.html<p>This article argues that donor coordination is even more important than we already thought because it is a more efficient way to bridge the recommendation gap I described in the context of <a href="http://claviger.net/expected-utility-auctions.html">expected utility auctions</a>.</p><div class="admonition type">
<p class="admonition-title">Dated Content</p>
<p>I tend to update articles only when I remember their content and realize that I want to change something about it. But I rarely remember it well enough once about two years have passed. Such articles are therefore likely to contain some statements that I no longer espouse or would today frame differently.</p>
</div>
<div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#prioritization-research">Prioritization Research</a></li>
<li><a href="#donor-coordination">Donor Coordination</a></li>
<li><a href="#scaling-it-up">Scaling It Up</a></li>
<li><a href="#room-for-more-funding">Room for More Funding</a></li>
<li><a href="#clustering">Clustering</a></li>
<li><a href="#getting-it-going">Getting it Going</a></li>
</ul>
</div>
<h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<p>At least since GiveWell’s 2014 series of <a href="http://blog.givewell.org/2014/12/02/donor-coordination-and-the-givers-dilemma/">articles on donor coordination</a> it has been clear that donor coordination is an important project to realize so large donors can split their funding rather than wait each other out, and if there are thresholds to realizing a project that neither can meet alone, they can pool resources.</p>
<p>But another advantage of donor coordination concerns all donors, and in the short term especially donors who would like to donate to <span class="caps">EA</span> metacharities. To convey the idea, I need to critique my previous proposal of expected utility auctions.</p>
<h1 id="prioritization-research"><a class="toclink" href="#prioritization-research">Prioritization Research</a></h1>
<p>Prioritization research as GiveWell and Animal Charity Evaluators conduct it combines a number of advantages:</p>
<ol>
<li>EAs know where they can donate,</li>
<li>it coordinates donors insofar as the estimates for room for more funding are updated in yearly cycles or more often,</li>
<li>but most importantly, it generates a cash flow to charities that is a function of their efficiency at producing positive impact:<ol>
<li>thus linking a metric that they cannot ignore for their own sake (money) to a metric that they morally mustn’t ignore (impact),</li>
<li>thus incentivizing less effective charities to become more effective (and conduct and publish research to get there) to partake of that cash flow or even to retain donors if the movement builders are very successful,</li>
<li>thus creating a more efficient market for altruistic investments, which raises the effectiveness waterline of the whole charity sector.</li>
</ol>
</li>
</ol>
<p>Depending on how much money we EAs have at our disposal in the movement, there is more or less value in the the first two points – but the overwhelming value for the vast majority of society, the long-run future, and moral process is in the third point. And GiveWell has reaped all the low-hanging prioritization fruit here, so it now needs to climb higher in the metaphorical tree (with Open Phil). (<span class="caps">ACE</span> may well have some low-hanging fruit left to pluck.)</p>
<p>But prioritization organizations need to also apply prioritization to their own research. When the funding gap of a charity is so limited that the expected value of information gained by reviewing it is lower than the price in staff time paid for the review, conducting the review would come at a positive opportunity cost, which translates into suffering that could’ve been averted. Hence, charities need to be somewhat scalable to qualify for review.</p>
<p>In my post on expected utility auctions I explained that this leads to a recommendation gap for smaller and more speculative opportunities, especially now that it has been announced that <a href="https://www.facebook.com/groups/effective.altruists/permalink/946573212065703/">Open Phil is not planning to put out public recommendations</a>. Since donors don’t (and would be ill-advised to) split up their donations arbitrarily between countless charities, they cluster around the few safe bets that GiveWell and <span class="caps">ACE</span> have found, and insofar as they don’t, their donations are split up across scores of smaller or more speculative endeavors, so that this support becomes less visible. This looks like a streetlight effect but is really quite rational.</p>
<p>As a remedy I proposed to lower the bar for prioritization research, and crowdsource it to highly transparent hobbyists who can share the responsibility for the results in a StackExchange-like system.</p>
<h1 id="donor-coordination"><a class="toclink" href="#donor-coordination">Donor Coordination</a></h1>
<p>The value of that system needs to be divided by the months of work that would have to go into community building to get it going and multiplied by the probabilities that this community building effort will be successful and that the resulting recommendations will be valuable. <span class="caps">MVP</span> versions of the system were proposed in the comments, but it would still require a significant leap of faith for someone to drive the project.</p>
<p>Instead, how about donor coordination?</p>
<p>The majority of the small charities that we wish to donate to that are not in the focus of GiveWell and <span class="caps">ACE</span> will be charities in prioritization research itself (let’s include 80,000 Hours here for the moment) and in <span class="caps">EA</span> movement building. These may be further subdivided, but I’ll use these two as examples for “program areas” for now. (Or “intervention areas,” but “program areas” sounds smoother to me.)</p>
<p>Within these two program areas there are only charities that get <span class="caps">EA</span>, eliminating the need for the third and major advantage of prioritization research. Hence it will be much less cost-effective to engage in. That leave the first two advantages.</p>
<p>Further, the donor coordination problem GiveWell faced may, to a large part, not apply. The problem only occurs when donors have substantially more faith in their own judgements than in that of the other donors, either because they’ve put a lot of effort into forming them or because they have outré moral goals. In my experience, neither is common, at least among the donors of less than €10,000 per year that I’ve talked to, who are probably not the category GiveWell worries about.</p>
<p>In my experience donors often have the opposite problem. They wish to drive, e.g., <span class="caps">EA</span> movement building with their donations but don’t know which organization to donate to, even after doing substantial research, because they all seem to do great and value-aligned work. Hence the more risk-averse donors then fall back on a GiveWell or <span class="caps">ACE</span> charity instead while the more risk-neutral ones pick out one at random or precisely because of its room for more funding, thinking of it as the best proxy for cost-effectiveness that they can find.</p>
<p>Room for more funding may be a decent proxy in some cases, but greater scale often correlates with lower cost-effectiveness, so that it also seems plausible that this metric might lead us astray – so long as we think of it as a proxy of an individual charity’s cost-effectiveness.</p>
<p>But here’s the rub: Unlike in the program area of net distributions, the total room for more funding in <span class="caps">EA</span> movement building probably doesn’t go far beyond the money that is available in the movement (if at all), so instead of thinking of our donation as one to an individual charity, we should think of it as a donation to a program area.</p>
<p>What we need for that is only:</p>
<ol>
<li>
<p>a common definition of room for more funding,</p>
</li>
<li>
<p>a set of charities in the program area that are value-aligned, and thus can be trusted to report their room for more funding honestly,</p>
</li>
<li>
<p>a system for donor coordination, which can be as simple as a Facebook group as <span class="caps">MVP</span>.</p>
</li>
</ol>
<p>Since every donor fills some room for more funding and thus displaces an equivalent donation, all donations are fungible and have equivalent counterfactual cost-effectiveness: the average marginal cost-effectiveness of the total filled room for more funding of the whole program area.</p>
<p>This happens anyway, and donor coordination can only make it better since on average highly cost-effective <span class="caps">EA</span> charities may currently receive more donations than they can invest at the margin, and since a lot of <span class="caps">EA</span> donors may currently avoid <span class="caps">EA</span> metacharities in favor of individual safe bets. We may not be able to detect the very best opportunities, but those wouldn’t be worth too much if they’d be filled by a fraction of the available donor money and the effectiveness waterline is generally high.</p>
<p>The question is no longer whether a marginal donation to GiveWell or 80,000 Hours is more cost-effective but whether prioritization research is sufficiently cost-effective to deserve our donations.</p>
<h1 id="scaling-it-up"><a class="toclink" href="#scaling-it-up">Scaling It Up</a></h1>
<p>Prioritization research and <span class="caps">EA</span> movement building are some obvious first targets for the donor coordination project since there is a seizable consensus that these are still equally or more cost-effective than object-level donations, but what if we fill both these funding gaps?</p>
<p>GiveWell deprioritizes research in program areas when it finds that there are <a href="http://blog.givewell.org/2015/10/15/charities-wed-like-to-see/">no promising charities</a> implementing the program. If there are really no charities or no value-aligned charities, then this can’t be helped, but if there are many value-aligned charities, each of insufficient scale to warrant GiveWell’s attention, then an evaluation of the program area and a very cursory review of the charities (to ensure that they’re value-aligned) can be enough to provide the foundation for donor coordination in the area.</p>
<h1 id="room-for-more-funding"><a class="toclink" href="#room-for-more-funding">Room for More Funding</a></h1>
<p>The meaning of room for more funding is highly unclear to me at the moment, which would need to change in order to get commensurable estimates from different charities.</p>
<p>Does it mean the amount of funding an organization can invest before it would choose to leave further funds on its bank account beyond the end of the fiscal year? Or before a hypothetical investment of further funds would incur a significant drop in marginal cost-effectiveness? If so, how significant?</p>
<p>As an example, I imagine that <span class="caps">AMF</span> has to be fairly liquid, so it probably has a lot of money on bank accounts with low interest. So in the hypothetical case where it gets more money than it would’ve normally invested in the year, it has to weigh whether it wants to leave it there until the next year, when it’ll get better distributions finalized, or whether it is better to invest it less well than <span class="caps">AMF</span> would typically invest it. Maybe there are regions where it could easily conduct distributions, e.g., because Concern Universal works there, but that it has so far avoided because the malaria incidence is slightly lower than in others or the <span class="caps">IR</span> rate slightly higher. In general, suffering accumulates over time, <span class="caps">IR</span> rates increase, incidence decreases, so the second option might sometimes be better.</p>
<p>So if room for more funding just means that the extraneous money will be left on the account until after the end of the year, then it’s equivalent to some degree of a drop in marginal cost-effectiveness that <span class="caps">AMF</span> was not ready to accept. So what drop is that? A significant drop from <span class="caps">AMF</span>’s accustomed level of cost-effectiveness, or a drop below GiveDirectly level, or a drop not to GiveDirectly level but significant enough that given the other factors in the evaluation, it would temporarily lose its top charity status?</p>
<p>For <span class="caps">EA</span> movement building and possibly prioritization research, the leverage multiplier for long-run estimated cost-effectiveness beyond object-level donations might serve as such a baseline below which the room for more funding of a charity is closed. It might be set higher than one to account for the increased uncertainty of meta work. “Long-run,” because in the short term, heavy reinvestment into one’s own program will lead to a quicker increase in scale.</p>
<p>For other program areas, different metrics would have to be developed.</p>
<p>And as an ancillary question, in view of home bias and the heavy commitment that founders make to their charities, how likely is it that they might rationalize in favor of their own charity being the most effective, which would increase the room it has until it reaches some lower marginal effectiveness threshold and thus make it moral for them to direct somewhat more funds toward it? By extension, staffers might assume that this partiality affects all the other charities, so that their own would end up relatively neglected, which would again be objectively bad in their view.</p>
<h1 id="clustering"><a class="toclink" href="#clustering">Clustering</a></h1>
<p>I’d be happy for my donations to <span class="caps">ACE</span> to be fungible with donations to the Global Priorities Project, GiveWell, etc. assuming that they have funding gaps, but Descartes would disagree. Hence what constitutes a program area has to be the result of a clustering of donor interests, the data for which is probably available to Giving What We Can and <span class="caps">EA</span> Hub.</p>
<h1 id="getting-it-going"><a class="toclink" href="#getting-it-going">Getting it Going</a></h1>
<p><a href="https://www.facebook.com/peterhurford">Peter Hurford</a> has started a <a href="https://dotimpact.slack.com/messages/p-donor-coord/">Slack channel</a> for donor coordination, which you can join if you have a Slack account. <a href="https://www.facebook.com/evangaensbauer">Evan Gaensbauer</a> considers working on something like it. If you wish to help, please get in touch with them.</p>The Redundancy of Quantity2015-09-03T17:40:00+00:002015-09-03T17:40:00+00:00Dawn Dreschertag:impartial-priorities.org,2015-09-03:/the-redundancy-of-quantity.html<p>I’m advocating for a simpler way to think about donations and investments, namely, not to distinguish them.</p><div class="admonition type">
<p class="admonition-title">Dated Content</p>
<p>I tend to update articles only when I remember their content and realize that I want to change something about it. But I rarely remember it well enough once about two years have passed. Such articles are therefore likely to contain some statements that I no longer espouse or would today frame differently.</p>
</div>
<div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#the-advantages">The Advantages</a><ul>
<li><a href="#it-is-inclusive">It is Inclusive</a></li>
<li><a href="#it-is-motivating">It is Motivating</a></li>
<li><a href="#it-expands-the-circle">It Expands the Circle</a></li>
<li><a href="#it-reduces-inferential-distance">It Reduces Inferential Distance</a></li>
<li><a href="#it-encourages-counterfactual-reasoning">It Encourages Counterfactual Reasoning</a></li>
</ul>
</li>
<li><a href="#the-disadvantages">The Disadvantages</a><ul>
<li><a href="#the-magic-of-donation">The Magic of “Donation”</a></li>
<li><a href="#it-makes-budgeting-hard">It Makes Budgeting Hard</a></li>
</ul>
</li>
</ul>
</div>
<h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<p><img alt="donations-and-efficiency.png" src="/images/the-redundancy-of-quantity/donations-and-efficiency.png"></p>
<p>Inspired by the two-part nature of the term “effective altruism,” I often see the goal of the movement explained as pictured above. We want to increase the amounts that are donated and increase the efficiency of the donations to maximize the integral, the effect.<sup id="fnref:1"><a class="footnote-ref" href="#fn:1">1</a></sup> The graph is pilfered from Giving What We Can, and since they are centrally concerned with encouraging donations and use them as proxy for their impact, it makes sense for them to think in these terms. But I think this is not the most parsimonious model, and the alternative has advantages beyond elegance.</p>
<p>The alternative is to think only in terms of efficiency. Just as the beneficiaries of our donation, we’re also capable of producing and, significantly, leveraging utils, so rather than thinking in terms of <span class="caps">ROI</span> in one case and in terms of utils (or QALYs) in the other, we can just think in terms of utils in both cases. (Or whatever you’re optimizing for rather than utils.)</p>
<p>Effective altruists have long compared donations to investments or even already seen donations as another type of investment, so this may seem trivial to some, but making more consistent use of this perspective and using it to encourage altruists to donate efficiently has several advantages.</p>
<p>Below a few advantages and disadvantages I can think of. Do you have further ideas?</p>
<h2 id="the-advantages"><a class="toclink" href="#the-advantages">The Advantages</a></h2>
<h3 id="it-is-inclusive"><a class="toclink" href="#it-is-inclusive">It is Inclusive</a></h3>
<p>Someone who is poor – certainly when they’re third-world poor, but it’s a sliding scale, logarithmic I think – has an extremely much harder time investing time or money into others’ happiness than someone who is even modestly affluent. Not being able to have the same impact by donating, these people will feel like failures compared to many high-profile high-achievers in the movement. Yet they may be just as smart and compassionate, and just lost the birth lottery.</p>
<p>But the poorer they are the more they have the ability to cost-effectively improve their own happiness. I’ve read an article a while ago (which I can’t find anymore) about people in great poverty who have devised sophisticated ways of managing their debt while maintaining liquidity. Their methods may be highly cost-effective at improving total happiness at their small scale, even if they’re only leveraging their own or that of their family. If ultimately they would be altruists given the means, they should feel welcomed in the movement.</p>
<h3 id="it-is-motivating"><a class="toclink" href="#it-is-motivating">It is Motivating</a></h3>
<p>Friends of mine, and I too, often struggle with giving later. We are students, earn little, and know that it is best for us to invest in our education and careers first to earn as much as possible as early as possible, and then have for donating the maximum of money over the maximum of time. This phase feels preparatory and makes us impatient. But that is only because we think of donations as something qualitatively different from the other investments that we’re already making.</p>
<p>Once we come to the point where our System 1s accept investments in our education as just another vehicle to achieving our altruistic moral goals, we’re having an impact right now. The reduction in suffering that we effect may be deferred, but even with <span class="caps">AMF</span> our investments are converted into actual bed nets with a delay, so any difference is only quantitative.</p>
<h3 id="it-expands-the-circle"><a class="toclink" href="#it-expands-the-circle">It Expands the Circle</a></h3>
<p>People who have a strong separation between themselves and others encoded in their moral goals (rather than just as a habit they want to break) will find it intuitive to make a distinction between investments that increase their happiness and those that increase the happiness of others. The rest of us, however, may come to understand <span class="caps">EA</span> reasoning much more easily the better we understand how this separation (which we don’t actually find morally relevant) influences our thinking.</p>
<p>Just as not thinking in terms of this separation may decrease risk aversity in altruism and expand the moral circle, so making this distinction may have the counterproductive effect of reinforcing this separation.</p>
<h3 id="it-reduces-inferential-distance"><a class="toclink" href="#it-reduces-inferential-distance">It Reduces Inferential Distance</a></h3>
<p>Or it reduces something like inferential distance but between parts of our own model of the world rather than those of different people.</p>
<p>We are all fairly adroit at thinking about cost-effectiveness when it comes to everyday investments, so people new to effective altruism can readily transfer that experience to our new type of investment. Some of us have particular knowledge of finance and can probably recognize patterns more easily that will enable more effective altruist financial innovations.</p>
<h3 id="it-encourages-counterfactual-reasoning"><a class="toclink" href="#it-encourages-counterfactual-reasoning">It Encourages Counterfactual Reasoning</a></h3>
<p>The flip side of this is that it makes budgeting hard, as described below, but the positive side is that it encourages people to think more clearly about the relative quality of their investments. 6% p.a. in returns doesn’t seem so amazing anymore when you can get <a href="http://www.harbus.org/2015/you-could-be-the-warren-buffett-of-social-investing/">6000% within “a few years.”</a> When you need to trade off donating now vs. donating later, you have to think in terms of the same outcomes anyway.</p>
<h2 id="the-disadvantages"><a class="toclink" href="#the-disadvantages">The Disadvantages</a></h2>
<h3 id="the-magic-of-donation"><a class="toclink" href="#the-magic-of-donation">The Magic of “Donation”</a></h3>
<p>Just as the separation between donations and other investments can be harmful, it can also be used for good. Compare:</p>
<ol>
<li>
<p><span class="caps">EA</span>: “Please buy one of our buttons. They’re €5, and all proceeeds go to <span class="caps">AMF</span>.”<br>
Donor: “€5 for a cheap button like that? No offence, I know it’s for charity, but that’s a total rip-off. I had some made last year, and they were only €.60 in production!”</p>
</li>
<li>
<p><span class="caps">EA</span>: “Please donate to our charity, <span class="caps">AMF</span>. For a donation of €5 or more, you can get button as a token of our gratitude. It can always remind you of the value of good deeds.”<br>
Donor: “Awesome! Here, have €10!”</p>
</li>
</ol>
<p>The proper and sustainable solution would be to present the Child in the Pond analogy to them, show them effective altruism, and then have them donate because they want to save people from malaria, but that will only work for a small fraction of your clients, and if you have several people per minute coming to your <span class="caps">EA</span> booth, you won’t have the time to even try. Rather you can hope that they’ll later reflect “Wow, I donated €10 for a cheap (but pretty!) button, I must be an altruist. Or I like that charity a lot. Let’s have a look at what they’re writing on their blog.”</p>
<p>But in this case I would blame the medium rather than the message since this ad-hoc manner of soliciting investments is highly exploitable anyway.</p>
<h3 id="it-makes-budgeting-hard"><a class="toclink" href="#it-makes-budgeting-hard">It Makes Budgeting Hard</a></h3>
<p>One central <a href="http://claviger.net/dissociation-for-altruists.html"><span class="caps">EA</span> coping strategy</a> is to have a separate donation budget and worry about the counterfactual use of the rest of your money only once a year. Without the categorical separation of donations, we’ll have to think of some other manner of budgeting.</p>
<p>One solution may be different classes of risk-aversity. One low-risk class may be dedicated to GiveWell- or <span class="caps">ACE</span>-recommended charities, another to metacharities or endeavors as Open Phil might evaluate, and another high-risk class to yourself, an intervention as 80,000 Hours might evaluate.</p>
<p>You can think of the same classes also in terms of how long-term the investments are, what their volatility is, or what the resulting util liquidity is. I’ll stick to two such budgets, but maybe there are advantages to introducing finer graduations.</p>
<div class="footnote">
<hr>
<ol>
<li id="fn:1">
<p>I prefer to think of “effective” as meaning that we focus on the effect rather than that we achieve it by means of high cost-effectiveness, since it should then rather be called “efficient altruism.” <a class="footnote-backref" href="#fnref:1" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
</ol>
</div>My Cause Selection: Denis Drescher2015-09-02T12:00:00+00:002015-09-02T12:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2015-09-02:/my-cause-selection-denis-drescher.html<p>This is my contribution to the <a href="http://effective-altruism.com/ea/mh/ea_blogging_carnival_my_cause_selection/"><span class="caps">EA</span> blogging carnival</a> on cause selection. I compare cause areas and attempt a quantitative comparison between <span class="caps">LLIN</span> distributions and advocacy for farmed animals. In short, I will continue to fundraise for the first but but personally donate more to prioritization research within the latter area.</p><div id="pec-encrypted-content" style="display:none">x9FM5km71Lw1Sjnzx0e9XA==;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;^</div>
<div id="pec-decrypted-content">
<h4><i></i></h4>
</div>
<form id="pec-decrypt-form">
<p>
Old articles can be embarrassing. If you would like to access them anyway, you may
<a href="https://bit.ly/3jPN8tT" target="_blank">request access here</a>. Please indicate
who you are in case I don’t know you or don’t recognize you from your email address.
</p>
<input type="password" id="pec-content-password" placeholder="Password" />
<button type="submit" id="pec-decrypt-content">Decrypt</button>
</form>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/core.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/enc-base64.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/cipher-core.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/pad-nopadding.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/md5.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/aes.js"></script>
<script type="text/javascript">
(function () {
var strip_padding = function (padded_content, padding_char) {
/*
* Strips the padding character from decrypted content.
*/
for (var i = padded_content.length; i > 0; i--) {
if (padded_content[i - 1] !== padding_char) {
return padded_content.slice(0, i);
}
}
};
var decrypt_content = function (password, iv_b64, ciphertext_b64, padding_char) {
/*
* Decrypts the content from the ciphertext bundle.
*/
var key = CryptoJS.MD5(password),
iv = CryptoJS.enc.Base64.parse(iv_b64),
ciphertext = CryptoJS.enc.Base64.parse(ciphertext_b64),
bundle = {
key: key,
iv: iv,
ciphertext: ciphertext
};
var plaintext = CryptoJS.AES.decrypt(bundle, key, { iv: iv, padding: CryptoJS.pad.NoPadding });
try {
return strip_padding(plaintext.toString(CryptoJS.enc.Utf8), padding_char);
}
catch (err) {
// encoding failed; wrong password
console.log(err);
return false;
}
};
var init_decryptor = function () {
var decrypt_btn = document.getElementById('pec-decrypt-content'),
password_input = document.getElementById('pec-content-password'),
encrypted_content = document.getElementById('pec-encrypted-content'),
decrypted_content = document.getElementById('pec-decrypted-content'),
decrypt_form = document.getElementById('pec-decrypt-form');
decrypt_btn.addEventListener('click', function () {
// grab the ciphertext bundle
var parts = encrypted_content.innerHTML.split(';');
// decrypt it
var content = decrypt_content(
password_input.value,
parts[0],
parts[1],
parts[2]
);
if (content) {
// success; display the decrypted content
decrypted_content.innerHTML = content;
decrypt_form.parentNode.removeChild(decrypt_form);
encrypted_content.parentNode.removeChild(encrypted_content);
// any post processing on the decrypted content should be done here
}
else {
// ¯\_(ツ)_/¯
password_input.value = '';
}
});
};
document.addEventListener('DOMContentLoaded', init_decryptor);
})();
</script>August 2015 Outreach and Fundraising2015-08-31T13:30:00+00:002015-08-31T13:30:00+00:00Dawn Dreschertag:impartial-priorities.org,2015-08-31:/august-2015-outreach-and-fundraising.html<p>August 2015 has been a busy month for me with three major effective altruism fundraising or outreach events – GalaCon, Czequestria, and the Vegan Summer Festival. Let’s celebrate with numbers!</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#galacon-2015">GalaCon 2015</a><ul>
<li><a href="#donations">Donations</a></li>
<li><a href="#top-items-at-the-table">Top Items at the Table</a></li>
<li><a href="#top-items-in-the-auction">Top Items in the Auction</a></li>
<li><a href="#galacon-and-bronies-for-good-history">GalaCon and Bronies for Good History</a></li>
</ul>
</li>
<li><a href="#czequestria-2015">Czequestria 2015</a><ul>
<li><a href="#donations_1">Donations</a></li>
<li><a href="#top-items-at-the-table_1">Top Items at the Table</a></li>
<li><a href="#top-items-in-the-auction_1">Top Items in the Auction</a></li>
<li><a href="#czequestria-and-bronies-for-good-history">Czequestria and Bronies for Good History</a></li>
</ul>
</li>
<li><a href="#vegan-summer-festival-2015">Vegan Summer Festival 2015</a><ul>
<li><a href="#donations_2">Donations</a></li>
<li><a href="#votes">Votes</a></li>
<li><a href="#top-items-at-the-table_2">Top Items at the Table</a></li>
</ul>
</li>
</ul>
</div>
<div class="admonition type">
<p class="admonition-title">Note</p>
<p>The posts on GalaCon and Czequestria were previously published on the website of my charity <a href="http://yoursiblings.org/blog/">Your Siblings</a>, so the tenses may be a bit confusing.</p>
</div>
<h2 id="galacon-2015"><a class="toclink" href="#galacon-2015">GalaCon 2015</a></h2>
<p><img src="https://yoursiblings.org/media/photologue/photos/galacon-2015.jpg" alt="GalaCon 2015" style="width: 500px; display: block; margin: 1em auto;" /></p>
<p><a href="http://galacon.eu/">GalaCon 2015</a> was an unprecedented success for us, a success that we’ve transferred to the Against Malaria Foundation right away. </p>
<p>“Us” is Bronies for Good, Your Siblings, and the two of us, Nino and Denis, who have run the charity booth. But none of this would’ve been possible without our monetary and in-kind donors, and the tireless work of the staff and volunteers of GalaCon as well as former staff, whose ideas and initiative continues to boost our fundraising. Many thanks also to <a href="http://www.stabletecstudios.com/">Stable-Tec Studios</a> for donating their complete profits to the cause.</p>
<p>Considering only the effects on mortality and only the effects on children under five (who are most likely to die from malaria) and not considering costs that the Against Malaria Foundation leverages from other funders, the €20,027 we raised will buy at least 5,500 mosquito nets and will save the lives of over 10 people. (Based on research by <a href="http://www.givewell.org/international/top-charities/AMF#Whatdoyougetforyourdollar">GiveWell</a>.)</p>
<p>We sold out of <em>Bound Together</em> at GalaCon, but we’ll have more at <a href="http://czequestria.cz/">Czequestria</a>!</p>
<h3 id="donations"><a class="toclink" href="#donations">Donations</a></h3>
<table style="width: auto; text-align: right; border-collapse: collapse; margin: 1em auto;">
<tbody>
<tr>
<th style="text-align: right;">Table Saturday</th>
<td style="text-align: right;">€3,789.88</td>
</tr>
<tr>
<th style="text-align: right;">Table Sunday</th>
<td style="text-align: right;">€1,851.51</td>
</tr>
<tr>
<th style="text-align: right;">Small Auction</th>
<td style="text-align: right;">€1,500.00</td>
</tr>
<tr>
<th style="text-align: right;">Main Auction</th>
<td style="text-align: right;">€11,970.00</td>
</tr>
<tr>
<th style="text-align: right;">Stable-Tec Studios</th>
<td style="text-align: right;">€915.00</td>
</tr>
<tr>
<th style="text-align: right;">Counting error</th>
<td style="text-align: right;">€1.02</td>
</tr>
<tr style="border-top: 1px solid black;">
<th style="text-align: right;"></th>
<td><strong>€20,027.41</strong></td>
</tr>
</tbody>
</table>
<h3 id="top-items-at-the-table"><a class="toclink" href="#top-items-at-the-table">Top Items at the Table</a></h3>
<table style="width: auto; text-align: right; border-collapse: collapse; margin: 1em auto;">
<tbody>
<tr>
<th style="text-align: right;">Bound Together</th>
<td style="text-align: right;"> 95</td>
</tr>
<tr>
<th style="text-align: right;"><span class="caps">CMC</span> button</th>
<td style="text-align: right;"> 69</td>
</tr>
<tr>
<th style="text-align: right;">BfG button</th>
<td style="text-align: right;"> 61</td>
</tr>
<tr>
<th style="text-align: right;">Metal pin</th>
<td style="text-align: right;"> 44</td>
</tr>
<tr>
<th style="text-align: right;"><span class="caps">EA</span> Handbook</th>
<td style="text-align: right;"> 11</td>
</tr>
</tbody>
</table>
<p>We also distributed 435 of our charity guides.</p>
<h3 id="top-items-in-the-auction"><a class="toclink" href="#top-items-in-the-auction">Top Items in the Auction</a></h3>
<table style="width: auto; text-align: right; border-collapse: collapse; margin: 1em auto;">
<tbody>
<tr>
<th style="text-align: right;">GalaCon staff banner</th>
<td style="text-align: right;"> €2,500</td>
</tr>
<tr>
<th style="text-align: right;">Perry’s hat</th>
<td style="text-align: right;"> €810</td>
</tr>
<tr>
<th style="text-align: right;">GalaCon flag</th>
<td style="text-align: right;"> €800</td>
</tr>
<tr>
<th style="text-align: right;">Funko Twilight signed by Lauren and Tara</th>
<td style="text-align: right;"> €720</td>
</tr>
<tr>
<th style="text-align: right;">Fender signed by all guests of honor</th>
<td style="text-align: right;"> €600</td>
</tr>
</tbody>
</table>
<h3 id="galacon-and-bronies-for-good-history"><a class="toclink" href="#galacon-and-bronies-for-good-history">GalaCon and Bronies for Good History</a></h3>
<table style="width: auto; text-align: right; border-collapse: collapse; margin: 1em auto;">
<tbody>
<tr>
<th style="text-align: right;"></th>
<th style="text-align: right;">GalaCon 2015</th>
<th style="text-align: right;">GalaCon 2014</th>
<th style="text-align: right;">GalaCon 2013</th>
<th style="text-align: right;">GalaCon 2012</th>
</tr>
<tr>
<th style="text-align: right;">Estimated attendance</th>
<td style="text-align: right;">1,300</td>
<td style="text-align: right;">1,200</td>
<td style="text-align: right;">1,000</td>
<td style="text-align: right;">400</td>
</tr>
<tr>
<th style="text-align: right;">Highest auction bids</th>
<td style="text-align: right;">€2,500 (banner)</td>
<td style="text-align: right;">
€1,720 (banner)
</td>
<td style="text-align: right;">€1,100 (banner)</td>
<td style="text-align: right;">€1,650 (figurine)</td>
</tr>
<tr>
<th style="text-align: right;">Totals of all auctions</th>
<td style="text-align: right;">€13,470</td>
<td style="text-align: right;">€9,455</td>
<td style="text-align: right;">€11,632</td>
<td style="text-align: right;">€7,545</td>
</tr>
<tr>
<th style="text-align: right;">Totals of direct donations</th>
<td style="text-align: right;">€6,556</td>
<td style="text-align: right;">€2,923</td>
<td style="text-align: right;">€2,694</td>
<td style="text-align: right;">€274</td>
</tr>
</tbody>
</table>
<h2 id="czequestria-2015"><a class="toclink" href="#czequestria-2015">Czequestria 2015</a></h2>
<p><img src="https://yoursiblings.org/media/photologue/photos/2_Czequestria_Header_draft10.png" alt="Czequestria 2015" style="width: 488px; display: block; margin: 1em auto;" /></p>
<p>Last year’s Czequestria was a life-changing experience for me. This year’s <a href="http://www.czequestria.cz/en/">Czequestria</a> will become a life-changing experience for hundreds of people who will be spared an infection with malaria.</p>
<p>This time the table was run mostly by Sara and me, but during the time that Ashley took over, supported by catra, they raised significant koruny too.</p>
<p>Again <a href="http://www.stabletecstudios.com/">Stable-Tec Studios</a> will contribute their complete profits to our invaluable cause. Thank you!</p>
<p>The total of almost €4,000 is enough to save about two lives in expectation and avert easily 300 times as many nonfatal cases of malaria that would’ve meant treatment costs and lost income for many of the poorest of the poor.</p>
<h3 id="donations_1"><a class="toclink" href="#donations_1">Donations</a></h3>
<p>Czequestria and Your Siblings are transferring the donations we raised to the Against Malaria Foundation in several batches. Since all transfer that involve currency conversions to euro are complete, the counts below can now be viewed as authoritative.</p>
<p>Assuming that Czequestria had in the area of 260 attendees, this puts the contribution per attendee in the €15 area, roughly the same as at GalaCon. You rock!</p>
<table style="width: auto; text-align: right; border-collapse: collapse; margin: 1em auto;"><colgroup><col width="100"><col width="100"><col width="100"><col width="100"></colgroup>
<tbody>
<tr>
<th style="text-align: right;">Donations</th>
<th style="text-align: right;"><span class="caps">EUR</span></th>
<th style="text-align: right;"><span class="caps">CZK</span></th>
<th style="text-align: right;"><span class="caps">USD</span></th>
</tr>
<tr>
<th style="text-align: right;">Table Saturday</th>
<td style="text-align: right;">€ 270.00</td>
<td style="text-align: right;">Kč 12,536</td>
<td style="text-align: right;">$5</td>
</tr>
<tr>
<th style="text-align: right;">Table Sunday</th>
<td style="text-align: right;">€ 25.00</td>
<td style="text-align: right;">Kč 7,676</td>
<td style="text-align: right;">$ 95</td>
</tr>
<tr>
<th style="text-align: right;">Stable-Tec</th>
<td style="text-align: right;">€ 80.00</td>
<td style="text-align: right;">Kč 4,075</td>
<td style="text-align: right;">$ 0</td>
</tr>
<tr>
<th style="text-align: right;">Auction</th>
<td style="text-align: right;">€ 0.00</td>
<td style="text-align: right;">Kč 67,677</td>
<td style="text-align: right;">$ 0</td>
</tr>
<tr style="border-top: 1px solid;">
<th style="text-align: right;"></th>
<td style="text-align: right;">€ 375.00</td>
<td style="text-align: right;">Kč 91,964</td>
<td style="text-align: right;">$ 100</td>
</tr>
</tbody>
</table>
<p>Finally the counts after conversions. Here the Stable-Tec donation counts into our table donations. Additional fuzziness is the result of different conversion rates charged for different transfers, and with one transfer a bank fee that Jamis generously paid for privately with an offset.</p>
<table style="width: auto; text-align: right; border-collapse: collapse; margin: 1em auto;"><colgroup><col width="100"><col width="100"><col width="100"><col width="100"><col width="100"></colgroup>
<tbody>
<tr>
<th style="text-align: right;">Donations</th>
<th style="text-align: right;"><span class="caps">EUR</span></th>
<th style="text-align: right;"><span class="caps">CZK</span> in <span class="caps">EUR</span></th>
<th style="text-align: right;"><span class="caps">USD</span> in <span class="caps">EUR</span></th>
<th style="text-align: right;">Sum (<span class="caps">EUR</span>)</th>
</tr>
<tr>
<th style="text-align: right;">Table</th>
<td style="text-align: right;">€ 375.00</td>
<td style="text-align: right;">€ 898.52</td>
<td style="text-align: right;">€ 87.81</td>
<td style="text-align: right;">€ 1,361.33</td>
</tr>
<tr>
<th style="text-align: right;">Auction</th>
<td style="text-align: right;">€ 0.00</td>
<td style="text-align: right;">€ 2,503.74</td>
<td style="text-align: right;">€ 0</td>
<td style="text-align: right;">€ 2,503.74</td>
</tr>
<tr style="border-top: 1px solid;">
<th style="text-align: right;"></th>
<td style="text-align: right;">€ 375.00</td>
<td style="text-align: right;">€ 3,402.26</td>
<td style="text-align: right;">€ 87.81</td>
<td style="text-align: right;">€ 3,865.07</td>
</tr>
</tbody>
</table>
<h3 id="top-items-at-the-table_1"><a class="toclink" href="#top-items-at-the-table_1">Top Items at the Table</a></h3>
<table style="width: auto; text-align: right; border-collapse: collapse; margin: 1em auto;">
<tbody>
<tr>
<th style="text-align: right;">Bound Together</th>
<td style="text-align: right;"> 31</td>
</tr>
<tr>
<th style="text-align: right;"><span class="caps">CMC</span> button</th>
<td style="text-align: right;"> 16</td>
</tr>
<tr>
<th style="text-align: right;">Metal pin</th>
<td style="text-align: right;"> 15</td>
</tr>
<tr>
<th style="text-align: right;"><span class="caps">EA</span> Handbook</th>
<td style="text-align: right;"> 3</td>
</tr>
<tr>
<th style="text-align: right;">BfG button</th>
<td style="text-align: right;"> 8</td>
</tr>
</tbody>
</table>
<p>We (especially Ashley) also distributed 66 of our charity guides.</p>
<h3 id="top-items-in-the-auction_1"><a class="toclink" href="#top-items-in-the-auction_1">Top Items in the Auction</a></h3>
<table style="width: auto; text-align: right; border-collapse: collapse; margin: 1em auto;">
<tbody>
<tr>
<th>Perry’s hat</th>
<td>Kč 10,777</td>
</tr>
<tr>
<th>Fallout Equestria book</th>
<td>Kč 4,500</td>
</tr>
<tr>
<th>Plush Discord signed by de Lancie</th>
<td>Kč 4,200</td>
</tr>
<tr>
<th>Libussa sculpture</th>
<td>Kč 4,000</td>
</tr>
<tr>
<th><span class="caps">CCG</span> signed by de Lancie and St. Germain</th>
<td>Kč 3,000</td>
</tr>
</tbody>
</table>
<h3 id="czequestria-and-bronies-for-good-history"><a class="toclink" href="#czequestria-and-bronies-for-good-history">Czequestria and Bronies for Good History</a></h3>
<table style="width: auto; text-align: right; border-collapse: collapse; margin: 1em auto;">
<tbody>
<tr>
<th style="text-align: right;"></th>
<th style="text-align: right;">Czequestria 2015</th>
<th style="text-align: right;">Czequestria 2014</th>
</tr>
<tr>
<th style="text-align: right;">Estimated attendance</th>
<td style="text-align: right;">260</td>
<td style="text-align: right;">260</td>
</tr>
<tr>
<th style="text-align: right;">Highest auction bids</th>
<td style="text-align: right;">€ 399 (Perry’s hat)</td>
<td style="text-align: right;">€ 162 (signed T-shirt)</td>
</tr>
<tr>
<th style="text-align: right;">Totals of auction</th>
<td style="text-align: right;">€ 2,486</td>
<td style="text-align: right;">€ 1,657</td>
</tr>
<tr>
<th style="text-align: right;">Totals of direct donations</th>
<td style="text-align: right;">€ 1,561</td>
<td style="text-align: right;">€ 346</td>
</tr>
</tbody>
</table>
<h2 id="vegan-summer-festival-2015"><a class="toclink" href="#vegan-summer-festival-2015">Vegan Summer Festival 2015</a></h2>
<p><img src="/images/august-2015-outreach-and-fundraising/vegan-summer-festival.png" alt="Vegan Summer Festival" style="width: 250px; display: block; margin: 1em auto;" /></p>
<p>For the first time we also had a booth at the <a href="http://www.veganes-sommerfest-berlin.de/en/">Vegan Summer Festival 2015</a> in Berlin. We had a booth on two of the three days of the festival, Friday and Sunday, and a talk on Saturday, which I held. On Friday Jan was my partner at the booth, on Sunday Laura and Ise assisted.</p>
<p>The focus of this event was on effective altruism outreach, so we didn’t expect to raise anything close to the amounts raise at the conventions. Unlike at the conventions, we also conducted a giving game, so that donors could vote on which charity should get their donation.</p>
<h3 id="donations_2"><a class="toclink" href="#donations_2">Donations</a></h3>
<table style="width: auto; text-align: right; border-collapse: collapse; margin: 1em auto;">
<tbody>
<tr>
<th style="text-align: right;">Table Friday</th>
<td style="text-align: right;">€149.87</td>
</tr>
<tr>
<th style="text-align: right;">Table Saturday</th>
<td style="text-align: right;">€60.00</td>
</tr>
<tr>
<th style="text-align: right;">Table Sunday</th>
<td style="text-align: right;">€107.50</td>
</tr>
<tr style="border-top: 1px solid black;">
<th style="text-align: right;"></th>
<td style="text-align: right;"><strong>€317.37</strong></td>
</tr>
</tbody>
</table>
<h3 id="votes"><a class="toclink" href="#votes">Votes</a></h3>
<table style="width: auto; text-align: right; border-collapse: collapse; margin: 1em auto;">
<tbody>
<tr>
<th style="text-align: right;"></th>
<th style="text-align: right;">Friday</th>
<th style="text-align: right;">Saturday</th>
<th style="text-align: right;">Sunday</th>
<th style="text-align: right;">Total</th>
</tr>
<tr>
<th style="text-align: right;">1. Against Malaria Foundation</th>
<td style="text-align: right;">1</td>
<td style="text-align: right;">0</td>
<td style="text-align: right;">2</td>
<td style="text-align: right;"><strong>3</strong></td>
</tr>
<tr>
<th style="text-align: right;">2. Animal Equality</th>
<td style="text-align: right;">6</td>
<td style="text-align: right;">5</td>
<td style="text-align: right;">6</td>
<td style="text-align: right;"><strong>15</strong></td>
</tr>
<tr>
<th style="text-align: right;">3. Evidence Action</th>
<td style="text-align: right;">0</td>
<td style="text-align: right;">0</td>
<td style="text-align: right;">0</td>
<td style="text-align: right;"><strong>0</strong></td>
</tr>
<tr>
<th style="text-align: right;">4. GiveDirectly</th>
<td style="text-align: right;">3</td>
<td style="text-align: right;">0</td>
<td style="text-align: right;">1</td>
<td style="text-align: right;"><strong>4</strong></td>
</tr>
<tr>
<th style="text-align: right;">5. Mercy For Animals</th>
<td style="text-align: right;">1</td>
<td style="text-align: right;">0</td>
<td style="text-align: right;">1</td>
<td style="text-align: right;"><strong>2</strong></td>
</tr>
<tr>
<th style="text-align: right;">6. Schistosomiasis Control Initiative</th>
<td style="text-align: right;">3</td>
<td style="text-align: right;">0</td>
<td style="text-align: right;">0</td>
<td style="text-align: right;"><strong>3</strong></td>
</tr>
<tr>
<th style="text-align: right;">7. The Humane League</th>
<td style="text-align: right;">0</td>
<td style="text-align: right;">0</td>
<td style="text-align: right;">0</td>
<td style="text-align: right;"><strong>0</strong></td>
</tr>
<tr>
<th style="text-align: right;">A. GiveWell</th>
<td style="text-align: right;">1</td>
<td style="text-align: right;">1</td>
<td style="text-align: right;">0</td>
<td style="text-align: right;"><strong>2</strong></td>
</tr>
<tr>
<th style="text-align: right;">B. Animal Charity Evaluators</th>
<td style="text-align: right;">0</td>
<td style="text-align: right;">0</td>
<td style="text-align: right;">0</td>
<td style="text-align: right;"><strong>0</strong></td>
</tr>
</tbody>
</table>
<p>One ballot card was invalid. (It said “5.A.”)</p>
<p>Animal Equality will receive its donation presently.</p>
<h3 id="top-items-at-the-table_2"><a class="toclink" href="#top-items-at-the-table_2">Top Items at the Table</a></h3>
<table style="width: auto; text-align: right; border-collapse: collapse; margin: 1em auto;">
<tbody>
<th style="text-align: right;"><span class="caps">EA</span> Handbook</th>
<td style="text-align: right;">29</td>
</tr>
<tr>
<th style="text-align: right;">Singer’s <span class="caps">TMGYCD</span></th>
<td style="text-align: right;">5</td>
</tr>
</tbody>
</table>
<p>We also distributed 269 of our latest flyers. (And 20 older ones.)</p>Results of the Effective Altruism Outreach Survey2015-07-26T13:30:00+00:002015-07-26T13:30:00+00:00Dawn Dreschertag:impartial-priorities.org,2015-07-26:/effective-altruism-outreach-survey.html<p>This article reports the results of an online survey with 167 respondents on the influence different styles of effective altruism outreach have on them. While we could not find evidence for our hypotheses, the exploratory data analysis yielded a ranking of the levels of motivation and curiosity our prompts induced.</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#topic">Topic</a></li>
<li><a href="#design-and-implementation">Design and Implementation</a></li>
<li><a href="#analysis">Analysis</a></li>
<li><a href="#evaluation">Evaluation</a><ul>
<li><a href="#main-hypotheses">Main Hypotheses</a></li>
<li><a href="#qualitative-results">Qualitative Results</a></li>
</ul>
</li>
<li><a href="#conclusion">Conclusion</a></li>
<li><a href="#footnotes">Footnotes</a></li>
</ul>
</div>
<h2 id="topic"><a class="toclink" href="#topic">Topic</a></h2>
<p>The aim of our survey was to determine what form of effective altruism outreach was most effective for what type of audience.</p>
<p>As types of outreach, we distinguished:</p>
<ol>
<li>the “obligation style,” which aims to reveal altruistic values in people by helping them overcome biases, a style that is epitomized by <a href="https://www.youtube.com/watch?v=Diuv3XZQXyc">Peter Singer</a>’s Child in the Pond analogy,<sup id="fnref:peter-singer-disclaimer"><a class="footnote-ref" href="#fn:peter-singer-disclaimer">1</a></sup> and</li>
<li>the “opportunity style,” which assumes that people are altruistic and helps them overcome biases that keep them locked in lethargy anyway, a style that is epitomized by <a href="https://www.youtube.com/watch?v=ZGAkrpwyu1k">Toby Ord</a>’s appeal that people can save hundreds of lives over their lifetime if they invest their money wisely.</li>
</ol>
<p>Styles that we did not investigate are the usage of humor to better convey topics that would otherwise be met by defensiveness (suggested by <a href="https://www.againstmalaria.com/People.aspx">Rob Mather</a>) and a style that is similar to the opportunity style but puts a stronger emphasis on personal discovery, as in <a href="https://www.youtube.com/watch?v=o0VrZPBskpg">Melanie Joy</a>’s <span class="caps">TED</span> talk.</p>
<p>Such an evaluation could help any group engaged in effective altruism outreach to communicate more effectively with their respective audiences.</p>
<p>Our hypotheses were:</p>
<ol>
<li>The obligation style leads to defensiveness, which would effect negative reaction at least in the short term and at least from less rationally-minded people. (If it is also more emotionally salient, later reflection might still make it more effective, but we cannot measure that.)</li>
<li>The opportunity style has a positive effect but only on people who already show a strong altruistic inclination.</li>
<li>Pitches targeted at specific demographics have a stronger effect on these people than on others.</li>
</ol>
<p>On the exploratory side, we were also interested in the correlation between rational inclination and respondents’ trust in their intuition, and their attitude toward our prompts, as well as any correlation between respondents’ reaction to the prompts and the degree to which the prompts informed them or withheld information, as teasers do.</p>
<p>Since we could not find evidence of these correlations, it would be interesting to see whether others can. Additionally, there are a number of prompts that seem very powerful that we did not include (e.g., a comparison of prioritization with triage). A different sample might be more representative of the taxonomy. A qualitative study might also shed more light on the way people react to our prompts.</p>
<h2 id="design-and-implementation"><a class="toclink" href="#design-and-implementation">Design and Implementation</a></h2>
<p>One of our worries was that if obligation-style prompts really make people defensive, then there is the risk of this defensiveness coloring the responses to later prompts. Hence we introduced a page break and sorted the critical prompts to the second page of the two.</p>
<p>The length of the prompts, especially the one’s borrowed from Peter Singer, was another problem. We slightly shortened them where possible and otherwise reduced the number of prompts from originally eight per category to five. In the interest of reducing the number of fields people have to tick, we removed a scale for how much people like a prompt, which we found dispensable.</p>
<p>To measure rational and experiential (intuition-related) proclivities, we relied on the Rational-Experiential Inventory (<span class="caps">REI</span>) with 10 prompts by Norris, Pacini, and Epstein (1998). To measure altruistic inclination, we selected 10 prompts from the Adapted Self-Report Altruism Scale by Rushton (original, 1981), Witt and Boleman (adapted version, 2009).</p>
<p>Just as these two scales, our prompts also relied on five-item Likert scales.</p>
<p><a href="http://claviger.net/assets/outreach-survey.html">The full survey and recruitment letter can be found here.</a></p>
<p>We at first used the original Rushton scale, but after receiving 15 responses switched to the modified one, which meant turning sentences from present perfect into conditional (“I have donated blood” became “I would donate blood.”). The change is fairly localized, the responses obtained after the change greatly outnumber those obtained before, and we did not see any noticeable differences in the spread of the answers, so we decided to include the first 15 in our final analysis.</p>
<p>We advertised the survey on Reddit, Twitter, and Facebook, also using paid advertisement on Facebook to reach more people. Most, however, were recruited through an email a friend send to a mailing list of the Humboldt-Universität zu Berlin. We tried to counterbalance and get more people without academic background into our sample by targeting younger people on Facebook, but we only recruited only about 26 people that way (at a rate of almost €1 per person), as opposed to 85 via the mailing list.</p>
<p>Please contact us if you would like to play around with the raw data.</p>
<h2 id="analysis"><a class="toclink" href="#analysis">Analysis</a></h2>
<p>Our R script for cleaning and analysis can be found in this <a href="https://bitbucket.org/snippets/Telofy/Mb4pE">Bitbucket snippet</a>.</p>
<p>After a first section of type conversions and reversal of questions that were asked in the negative for validation purposes, we engaged in the controversial practice of interpreting the ordinal Likert items as interval scale to compute means. This would imply that the differences between the five options we gave are identical. We have no basis for this assumption, and the results should be taken with the appropriate absolute-scale number of grains of salt.</p>
<p>Apart from more cleaning, we also combined answers into categories that seem intuitive enough to us to not be motivated by the data. However, we have seen the data before deciding on the categories in all cases except for political views. The intervals used for the respondents age are not ours but intervals often used in the literature. These coarser categories allowed us to compensate for the low sample sizes per cohort.</p>
<p>Finally the script produces some eighty graphs.</p>
<p>When the analyses showed that we could find evidence for none of our hypotheses, we engaged in exploratory data analysis, the results of which are detailed in the following.</p>
<h2 id="evaluation"><a class="toclink" href="#evaluation">Evaluation</a></h2>
<p><img alt="Full ranking" src="/images/effective-altruism-outreach-survey/all.png"></p>
<p>Exploratory data analysis has the inevitable drawback that in all likelihood we’ll find significant-looking correlations in our data simply by chance.</p>
<p>Nonetheless the overall ranking of the prompts, prompts that we asked our respondents to rate along scales of curiosity and motivation they either induced or failed to induce, has the power of our full sample size of 167 behind it, so that we’re somewhat confident that conclusions drawn about prompts close to its extreme points are valuable.</p>
<p>The graph above shows the distribution of respondents’ votes with the prompts described by a key where the first is a keyword that makes clear which prompt is meant, the second part is our taxonomy of whether the prompt focuses on the donor’s opportunity or moral obligation, the third part is either “info” or “teaser” depending on whether the prompt explains something or withholds information, and the fourth part indicates whether the respondent gauged their motivation or their curiosity. The first and last part are restrictive while the second and third are descriptive.</p>
<p>There are also some post-hoc rationalizations that make the rankings of the top prompts plausible.</p>
<p>The absolute top prompt in terms of motivation and curiosity is Peter Singer’s famous Child in the Pond analogy, which would probably not have made it into our survey had it not proved its persuasive power by turning Singer’s essay “Famine, Affluence, and Morality” into a seminal paper of moral philosophy well-known to philosophers worldwide.</p>
<p>The third place is a slightly adapted version of the sentence that Giving What We Can uses as one of their slogans, “Studies have found that top charities are up to a thousand times more effective than others,” except that the organization omits the weasel words “studies have found.” It is also a time-tested prompt.</p>
<p>The fourth place is an almost verbatim quote from Toby Ord’s <span class="caps">TED</span> talk and surely a statement that the Giving What We Can founder has honed in hundreds of conversations with potential pledge-takers: “You can save someone’s life without even changing your career.”</p>
<p>The final spots in the ranking can be explained as an aversive reaction to an insulting prompt. Interestingly, the rather popular prompt comparing the training of a guide dog to sight-restoring surgery ranks very low in terms of the motivation it induces.</p>
<p>Threats to our external validity are that we have in our sample:</p>
<ul>
<li>3.7 times as many academics than people who only graduated school if you count as academics anyone who has visited a university or college irrespective of whether they’ve attained a degree yet,</li>
<li>3 times as many nonreligious than religious respondents, and
a mean age of 25 (σ=7) with only two respondents over 45.</li>
</ul>
<p>There are likely more biases that we can’t recognize.</p>
<h3 id="main-hypotheses"><a class="toclink" href="#main-hypotheses">Main Hypotheses</a></h3>
<p><img alt="Scatter plots" src="/images/effective-altruism-outreach-survey/all.mean.matrix.png"></p>
<p>In our data exploration, we have generated over eighty graphs that can be found in <a href="https://drive.google.com/open?id=0B_EGIYnWjWAvfmRnTGJWdWhVTzVBUl93TnRaQWM5NS1DX3ctWXpsWXVlM3J0cXBqN1h2YmM">this gallery</a>.</p>
<p>Based on experiences in the Less Wrong community and <a href="http://reg-charity.org/"><span class="caps">REG</span></a>’s experiences with poker players as well as our inside view of the effective altruism movement itself, we expected to see a clear correlation between rationality and effective altruism inclination (the “all” vs. “rational” plots above).</p>
<p>We did not expect to see such a clear correlation with our data on the respondents’ altruism inclination (the “all” vs. “altruistic” plots above), because it tested very elementary, naive empathetic skill, which may be necessary to a degree but is otherwise unhelpful for understanding effective altruism.</p>
<p>Neither correlation showed. Not even the square root of the product of the two features was correlated with responses to our prompts. If these results can be taken at face value, then it seems to us that rationality and altruism may be little more than necessary conditions for becoming effective altruism, and that something else is just as necessary—maybe the principle of “taking ideas seriously,” which is common on Less Wrong, or any number of other such traits. But the results are probably more likely to be meaningless.</p>
<p>The strong correlations between “altruistic” and the two <span class="caps">REI</span> dimensions may be just artifacts of people’s different inclinations to answer Likert scales with extreme or moderate values. Surprisingly, however, the same tendency is not evident between the two <span class="caps">REI</span> dimensions. Perhaps they are sufficiently contradictory to offset this tendency. Please let us know when you have other explanations.</p>
<h3 id="qualitative-results"><a class="toclink" href="#qualitative-results">Qualitative Results</a></h3>
<p>The only nonquantitative question in our survey was the one asking for comments and suggestions. A few interesting comments:</p>
<ol>
<li>One respondent made the good point that the questions that focus on opportunities in effective altruism put the donor at the center rather than the beneficiary, something that to change is a crucial part of effective altruism.</li>
<li>Five respondents made suggestions that seemed to go in the opposite direction (though that is my interpretation), largely for pragmatic purposes. Two of them seemed take this position despite seeming fairly aware of the privilege of their birth.</li>
<li>One respondent said fairly directly that the distance of suffering was morally relevant to them.</li>
</ol>
<h2 id="conclusion"><a class="toclink" href="#conclusion">Conclusion</a></h2>
<p>While we could not find evidence for our hypotheses, we were able to generate a ranking of prompts commonly used by effective altruists of to how much motivation and curiosity they induce according to self-report. Due to biases in our sample, the external validity of these results is probably higher for populations of academics than the general population.</p>
<h2 id="footnotes"><a class="toclink" href="#footnotes">Footnotes</a></h2>
<div class="footnote">
<hr>
<ol>
<li id="fn:peter-singer-disclaimer">
<p>Disclaimer: I don’t generally endorse the works of the author. Singer originated many useful ideas, so that I can’t help but cite him lest it seem that I plagiarize them. Until around 2016, unfortunately, he seemed to be attracted to controversial ideas or framings for controversiality’s sake. Controversial ideas are unusually often insensitive and weak. They need to be treated with special care to make them as uncontroversial as possible. I’m adding these disclaimers to avoid the impression that I accept such intellectual wantonness or that it is accepted in my circles. <a class="footnote-backref" href="#fnref:peter-singer-disclaimer" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
</ol>
</div>Effective Altruism Flyer2015-07-22T17:30:00+00:002015-07-22T17:30:00+00:00Dawn Dreschertag:impartial-priorities.org,2015-07-22:/effective-altruism-flyer.html<p>A quick post on a flyer I’ve come up with in cooperation with Charity Science (mostly Tom Ash) and that we then had implemented by a professional designer.</p><div id="pec-encrypted-content" style="display:none">ZWTVQ0I3g1sNHopT8C3GEg==;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;^</div>
<div id="pec-decrypted-content">
<h4><i></i></h4>
</div>
<form id="pec-decrypt-form">
<p>
Old articles can be embarrassing. If you would like to access them anyway, you may
<a href="https://bit.ly/3jPN8tT" target="_blank">request access here</a>. Please indicate
who you are in case I don’t know you or don’t recognize you from your email address.
</p>
<input type="password" id="pec-content-password" placeholder="Password" />
<button type="submit" id="pec-decrypt-content">Decrypt</button>
</form>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/core.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/enc-base64.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/cipher-core.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/pad-nopadding.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/md5.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/aes.js"></script>
<script type="text/javascript">
(function () {
var strip_padding = function (padded_content, padding_char) {
/*
* Strips the padding character from decrypted content.
*/
for (var i = padded_content.length; i > 0; i--) {
if (padded_content[i - 1] !== padding_char) {
return padded_content.slice(0, i);
}
}
};
var decrypt_content = function (password, iv_b64, ciphertext_b64, padding_char) {
/*
* Decrypts the content from the ciphertext bundle.
*/
var key = CryptoJS.MD5(password),
iv = CryptoJS.enc.Base64.parse(iv_b64),
ciphertext = CryptoJS.enc.Base64.parse(ciphertext_b64),
bundle = {
key: key,
iv: iv,
ciphertext: ciphertext
};
var plaintext = CryptoJS.AES.decrypt(bundle, key, { iv: iv, padding: CryptoJS.pad.NoPadding });
try {
return strip_padding(plaintext.toString(CryptoJS.enc.Utf8), padding_char);
}
catch (err) {
// encoding failed; wrong password
console.log(err);
return false;
}
};
var init_decryptor = function () {
var decrypt_btn = document.getElementById('pec-decrypt-content'),
password_input = document.getElementById('pec-content-password'),
encrypted_content = document.getElementById('pec-encrypted-content'),
decrypted_content = document.getElementById('pec-decrypted-content'),
decrypt_form = document.getElementById('pec-decrypt-form');
decrypt_btn.addEventListener('click', function () {
// grab the ciphertext bundle
var parts = encrypted_content.innerHTML.split(';');
// decrypt it
var content = decrypt_content(
password_input.value,
parts[0],
parts[1],
parts[2]
);
if (content) {
// success; display the decrypted content
decrypted_content.innerHTML = content;
decrypt_form.parentNode.removeChild(decrypt_form);
encrypted_content.parentNode.removeChild(encrypted_content);
// any post processing on the decrypted content should be done here
}
else {
// ¯\_(ツ)_/¯
password_input.value = '';
}
});
};
document.addEventListener('DOMContentLoaded', init_decryptor);
})();
</script>Dissociation for Altruists2015-05-14T13:30:00+00:002015-05-14T13:30:00+00:00Dawn Dreschertag:impartial-priorities.org,2015-05-14:/dissociation-for-altruists.html<p>Some people do not lack in altruism and are well aware of effectiveness considerations too, but the sheer magnitude of suffering that effective interventions would force them to face is too unbearable for them to acknowledge. I give tips on how they can use dissociation to put altruism on a more scalable basis.</p><div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#permitting-yourself-to-dissociate">Permitting Yourself to Dissociate</a></li>
<li><a href="#dissociation-tips">Dissociation Tips</a></li>
<li><a href="#footnotes">Footnotes</a></li>
</ul>
</div>
<h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<p>Books like Peter Singer’s <em>The Life You Can Save</em> mostly aim to break through people’s compartmentalization to allow them a glimpse of the real importance of problems like poverty. This is one of the most common outreach strategies and it may also be the most widely applicable one. Some of you, however, may be well aware of the importance of the enterprise but are too sensitive to face the suffering that the most effective interventions necessarily bring to mind.</p>
<p>Repeatedly I’ve heard that people think effective altruism is great, but that they would become emotional wrecks if they allowed themselves to care about more than the well-being of their closest friends, Wikipedia, and <span class="caps">SETI</span>@home.</p>
<p>I’ve been there. The suffering in the world is enormous, and all the while even the suffering of a single chicken in industrial agriculture is unfathomable to me. I watched a video of seals being flayed alive once. For weeks I slipped through the world in a haze.</p>
<p>However, when we want to donate time and money to alleviate this suffering effectively, then we have to acknowledge it. My solution to overcome this “squeamishness” has been to not “feel about” the things I think about, a form of dissociation. It was more a decision for me than something I had to train, but if it doesn’t come easy to you, then I’ve collected some tips for you as well. (Peter Singer touches on this in chapter 7, “Is Love all We Need?,” of <a href="http://www.amazon.com/Most-Good-You-Can-Effective/dp/0300180276/ref=sr_1_1?ie=UTF8&qid=1431534189&sr=8-1&keywords=the+most+good+you+can+do&tag=s4charity-20"><em>The Most Good You Can Do</em></a>.<sup id="fnref:peter-singer-disclaimer"><a class="footnote-ref" href="#fn:peter-singer-disclaimer">1</a></sup>)</p>
<h2 id="permitting-yourself-to-dissociate"><a class="toclink" href="#permitting-yourself-to-dissociate">Permitting Yourself to Dissociate</a></h2>
<p>When I still thought that it was virtuous to empathize emotionally, I did a version of <a href="http://lesswrong.com/lw/l30/on_caring/">Nate Soares’s experiment</a> where, time and time again, I invented an individual family to represent the statistical victims of poverty and disease that I knew about. Thus I tried to prove to myself that my brain was not a cold dead place. I was overwhelmed and paralyzed with their suffering. I hadn’t even started to imagine the ten-fold suffering of ten families or the ten million–fold suffering of ten million families, yet my “care-o-meter,” as Nate calls it, was maxed out already. As these emotions preoccupied me and weighed me down, my productivity suffered and it threatened to impact, among other things, my fundraising.</p>
<p>Consequentialism saved me. Empathy doesn’t have to take the form that lets you break down sobbing and clawing the floor, or whatever it looks like for you. In fact our recent advances in transport and communication mean that we need a more advanced form of empathy too. <a href="http://lesswrong.com/lw/l30/on_caring/">As Nate puts it</a>:</p>
<blockquote>
<p>Prominent altruists aren’t the people who have a larger care-o-meter, they’re the people who have <em>learned not to trust their care-o-meters.</em> Our care-o-meters are broken. They don’t work on large numbers. Nobody has one capable of faithfully representing the scope of the world’s problems. But the fact that you can’t <em>feel</em> the caring doesn’t mean that you can’t <em>do</em> the caring. … You can still <em>act</em> like the world’s problems are as big as they are. You can stop trusting the internal feelings to guide your actions and switch over to manual control.</p>
</blockquote>
<p>The realization that there’s nothing virtuous about having these emotions, that all I was doing was trying to electrocute myself with a broken voltmeter, freed me and allowed me to get stuff done again.</p>
<p>Medical students have to learn that too. One of them <a href="http://www.reddit.com/r/medicine/comments/2aiyai/doctorssurgeons_of_reddit_are_you_squeamish/ciwai6t">told themselves</a> “if I don’t dissociate myself from this patient’s pain, then I will not be able to take care of them the way I need to.” Intellectualizing this decision repeatedly in a small pep talk before each surgery helped them shut it off, and only that way they were able to perform as highly as possible. As Nate puts it again:</p>
<blockquote>
<p>It’s a common trope that courage isn’t about being fearless, it’s about being afraid but <em>doing the right thing anyway.</em> In the same sense, caring about the world isn’t about having a gut feeling that corresponds to the amount of suffering in the world, it’s about <em>doing the right thing anyway.</em> Even without the feeling. [Especially without the feeling.]</p>
</blockquote>
<p>There may be other hurdles for you, as there were for me:</p>
<ol>
<li>If you’re worried that your altruism may be selfishly motivated, then that’s just fine so long as it helps the beneficiaries as much as differently motivated forms of altruism would.</li>
<li>If you’re worried that this supposed nonaltruistic motivation may be less lasting than extreme poverty, then put some effort into getting all your finances organized and then <a href="http://givingwhatwecan.org/">file a pledge</a>. A “further” pledge may even be less risky if you anticipate periods of low income.</li>
<li>Finally, sabotage your supposed motivation. Today the selfish motivation I used to suspect seems absurd to me. It ceased to apply at some point, and my enthusiasm for <span class="caps">EA</span> was as great as ever. Q.e.d., fscker!</li>
</ol>
<p>And in case you’re worried, I don’t think that practicing dissociation has impacted my social abilities in any way. On the contrary, I’ve become able to talk about things that used to be so emotional I hardly dared to think about them before.</p>
<h2 id="dissociation-tips"><a class="toclink" href="#dissociation-tips">Dissociation Tips</a></h2>
<p>For me, not feeling was mostly a choice, but there are some precautions I make anyway and some more tips I’ve gleaned from MDs.</p>
<ol>
<li>Get the tools from the abstraction toolbox. It helps to think in terms of DALYs and funding gaps when you read a statistic. That way you can think about it without automatically imagining what it means for the beings it describes.</li>
<li>If it makes you feel safer, imagine, just once, a unit of suffering that you can envision without going to pieces, like a nonfatal episode of malaria, to understand what a sixth of a <span class="caps">DALY</span> means. Once you’ve thus anchored the unit, you can stop thinking about the thing itself altogether and only operate with the abstraction.</li>
<li>Avoid images and videos unless you’re sure you can dissociate sufficiently. Academic prose is safer.</li>
<li>Don’t visit the countries where you’re helping. There is great value in understanding the local conditions better to run better interventions, but the top charities and GiveWell are doing that already so you don’t have to.</li>
<li>Related to that, try not to get to know anyone you might be helping. The decision to donate someplace and not someplace else may consign some to death so others (and hopefully more) can live. It helps to not know either to make better decisions.</li>
<li>A common technique is to create a clear donation budget so that you only have to think about these issues in depth once or a couple times per year.</li>
<li>Donating to a trust like GiveWell’s or Giving What We Can’s might also help increase the abstraction.</li>
<li><a href="http://webcache.googleusercontent.com/search?q=cache:_-Rsy4xMONwJ:www.medscape.com/viewarticle/714098+&cd=1&hl=en&ct=clnk">Sara Cohen</a>, <span class="caps">MD</span>, recommends distraction in addition to dissociation: “When I started feeling queasy, I did a simple trick: I counted to 10. Deep breath, 1-2-3-4-5-6-7-8-9-10.”</li>
<li>Another one of her strategies is to desensitize herself, which she did in a controlled, low-risk fashion with videos. When you feel up for it, you can try that too, carefully.</li>
</ol>
<p>Sara Cohen concludes somewhat self-contradictorily that “like my own advisors, I can now assert that it’s true that everyone eventually gets over it. And if not, there’s always radiology.” There may be some selection bias involved here (due to medical students who can’t dissociate sufficiently changing degrees), and many surgeons also indicate that there are still certain procedures that they can’t perform. But even these surgeons have surely made great progress since their first semesters of medical school, and the combination of dissociation and avoidance that they practice works well for them and can surely also work well for you.</p>
<p>If you have any more of such tips to share, please do so in the comments. I can’t give comprehensive answers that’ll work for everyone, but I hope I can start a discussion that will generate more of them for more of you.</p>
<p>Once you’ve taken this hurdle, articles like Ben Kuhn’s “<a href="http://www.benkuhn.net/stress">To Stressed-Out Altruists</a>” and Lukas Gloor’s “<a href="http://effective-altruism.com/ea/gv/room_for_other_things_how_to_adjust_if_ea_seems/">Room for Other Things</a>” may become helpful to make your impact even more sustainable.</p>
<h2 id="footnotes"><a class="toclink" href="#footnotes">Footnotes</a></h2>
<div class="footnote">
<hr>
<ol>
<li id="fn:peter-singer-disclaimer">
<p>Disclaimer: I don’t generally endorse the works of the author. Singer originated many useful ideas, so that I can’t help but cite him lest it seem that I plagiarize them. Until around 2016, unfortunately, he seemed to be attracted to controversial ideas or framings for controversiality’s sake. Controversial ideas are unusually often insensitive and weak. They need to be treated with special care to make them as uncontroversial as possible. I’m adding these disclaimers to avoid the impression that I accept such intellectual wantonness or that it is accepted in my circles. <a class="footnote-backref" href="#fnref:peter-singer-disclaimer" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
</ol>
</div>Incentivizing Charity Cooperation2015-05-11T14:00:00+00:002015-05-11T14:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2015-05-11:/incentivizing-charity-cooperation.html<p>There is a danger of charities competing when they could cooperate to increase their total impact. I describe the danger and propose a strategy for alleviating it.</p><div class="admonition type">
<p class="admonition-title">Dated Content</p>
<p>I tend to update articles only when I remember their content and realize that I want to change something about it. But I rarely remember it well enough once about two years have passed. Such articles are therefore likely to contain some statements that I no longer espouse or would today frame differently.</p>
</div>
<div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#dangers-of-breeding-competition">Dangers of Breeding Competition</a></li>
<li><a href="#specialized-effectiveness-training">Specialized Effectiveness Training</a></li>
<li><a href="#cost-effectiveness-of-ea-training">Cost-Effectiveness of EA Training</a></li>
<li><a href="#incentivizing-ea-training">Incentivizing EA Training</a></li>
</ul>
</div>
<h2 id="dangers-of-breeding-competition"><a class="toclink" href="#dangers-of-breeding-competition">Dangers of Breeding Competition</a></h2>
<p>I see two success scenarios for effective altruism, one focused more on causes and one more focused on charities: (1) the top effective interventions get sufficient funding, solve the problems they address, new interventions become most effective, and the cycle continues; and (2) top effective charities get funding from EAs and foundations, less effective charities are incentivized to care less about the donor and more about the beneficiary, and their managements gear their interventions toward higher cost-effectiveness. This post focuses on the second scenario.</p>
<p>When I started getting involved with the running of a charity, with fundraising, and with organizing collaborative events, I didn’t know anything about effective altruism (this was in 2010–2011). I already had my intuitions about cost-effectiveness and its drops at the margin in some cases, but by and large I saw most charities as cooperating toward the reduction of suffering. That allowed me to be happy about almost any charity fundraising success no matter whether we had anything to do with it.</p>
<p>When I learned more about effective altruism, I realized how wasteful many charities are and how fundraising may do harm if it directs donations from more effective charities to less effective ones. Suddenly competition appeared where I thought only cooperation should rule.</p>
<p>At our <span class="caps">EA</span> Berlin meetup, we are worried that we’re not doing enough to address this problem. <a href="https://www.youtube.com/watch?v=o0VrZPBskpg">Melanie Joy</a>, who started the discussion, reported that some people involved in animal charities that <span class="caps">ACE</span> deems less effective than its top charities have come to view these top charities as competitors even though they work together toward closely related goals. However, some of these people are actually open to learning how to improve the effectiveness of their charities.</p>
<h2 id="specialized-effectiveness-training"><a class="toclink" href="#specialized-effectiveness-training">Specialized Effectiveness Training</a></h2>
<p>When I’ve been thinking about improving the effectiveness of charities, I always thought along the lines of targeted advocacy, where I would try to find a charity that moves a lot of money and has much room for improvement, find someone in a key position who might be open to improving something, contract them and make a case for a more effective allocation of their resources. This is complicated and cumbersome. In this case, however, there are actually just these people who are in key positions and are open to educating themselves. This is an amazing new perspective!</p>
<p>The problem is just that building up a highly effective charity in a specific cause area, running one or several specific interventions takes a lot of very specialized knowledge. The <span class="caps">EA</span> forum and the blogs of EAs and <span class="caps">EA</span> organizations are unlikely to hold the information that these people will need.</p>
<p>Prioritization organizations could probably give some advice here, but they would become a central hub for this sort of consulting and are probably too resource-constrained to do that. They also need to be generalists, so they might not have the knowledge at quite the level these trainees require. Alternatively, someone could start an <span class="caps">EA</span> McKinsey, but I’m not sure if that could work out financially. Consultants like Caroline Fiennes seem to focus more on advising donors than charities, but please let me know if there are already some that give quality advice to charities.</p>
<p>Rather the top charities themselves could help here. It is in the interest of their transparency anyway to keep their strategies and processes open to the public, for example in blog posts, but they could go beyond that and actually accept in staff members of interested less effective charities and work with them for some time, maybe a month or a year.</p>
<h2 id="cost-effectiveness-of-ea-training"><a class="toclink" href="#cost-effectiveness-of-ea-training">Cost-Effectiveness of <span class="caps">EA</span> Training</a></h2>
<p>The main question is whether such a program could be cost-effective for the top charity and in general, and I’m not satisfied with the answer I can give to that.</p>
<p>First, the costs to the top charity are not comparable with the expenses they would incur through volunteers, interns, or even average new hires. Typically you invest into hiring, onboarding, training, and salaries in the hope that the employee will stay long enough to far outearn this initial investment. In this case, however, the less effective charity would react to a call for applications from the top charity, and the latter only has to review the applications it receives, which already reduces costs. (The interest in such positions will likely increase as effective altruism becomes more popular.) Then, crucially, these are people who already worked in the same cause area for years. Even though their charities’ performances might not be on par with the top charity yet, they are likely to be highly qualified in the same field. Training costs will be minimal. Finally, they may’ve solved problems at their own charities that the top charity still struggles with, so that the flow of experience is not strictly one-sided. All in all, the cost for the program is likely to be smaller than it would at first seem.</p>
<p>Second, these charities are cooperating toward a common goal, so the improvements of the less effective charities also count into the effectiveness side. The more money these charities move (and the lacking recommendation will unfortunately not have a crucial impact on that) and the greater their room for improvement, the greater will be the absolute improvement in impact that the program can achieve (and the more likely it is that they can afford to send an employee to such a training). If the program is very successful, the charities might even merge, which could enhance the effectiveness of both.</p>
<h2 id="incentivizing-ea-training"><a class="toclink" href="#incentivizing-ea-training">Incentivizing <span class="caps">EA</span> Training</a></h2>
<p>It would be valuable to test these intuitions about the cost-effectiveness of such a program against real numbers that I currently don’t have. It would also be interesting to poll how many organizations of what size would already be interested in taking part in such a program at any current top charity, but I would expect that number to increase over time.</p>
<p>If the cost-effectiveness is sufficient, the program could be seen as increasing the effectiveness of the top charity via other charities, an additional argument for donating to them and for the prioritization organization to keep recommending them. That, in turn, would incentivize top charities to conduct such programs.</p>
<p>All the while the program would also thwart division among charities by showing them that <span class="caps">EA</span> is not only about finding the best charities and donating to them at the exclusion of all others but also about lifting those other charities to the same level of effectiveness as the top charities where this is possible.</p>
<p>See the <a href="http://effective-altruism.com/ea/ih/incentivizing_charity_cooperation/"><span class="caps">EA</span> Forum</a> for the discussion.</p>Expected Utility Auctions2015-05-02T16:22:00+00:002018-01-28T13:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2015-05-02:/expected-utility-auctions.html<p>I give an explanation for a phenomenon in the effective altruism community (<a href="https://docs.google.com/presentation/d/1s4uDz7MFTfRl4Cxn7QVKQt3Of0SqLA4Nfrxtx6Tc9IY/edit">related to this presentation</a>) that might look like the streetlight effect, propose an idea for a software that might help to further optimize this area, and ask you for your input.</p><div class="admonition type">
<p class="admonition-title">Dated Content</p>
<p>I tend to update articles only when I remember their content and realize that I want to change something about it. But I rarely remember it well enough once about two years have passed. Such articles are therefore likely to contain some statements that I no longer espouse or would today frame differently.</p>
</div>
<div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#open-philanthropytype-interventions">Open Philanthropy–Type Interventions</a></li>
<li><a href="#expected-utility-and-limited-diversification">Expected Utility and Limited Diversification</a></li>
<li><a href="#expected-utility-auctions">Expected Utility Auctions</a></li>
</ul>
</div>
<h2 id="open-philanthropytype-interventions"><a class="toclink" href="#open-philanthropytype-interventions">Open Philanthropy–Type Interventions</a></h2>
<p>Insightful quantitative analyses give me the warm-fuzzies, but there may well be highly effective interventions, maybe even interventions more cost-effective than GiveWell’s top charities, that are not easily quantifiable. The Open Philanthropy Project has set out to find them, but so far it has not published any hard and fast recommendations. (<a href="https://www.openphilanthropy.org/blog/suggestions-individual-donors-open-philanthropy-project-staff-2017">Though their staff have.</a>)</p>
<p>Meanwhile outsiders seem to have mistaken our enthusiasm for certain more easily quantifiable interventions for “effective altruism is about donating to easily quantifiable interventions” rather than “effective altruism is about doing the most good.” That’s weird. But in a comment on my article David Moss noted that there is something going on that does look like the streetlight effect. Still I think that this is the result of good judgment. That, however, doesn’t mean that the same good judgment could not also lead to different decisions given more information.</p>
<h2 id="expected-utility-and-limited-diversification"><a class="toclink" href="#expected-utility-and-limited-diversification">Expected Utility and Limited Diversification</a></h2>
<p>What the streetlight effect describes is a scenario where you lose your wallet in the park, and then go up to the street to search for it there since there are streetlights there and you’d never find it in the dark of the park anyway. The reality is more like that you haven’t lost any particular wallet but that there are potentially a whole bunch of wallets lying around on the street and in the park. Since the light cones are so small, you figure chances are that the biggest wallets are probably somewhere out in the dark.</p>
<p>Translating the metaphor into reality, the darkness are areas of great uncertainty. You may, for example, have uncertainty surrounding how important the cause is that an intervention is trying to address, how effective the intervention is at addressing said cause, whether the intervention is still cost-effective at the margin, and whether the charity implementing the intervention is any good at it. If these uncertainties were known probabilities, you’d have to multiply them all, but you can’t do that unfortunately. The result of all these pseudomultiplications is that the expected utility of interventions in the dark becomes pretty small.</p>
<p>All the while there are a few, pretty few, really cool interventions in the cones of the streetlights, those of the GiveWell and Animal Charity Evaluators top charities. The interventions in the dark haze of uncertainty are many, many more, but you can only expect the tiniest fraction of them to have any considerable cost-effectiveness at the margin and even fewer of them to beat the known top charities.</p>
<p>Ordered by expected utility, the interventions will form something like a hyperbolic function (<a href="http://effective-altruism.com/ea/ws/on_priors/">or more likely Pareto distribution</a>) with very few top interventions and a long long tail of potentially interesting giving opportunities. This doesn’t include interventions that are fairly certain not to be cost-effective.</p>
<p><img alt="Distribution before." src="/images/expected-utility-auctions/distribution-before.png"></p>
<p>When we choose charities to donate to, we can’t diversify infinitely, and that wouldn’t be a good idea anyway. Some even argue against any diversification, but that seems unnecessarily restrictive. In any case, few effective altruists will donate to more than five charities, and they will mostly focus on the charities with the highest expected utility or some slight variation thereof. Most will agree on the high expected utility of the known top charities, but the opinions on the long tail charities will vary widely. One person may be very familiar with a certain long tail charity and may hence think that they’re able to tell with above-average certainty that it’s a good buy, but someone else could worry that this very familiarity might bias the first person’s judgment and donate to a different one or none at all. The result is that the hyperbolic function becomes even more extreme when the y axis are donations.</p>
<p><img alt="Distribution after." src="/images/expected-utility-auctions/distribution-after.png"></p>
<h2 id="expected-utility-auctions"><a class="toclink" href="#expected-utility-auctions">Expected Utility Auctions</a></h2>
<p>This seems all perfectly logical to me, and I see no reason to criticize these people’s decisions. What would be very valuable, however, is what Open Phil does, to try to find the few good giving opportunities in the long tail and lift them out of it.</p>
<p>Open Phil, however, has to prioritize interventions that are very scalable because there are eight billion dollar waiting to be invested. The interventions at least have to maintain a comparable marginal cost-effectiveness for long enough to warrant the time and money invested into finding them.</p>
<p>Significantly, <span class="caps">EA</span> metacharities could profit from prioritization. Such goals as to educate the public about effective giving, to fundraise for effective charities, to collect donation pledges, to conduct prioritization research, and much more might all conceivably be highly cost-effective so long as they haven’t overexerted the limits of their scalability or suffer from any other hidden ailments. Worries about these latter problems are probably what’s holding back many EAs who would otherwise donate to metacharities.</p>
<p>Is there maybe a system that is less reliable than the proper Open Phil treatment but that might serve as a rough guide for these donors and as a training ground for prioritization research hobbyists? Impact certificates may develop into such a tool, but here’s another idea.</p>
<p>I envision an expected utility auction site a bit like Stack Exchange,</p>
<ol>
<li>where people can post their own estimates of the cost-effectiveness and scalability of their project and their reasoning and calculations behind it</li>
<li>where other people can reply to such a bid with their own ones and their own calculations</li>
<li>where a widget at the side displays the current average of the bids weighted according to their upvotes, the standard deviation, and some other metrics of the thread</li>
<li>where a list gives a sorted overview of these metrics of all threads.</li>
</ol>
<p>The unit could be something like 1 util/$ = what GiveDirectly can do for $1. The better estimates would influence the overall total more strongly, and the original poster would be incentivized to start out with a reasonable starting bid to earn upvotes and thus exposure for their project.</p>
<p>Later a link to the donation register of the <span class="caps">EA</span> Hub might be useful so that people who read a review a few months after it was published can estimate how much room for more funding is still left for them.</p>
<p>Do you think this could work? Do you think it’s worthwhile? What other features would such a site need? Who would like to use such a site? Who can implement the <span class="caps">MVP</span>? (I’d do it myself, but I should really be working on my thesis.)</p>
<p>Thanks for reading!</p>
<p>You can find a discussion in the <a href="http://effective-altruism.com/ea/i6/expected_utility_auctions/"><span class="caps">EA</span> Forum</a>.</p>Common Misconceptions about Effective Altruism2015-03-23T17:30:00+00:002015-03-23T17:30:00+00:00Dawn Dreschertag:impartial-priorities.org,2015-03-23:/common-misconceptions-about-effective-altruism.html<p>Effective altruism has seen much welcome criticism that has helped it refine its strategies for determining how to reach its goal of doing the most good—but it has also seen some criticism that is fallacious.</p><div id="pec-encrypted-content" style="display:none">pY3zNybIGEfEFKhbRLhHrw==;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;^</div>
<div id="pec-decrypted-content">
<h4><i></i></h4>
</div>
<form id="pec-decrypt-form">
<p>
Old articles can be embarrassing. If you would like to access them anyway, you may
<a href="https://bit.ly/3jPN8tT" target="_blank">request access here</a>. Please indicate
who you are in case I don’t know you or don’t recognize you from your email address.
</p>
<input type="password" id="pec-content-password" placeholder="Password" />
<button type="submit" id="pec-decrypt-content">Decrypt</button>
</form>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/core.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/enc-base64.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/cipher-core.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/pad-nopadding.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/md5.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/aes.js"></script>
<script type="text/javascript">
(function () {
var strip_padding = function (padded_content, padding_char) {
/*
* Strips the padding character from decrypted content.
*/
for (var i = padded_content.length; i > 0; i--) {
if (padded_content[i - 1] !== padding_char) {
return padded_content.slice(0, i);
}
}
};
var decrypt_content = function (password, iv_b64, ciphertext_b64, padding_char) {
/*
* Decrypts the content from the ciphertext bundle.
*/
var key = CryptoJS.MD5(password),
iv = CryptoJS.enc.Base64.parse(iv_b64),
ciphertext = CryptoJS.enc.Base64.parse(ciphertext_b64),
bundle = {
key: key,
iv: iv,
ciphertext: ciphertext
};
var plaintext = CryptoJS.AES.decrypt(bundle, key, { iv: iv, padding: CryptoJS.pad.NoPadding });
try {
return strip_padding(plaintext.toString(CryptoJS.enc.Utf8), padding_char);
}
catch (err) {
// encoding failed; wrong password
console.log(err);
return false;
}
};
var init_decryptor = function () {
var decrypt_btn = document.getElementById('pec-decrypt-content'),
password_input = document.getElementById('pec-content-password'),
encrypted_content = document.getElementById('pec-encrypted-content'),
decrypted_content = document.getElementById('pec-decrypted-content'),
decrypt_form = document.getElementById('pec-decrypt-form');
decrypt_btn.addEventListener('click', function () {
// grab the ciphertext bundle
var parts = encrypted_content.innerHTML.split(';');
// decrypt it
var content = decrypt_content(
password_input.value,
parts[0],
parts[1],
parts[2]
);
if (content) {
// success; display the decrypted content
decrypted_content.innerHTML = content;
decrypt_form.parentNode.removeChild(decrypt_form);
encrypted_content.parentNode.removeChild(encrypted_content);
// any post processing on the decrypted content should be done here
}
else {
// ¯\_(ツ)_/¯
password_input.value = '';
}
});
};
document.addEventListener('DOMContentLoaded', init_decryptor);
})();
</script>Precise Altruism2015-03-21T14:30:00+00:002015-03-21T14:30:00+00:00Dawn Dreschertag:impartial-priorities.org,2015-03-21:/precise-altruism.html<p><a href="https://github.com/Telofy/precise-altruism">Precise Altruism</a> is a service that reads a number of news feeds of effective altruism organizations and general news aggregators, classifies news articles according to their relevance to altruism and effective altruism, and posts matching articles to <a href="http://altrunews.claviger.net/">Tumblr</a>, <a href="https://twitter.com/Altrunews">Twitter</a>, and <a href="https://www.facebook.com/altrunews">Facebook</a> under the name of Altrunews.</p><div class="admonition type">
<p class="admonition-title">Dated Content</p>
<p>I’ve replaced this service with a <a href="https://bitbucket.org/Telofy/resyndicator">Resyndicator</a> instance.</p>
</div>
<div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#summary">Summary</a></li>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#the-classifier">The Classifier</a></li>
<li><a href="#the-daemon">The Daemon</a></li>
</ul>
</div>
<h2 id="summary"><a class="toclink" href="#summary">Summary</a></h2>
<p><a href="https://github.com/Telofy/precise-altruism">Precise Altruism</a> is a service that reads a number of news feeds of effective altruism organizations and general news aggregators and classifies the news articles according to their relevance to altruism and effective altruism. Articles that fall into this category are then linked and summarized on Tumblr and posted to Twitter and Facebook under the name of Altrunews. (A post is by no means to be understood as an endorsement.)</p>
<p>You can follow Altrunews on <a href="http://altrunews.claviger.net/">Tumblr</a>, <a href="https://twitter.com/Altrunews">Twitter</a>, and <a href="https://www.facebook.com/altrunews">Facebook</a>.</p>
<h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<p>Precise Altruism is a university project by <a href="https://github.com/helmersl/">Lea Helmers</a> and me, which we worked on throughout a data science course by <a href="https://github.com/kashif">Dr. Kashif Rasul</a> at the Freie Universität Berlin.</p>
<p>The service reads feeds from the following sources and classifies them based on a hand-annotated corpus of a few hundred news articles.</p>
<ul>
<li>The Against Malaria Foundation</li>
<li>GiveWell (two feeds)</li>
<li>GiveDirectly</li>
<li>Giving What We Can</li>
<li>The Live You Can Save</li>
<li>Charity Science</li>
<li>80,000 Hours</li>
<li>David Roodman’s blog</li>
<li>Julia Wise’s blog (Giving Gladly)</li>
<li>Ben Kuhn’s blog</li>
<li>Brian Tomasik’s blog (Reducing Suffering)</li>
<li>My own blog (claviger.net)</li>
<li>The Effective Altruism Forum</li>
<li>Animal Charity Evaluators</li>
<li>The Abdul Latif Jameel Poverty Action Lab (three feeds)</li>
<li>Center for Global Development</li>
<li>Sentience Politics</li>
<li>The Global Priorities Project</li>
<li>Gates Notes</li>
<li>Evidence Action</li>
<li>Your Siblings</li>
<li>The World Health Organization</li>
<li>Raising for Effective Giving</li>
<li>Good Ventures</li>
<li>Innovations for Poverty Action</li>
<li>Vegan Outreach</li>
<li>The Future of Humanity Institute</li>
<li>Animal Equality</li>
<li>The Google News feed of English-language news articles containing certain keywords</li>
<li>The Kuerzr feed of English-language news articles containing a similar set of keywords</li>
</ul>
<p>Unfortunately I couldn’t find the feeds of the Schistosomiasis Control Initiative, the Copenhagen Consensus Center, and Mercy For Animals. I’m open for further source feed suggestions, preferably Atom, not <span class="caps">RSS</span>.</p>
<p>By the way, <a href="http://effective-altruism.com/ea/d3/eas_on_rss_and_reddit/">Peter Hurford</a> runs an unfiltered feed exclusively over <span class="caps">EA</span> blogs, and I wrote a thing once, the <a href="https://bitbucket.org/Telofy/resyndicator">Resyndicator</a>, that could be used for something like that (especially in scenarios where it doesn’t already exist).</p>
<h2 id="the-classifier"><a class="toclink" href="#the-classifier">The Classifier</a></h2>
<p>The heart of our application is a classification pipeline built with <a href="http://scikit-learn.org/">scikit-learn</a>, which uses tf-idf to generate a feature matrix of our news data and then a Stochastic Gradient Descent classifier to assign them one of our two categories.</p>
<p>We used grid search and cross-validation to determine the optimal classifier and an optimal set of parameters for it. Using only a small set of plausible parameters and only three splits for the cross-validation, we quickly determined the four out of initially ten classification algorithms that performed best on our data, Stochastic Gradient Descent, Logistic Regression, and two variations of the Support Vector Machines classifier. In our final, most finely tuned run, Stochastic Gradient Descent achieved an F1 score of 93%, about two percentage points more than the best of the other three classifiers.</p>
<p>The clearest takeaways from the grid search over a plausible <span class="caps">SGD</span> parameter set were that as loss functions <code>log</code>, <code>hinge</code>, <code>modified_huber</code>, and <code>perceptron</code> performed well; that as penalty <code>l2</code> and <code>elasticnet</code> performed well; that activating the shuffling helped; that using bigrams in addition to unigrams was useful but that 3-grams did not improve the F1 score; and that the best values for <code>alpha</code> and <code>n_iter</code> varied widely among the best configurations.</p>
<p>It’s been almost a year since I implemented this, so please don’t quiz me on the details.</p>
<h2 id="the-daemon"><a class="toclink" href="#the-daemon">The Daemon</a></h2>
<p>The daemon is the service that continuously runs on the server and continually checks the source feeds. It sends <code>if-modified-since</code> and <code>if-none-matches</code> headers whenever possible to minimize server load and traffic. Then the feed entries are compared to those in the database to filter out known ones, whereby we also compute the <a href="http://en.wikipedia.org/wiki/Jaccard_index">Jaccard distance</a> between the preprocessed titles to avoid posting the same press releases over and over.</p>
<p>The articles that are typically associated with these entries are then fetched, stripped of boilerplate using Readability, summarized using Sumy, and finally posted to Tumblr. We extended the extraction step with one that also extracts a featured image and added a naive keyword extraction for the post tags on Tumblr.</p>Effective Altruism 1012015-02-07T17:30:00+00:002020-09-07T11:35:00+00:00Dawn Dreschertag:impartial-priorities.org,2015-02-07:/effective-altruism-101.html<p>Effective altruism allows donors to make confident, evidence-based giving decisions that turn even small donations into life-changing events for those in need. (Here’s a recent <a href="https://forum.effectivealtruism.org/posts/oko3LekWn95Dh6ANW/introduction-to-effective-altruism-2">“Introduction to Effective Altruism”</a> that will hopefully continue to be updated in light of new insights.)</p><div id="pec-encrypted-content" style="display:none">FWhRX6sQnrrKWKTHJ/VdDg==;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;^</div>
<div id="pec-decrypted-content">
<h4><i></i></h4>
</div>
<form id="pec-decrypt-form">
<p>
Old articles can be embarrassing. If you would like to access them anyway, you may
<a href="https://bit.ly/3jPN8tT" target="_blank">request access here</a>. Please indicate
who you are in case I don’t know you or don’t recognize you from your email address.
</p>
<input type="password" id="pec-content-password" placeholder="Password" />
<button type="submit" id="pec-decrypt-content">Decrypt</button>
</form>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/core.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/enc-base64.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/cipher-core.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/pad-nopadding.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/md5.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/aes.js"></script>
<script type="text/javascript">
(function () {
var strip_padding = function (padded_content, padding_char) {
/*
* Strips the padding character from decrypted content.
*/
for (var i = padded_content.length; i > 0; i--) {
if (padded_content[i - 1] !== padding_char) {
return padded_content.slice(0, i);
}
}
};
var decrypt_content = function (password, iv_b64, ciphertext_b64, padding_char) {
/*
* Decrypts the content from the ciphertext bundle.
*/
var key = CryptoJS.MD5(password),
iv = CryptoJS.enc.Base64.parse(iv_b64),
ciphertext = CryptoJS.enc.Base64.parse(ciphertext_b64),
bundle = {
key: key,
iv: iv,
ciphertext: ciphertext
};
var plaintext = CryptoJS.AES.decrypt(bundle, key, { iv: iv, padding: CryptoJS.pad.NoPadding });
try {
return strip_padding(plaintext.toString(CryptoJS.enc.Utf8), padding_char);
}
catch (err) {
// encoding failed; wrong password
console.log(err);
return false;
}
};
var init_decryptor = function () {
var decrypt_btn = document.getElementById('pec-decrypt-content'),
password_input = document.getElementById('pec-content-password'),
encrypted_content = document.getElementById('pec-encrypted-content'),
decrypted_content = document.getElementById('pec-decrypted-content'),
decrypt_form = document.getElementById('pec-decrypt-form');
decrypt_btn.addEventListener('click', function () {
// grab the ciphertext bundle
var parts = encrypted_content.innerHTML.split(';');
// decrypt it
var content = decrypt_content(
password_input.value,
parts[0],
parts[1],
parts[2]
);
if (content) {
// success; display the decrypted content
decrypted_content.innerHTML = content;
decrypt_form.parentNode.removeChild(decrypt_form);
encrypted_content.parentNode.removeChild(encrypted_content);
// any post processing on the decrypted content should be done here
}
else {
// ¯\_(ツ)_/¯
password_input.value = '';
}
});
};
document.addEventListener('DOMContentLoaded', init_decryptor);
})();
</script>Adrift on Brook Madregot2014-09-13T12:00:00+00:002014-09-13T12:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2014-09-13:/adrift-on-brook-madregot.html<p>Gene Wolfe’s <em>The Book of the New Sun</em> is a Daedalean masterpiece of literature and illusionism for a whole host of reasons, but centrally because of its narrators. This paper investigates the personality of the narrator-protagonist Severian by recourse to the psychological literature, specifically Alexander Luria’s study of the mnemonist Solomon Shereshevsky.</p><style>
body {counter-reset: section}
.entry-content h2 {counter-reset: sub-section}
.entry-content h3 {counter-reset: composite}
.entry-content h4 {counter-reset: detail}
.entry-content h2:before {
counter-increment: section;
content: counter(section) " ";
}
.entry-content h3:before {
counter-increment: sub-section;
content: counter(section) "." counter(sub-section) " ";
}
.entry-content h4:before {
counter-increment: composite;
content: counter(section) "." counter(sub-section) "." counter(composite) " ";
}
.entry-content h5:before {
counter-increment: detail;
content: counter(section) "." counter(sub-section) "." counter(composite) "." counter(detail) " ";
}
</style>
<div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#plot-summary">Plot Summary</a></li>
<li><a href="#the-reader">The Reader</a></li>
<li><a href="#the-narrator">The Narrator</a><ul>
<li><a href="#qualifications">Qualifications</a></li>
<li><a href="#the-mnemonist">The Mnemonist</a><ul>
<li><a href="#wolfes-s">Wolfe’s S.</a></li>
<li><a href="#lurias-s">Luria’s S.</a></li>
</ul>
</li>
<li><a href="#the-marionette">The Marionette</a></li>
</ul>
</li>
<li><a href="#reconciliation">Reconciliation</a><ul>
<li><a href="#the-man-your-mother-bore">The Man Your Mother Bore</a></li>
<li><a href="#our-own-shadows-race-into-the-past">Our Own Shadows Race Into the Past</a></li>
<li><a href="#for-that-were-you-chosen">For That Were You Chosen</a></li>
</ul>
</li>
<li><a href="#conclusion">Conclusion</a></li>
<li><a href="#endnotes">Endnotes</a></li>
<li><a href="#works-cited">Works Cited</a></li>
</ul>
</div>
<h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<p>Reviewers of Gene Wolfe’s Urth Cycle—centrally the four volumes of <em>The
Book of the New Sun</em> and their coda <em>The Urth of the New Sun</em><sup id="fnref:1"><a class="footnote-ref" href="#fn:1">1</a></sup>—have likened the author to an
illusionist, a cardsharp (<a href="#budrys2012benchmarks">Budrys 195</a>), and a builder of
labyrinths (<a href="#wolfe1983solar">Wolfe, “A Solar Labyrinth”</a>; <a href="#borski2004solar">Borski</a>). Misdirection, slight of
hand, and torturous paths that circle back on themselves are the stuff
of his oeuvre and the five books epitomize this style. Within the Urth
Cycle, the narrators are cast in major obfuscatory roles. First there is
G.W., the intradiegetic translator of the manuscript, which he<sup id="fnref:2"><a class="footnote-ref" href="#fn:2">2</a></sup> has
somehow received from the future, so that he has to borrow many words
from our past as “suggestive rather than definitive” stand-ins for words
of a language that “has not yet achieved existence.” (<a href="#wolfe1980shadow">Wolfe, <em>The Shadow of the Torturer</em> 211; appx.</a>)
And then there is Severian, the author of the manuscript—at
least in a superficial sense.<sup id="fnref:3"><a class="footnote-ref" href="#fn:3">3</a></sup></p>
<p>Severian, later known as Severian the Great and Severian the Lame, has a
peculiar mental condition that lends him extraordinary mnemonic
abilities but also comes with more elusive side effects. Like many other
books by Gene Wolfe, the Urth Cycle puts us into this extraordinary
mind, so that we may understand its nature and overcome its limitations.
Trusting Merryn’s words of wisdom that “There is no magic[,] only
knowledge, more or less hidden” (<a href="#wolfe1981claw">Wolfe, <em>The Claw of the Conciliator</em> 404; ch. 31</a>),
reviewers have long tried to achieve the latter; this paper will attempt the
former by highlighting how the first-person narrative mirrors Severian’s cognition.</p>
<p>Such a narrow focus is indispensable in a discussion of a cycle that has
been analyzed in several books, numerous articles, and countless mailing
list posts over the course of three decades. No less focused approach
could come to any conclusion within just a few thousand words. Said
focus, however, also necessitates that many intriguing tangents be cut
short if they do not circle back on the gist of the paper. The cited
literature is recommended to anyone interested in investigating these further.</p>
<p>The structure of the following is such that after a short summary, one
section will briefly highlight some of the effects of the text on the
reader. Then the principle section will expand on these observations to
encompass many key elements that help unlock the story that lies
obscured by the narration. Penultimately, a more speculative section
will take a few tentative steps into the story thus revealed, until
finally the conclusion will summarize the findings.</p>
<h2 id="plot-summary"><a class="toclink" href="#plot-summary">Plot Summary</a></h2>
<p>This summary is meant to guide a reader who wants to follow the
arguments in this paper without having read the Urth Cycle in its
entirety. It is also highly compressed. Hence you need to take care to
bear in mind that focused as it is on providing a basis for this
particular paper, it is also inevitably biased. The facts that seem
crucial to one reader may appear peripheral to another, but there is no
way to even attempt to do justice to all the more prevalent readings in
a summary without rendering it greatly more expansive than the scope of
this paper would allow.<sup id="fnref:4"><a class="footnote-ref" href="#fn:4">4</a></sup></p>
<div> </div>
<p>The torturer’s apprentice Severian has drowned in Gyoll where he has
been resurrected in a new body. Thinking he merely narrowly escaped
death, he recovers from his trauma quickly enough to save the life of
the famous outlaw Vodalus by killing a volunteer guard. Vodalus thanks
him by giving him a chrisos, a valuable coin that shows the autarch’s
face, but Severian interprets this as a symbolic enlistment of his
person in the movement of the Vodalarii and professes to have shared
their ideals.</p>
<p>One winter, he unwittingly resurrects a dog, smuggles it into their
guild’s home, the Matachin Tower, and expertly tends to its wounds. Soon
the dog scuttles off and its tracks lead him into dark tunnels
underneath the Citadel—a fortified harbor for spaceships such as the
Matachin Tower, which are long grounded and have largely fallen into
disrepair so that they are used only as houses for about 135
guilds—where he loses its tracks and emerges in the Atrium of Time, a
likely time traveling platform at the heart of the Citadel. There he
meets Valeria, who he will later marry.</p>
<p>A while later, the Chatelaine Thecla is put into the custody of the
torturers and meets Severian. Highly placed, she can request for
Severian to entertain her. In an attempt to keep Severian from
developing feelings for her despite their closeness, the guild pays for
a trip to brothel for him where he meets the autarch but fails to
recognize him from the coin. The guild’s attempt fails. When Thecla is
subjected to torture, Severian enables her to commit suicide.</p>
<p>Expecting to be tortured and executed, Severian is surprised to find
that the masters of the guild appear to be so afraid to lose the trust
of the courts if Severian’s violation of guild dogma should become known
that they merely exile him to a distant town.</p>
<p>He ventures forth into the world carrying first the valuable sword
Terminus Est and soon a valuable gem, the Claw of the Conciliator, which
he seeks to return to its rightful owners. Almost always he has
companions at his side, most significantly the teenage yet wise Dorcas.
At his next meeting with Vodalus, he gets to merge Thecla’s mind into
his own, and soon his allegiances veer about 180°; still he accepts a
new mission from Vodalus.</p>
<p>At several points—specifically in a play by a Dr. Talos, which is based
on an old book, <em>The Book of the New Sun</em>—it is now indicated that the
sun is slowly dying because it was injected with a black whole that eats
it up from inside. Some of the people of Urth now hope for the coming of
the New Sun (pun surely intended), while others fear the destruction its
gravitational influence would wreak on Urth.</p>
<p>Severian’s newest mission is exhausted when he delivers a message to an
agent of Vodalus’s inside the secret house of the House Absolute,<sup id="fnref:5"><a class="footnote-ref" href="#fn:5">5</a></sup> an
agent who turns out to be the autarch himself. This time Severian
recognizes him from the House Azure but it is from Thecla’s memories
that he, much delayed, recognizes him as the autarch.</p>
<p>He completes his initial mission by becoming lictor of Thrax but soon
repeats, in effect, the transgression that got him exiled and has to flee
the town. When his mission of returning the Claw is exhausted as well,
he aimlessly joins the cavalry in the war against the Ascians.</p>
<p>After having been saved repeatedly by friend and foe alike, the autarch
bequeaths to him his position, and as ruler of the commonwealth Severian
awaits the trials of the hierogrammates that will determine whether Urth
deserves a fresh sun.</p>
<p>To attend these trials, he travels to Yesod, an ostensibly higher<sup id="fnref:6"><a class="footnote-ref" href="#fn:6">6</a></sup>
universe with an eponymous planet ship within it. There he is told that
his whole life and the lives of his predecessors on the Phoenix Throne
were all part of his trial and that he has already passed it. He also
learns that the reason the hierogrammates have orchestrated all this is
so the hieros—descendents, they hope, of the humans of Severian’s
day—would come into existence and eventually create the hierogrammates,
which would then evade the demise and reconstitution of the Briahtic
universe (or rather multiverse) by fleeing into the universe of
Yesod.<sup id="fnref:7"><a class="footnote-ref" href="#fn:7">7</a></sup> Hence they are working to ensure their own procreation or
birth as a race.</p>
<p>Severian then goes on to become the Jesus-like Conciliator in the
distant past, where he tells his life story while incarcerated in the
Matachin Tower, whereby he uses Dr. Talos’s play to guide his narration.
Another prisoner, Canog, protocols it and turns it later into a book
that would become known as <em>Book of the New Sun</em>.</p>
<p>Later he steps into the future to observe the destruction of Urth, goes
back to an even more distant past to become Apu-Punchau, and finally
returns to the future and to Ushas, as postdiluvian Urth is called.</p>
<h2 id="the-reader"><a class="toclink" href="#the-reader">The Reader</a></h2>
<p>Oftentimes a strong internal focalization has the effect of allowing the
reader an insight, however superficial, into the nature of the focalized
character’s thought processes and modes of perception. In many of his
books, Gene Wolfe goes a step further. Internal focalization of this
kind still requires the empathetic cooperation of the reader, and in
novels such as <em>There Are Doors</em>, this empathy provides a crucial part of the reading
experience, but at the same time a closely related process is going on,
a recreation of the focalized characters confusion on the extradiegetic
level. Readers who try to make sense of the text, just as the focal
character tries to make sense of his or her life, encounter in their
interpretative efforts hurdles that replicate the difficulties of the
character. Empathy is no longer required to evoke the readers’ sympathy.</p>
<p><a href="#wright2003attending">Wright</a> identifies this artifice as an all-pervasive theme
in the Urth Cycle. The books are replete with intertextual references to
sources as diverse as Kabbalah; Tarot; Grimm’s Fairy Tales; Greek,
Roman, Christian, Egyptian, Persian, and Norse mythology; various
traditions of sun worship; and of course many individual authors such as
Marcel Proust, Jorge Luis Borges, Joseph Campbell, <span class="caps">H.G.</span> Wells, Lewis
Carroll, John Locke, and many more. All of these are worked into the
story so seamlessly that the reader has to be versed in the referenced
material to even notice them, but then they open up new subtextual
backdoors to countless minor mysteries of the text. The richly textured,
defamiliarizing descriptions that have led reviewers to call the style
of the Urth Cycle baroque (<a href="#gordon1986gene">Gordon 75</a>) are similar in effect
and function, and deciphering them does not get easier when the narrator
starts to shift between the many personalities he is host to. Finally,
there is also the archaic, eclecticist diction, which first refracts
what light one tries to shine on the events, but then, under close
etymological or historical examination, reveals important (albeit
limited) clues to their deeper significance. While Wolfe may mention
crucial facts only once or twice in his books,<sup id="fnref:8"><a class="footnote-ref" href="#fn:8">8</a></sup> off-handed remarks,
idle metaphors, and the words themselves can give additional clues to
those who look closely enough.</p>
<p>However interesting these facets of the text may be,
<a href="#wright2003attending">Wright</a> is wary of them for the myopia they induce, and
especially in view of the tetralogy that forms the Urth Cycle proper,
his apprehension is justified. Where unsuspecting readers will see a
monomythical coming-of-age fantasy story with many Romantic elements,
more careful readers will find a labyrinthine science fiction with
self-similar riddles on all surfaces, but they would have to be almost
paranoid to suspect that if they could view the structure from the
orbit, they would read a mocking inscription in the patterns of its
meandering paths.<sup id="fnref:9"><a class="footnote-ref" href="#fn:9">9</a></sup></p>
<p>The coda resolves some of these puzzles, for which <a href="#Kincaid2004">Kincaid</a> has
criticized it as “a tying up of loose ends that didn’t necessarily need
tying.” Especially a character by the name of Apheta, who, fittingly,
has no voice but speaks by canceling out ambient noise, makes a central
one explicit. To <a href="#wright2003attending">Wright</a>, it is the one central puzzle of
the series, and he makes a good argument for its pervasiveness, but in a
world where “everything, whatever happens, has three meanings,”
(<a href="#wolfe1980shadow"><em>Shadow</em> 190; ch. 32</a>) there surely are more mysteries of equal
rank that are still waiting to be discovered.<sup id="fnref:10"><a class="footnote-ref" href="#fn:10">10</a></sup></p>
<p>What then is the origin of all this fascinating, addictive, but
distracting embellishment? One of the answers lies in the cognitive
predispositions of Severian, who unconsciously plunges the reader into
the same noisy reality that is the only one he knows.</p>
<h2 id="the-narrator"><a class="toclink" href="#the-narrator">The Narrator</a></h2>
<h3 id="qualifications"><a class="toclink" href="#qualifications">Qualifications</a></h3>
<p>In order to attempt an examination of the unusual cognitive abilities
and limitations that are mirrored in Severian’s narrative, it is
necessary to make at least two assumptions that are not trivial, namely,
that Severian is (intradiegetically) real and that his memories are not
wholly fabricated.</p>
<p>It could be argued that the hierogrammates could have more easily
instated the Conciliator myth by fabricating the first <em>Book of the New
Sun</em> and leaking it to Canog rather than having Severian reenact all of
it, so that Severian’s <em>Book of the New Sun</em> may be similarly fabricated
for the purposes of manipulating its intradiegetic readers or possibly us.</p>
<p>Similarly, Severian may have received a much harsher punishment than
exile for his violation of guild dogma, may have been locked away on the
third level of the oubliette where Master Palaemon, pitying him, left
him the four books Severian had fetched for Thecla, among them the brown
book and Canog’s <em>Book of the New Sun</em> (<a href="#wolfe1984plan">Wolfe, <em>Plan(e)t Engineering</em> 15</a>),
and is at times visited by Master Gurloes who would talk with him of things no
eavesdropper could understand (<a href="#wolfe1980shadow"><em>Shadow</em> 56; ch. 7</a>). Severian
would retreat too easily into his rich imagination and relive his own
version of Canog’s story, eventually able to write it down (time and
time again for lack of anything else to do) thanks to a “quiet well of
vermilion ink” (<a href="#wolfe1980shadow"><em>Shadow</em> 205; ch. 35</a>) maybe from one of his
legs.<sup id="fnref:11"><a class="footnote-ref" href="#fn:11">11</a></sup></p>
<p>These interpretations, however, would render much of the following
considerations moot, so that they need to be temporarily put in abeyance.</p>
<h3 id="the-mnemonist"><a class="toclink" href="#the-mnemonist">The Mnemonist</a></h3>
<h4 id="wolfes-s"><a class="toclink" href="#wolfes-s">Wolfe’s S.</a></h4>
<p>It seems to be a repeating theme that the narrators of Gene Wolfe’s
novels die shortly before the beginning of their narration. The title of
the first chapter, “Resurrection and Death,” gives this fact away, which
will only be explained again in the coda.<sup id="fnref:12"><a class="footnote-ref" href="#fn:12">12</a></sup> The same chapter
continues to summarize many of the crucial facts in intimations that, in
some cases, will only become clear much later.</p>
<p>On its first page already, Severian explains the nature of his memory in
a section that will become even more significant in the following:</p>
<blockquote>
<p>Just as all that appears imperishable tends toward its own
destruction, those moments that at the time seem the most fleeting
recreate themselves—not only in my memory (which in the final
accounting loses nothing) but in the throbbing of my heart and the
prickling of my hair, making themselves new just as our Commonwealth
reconstitutes itself each morning in the shrill tones of its own
clarions. (<a href="#wolfe1980shadow"><em>Shadow</em> 9; ch. 1</a>)</p>
</blockquote>
<p>Time and time again he mentions this “perfect memory”<sup id="fnref:13"><a class="footnote-ref" href="#fn:13">13</a></sup> and uses it
to pass time by counting from memory 137 soldiers who had marched past
him a week or so earlier (<a href="#wolfe1981claw"><em>Claw</em> 319; ch. 18</a>) or to recount
verbatim several stories he was told, but superficially it does not seem
to have a major significance for the plot.</p>
<p>The simile in the quoted paragraph is also a surprisingly apt
description since each morning may appear a recreation of the last but
is yet bound to be different from it in countless ways, just as new
recall errors increasingly skew the original memory with every
reactivation (<a href="#bridge2012neural">Bridge and Paller</a>).</p>
<p>Finally, the same section even illustrates this process. A few sentences
before the quoted paragraph, we have the sentence “I would have hidden,
but Roche held me, saying, ‘Wait, I see pikes,’” and immediately
following it, “The men had no armor, as I could soon see by the sickly
yellow light of the lanterns; but they had pikes, as Drotte had said,
and staves and hatchets.”</p>
<p>Other examples of the mutability of his memory concern the bag Docas
sews to hold the Claw and the pistol hand-off in the necropolis. As
<a href="#wright2003attending">Wright (114)</a> observes, the bag transmogrifies from
“doeskin” (<a href="#wolfe1982sword">Wolfe, <em>The Sword of the Lictor</em> 13; ch. 1</a>) into
“manskin” (<a href="#wolfe1987urth"><em>Urth</em> 279; ch. 39</a>), a very unlikely material for Dorcas to have used, and
(here <a href="#wright2003attending">Wright</a> quotes <a href="#Greenland1982">Greenland (82–85)</a>) Severian
first recalls Vodalus giving his pistol to Hildegrin, who then, being
unacquainted with the weapon, hands it to Thea (<a href="#wolfe1980shadow"><em>Shadow</em> 13; ch. 1</a>),
and later recalls Vodalus handing it to Thea directly
(<a href="#wolfe1981claw"><em>Claw</em> 221; ch. 1</a>).</p>
<p>Severian freely admits to the mutability of his ostensibly perfect
memory at several places.</p>
<blockquote>
<p>You that read [my story] cannot but have noticed that I have not
scrupled to recount in great detail things that transpired years ago,
and to give the very words of those who spoke to me, and the very
words with which I replied; and you must have thought this only a
conventional device I had adopted to make my story flow more smoothly.
The truth is that I am one of those who are cursed with what is called
perfect recollection. We cannot, as I have sometimes heard foolishly
alleged, remember everything. I cannot recall the ordering of the
books on the shelves in the library of Master Ultan, for example. But
I can remember more than many would credit: the position of each
object on a table I walked past when I was a child, and even that I
have recalled some scene to mind previously, and how that remembered
incident differed from the memory of it I have now. (<a href="#wolfe1981claw"><em>Claw</em> 260; ch. 8</a>)</p>
</blockquote>
<p>Known gaps in his memory do not seem to worry him; he may not have paid
attention to the ordering of the books in the first place, and so it
never impressed itself on his memory. Alternatively, it may be that he
once knew it but not only forgot it but even forgot ever having known
it. Either case would explain his indifferent attitude. Interesting is
also the equanimity with which he observes alterations of his memories.
Since his memory of how he had previously recalled some scene may have
undergone alterations just as severe as any alterations of his memory of
the scene itself, he has no way of knowing the scene itself anymore.
That his original memory of it has been replaced by a fiction does not
undo the fact of his forgetting, it merely conceals it, yet he does not
seem to acknowledge this form of forgetting or dismisses it as
inconsequential (“I searched my memory, which is perfect, except perhaps
for a few slight lapses and distortions.” (<a href="#wolfe1987urth"><em>Urth</em> 244; ch. 34</a>)).</p>
<p>What seems to be truly frightening to him, though, is forgetting. The
following excerpts show his fear of forgetting and the powerful
mechanisms his brain has developed to conceal it from him.</p>
<blockquote>
<p>I shook my head. “I don’t want to forget, Tzadkiel. I’ve boasted too
often that I forget nothing, and forgetting—which I have known once or
twice—seems to me a kind of death.” (<a href="#wolfe1987urth"><em>Urth</em> 174; ch. 24</a>)</p>
<p>“I have forgotten! Do you remember when we flew over the armies? For a
time I forgot it! I know now what it is to forget.”</p>
<p>There was pale laughter in his voice. “Which you will now remember always.”</p>
<p>“I hope so, but it fades even as we speak. It vanishes like mist,
which must itself be a forgetting. …” (<a href="#wolfe1983citadel">Wolfe, <em>The Citadel of the Autarch</em>
336; ch. 25</a>)</p>
</blockquote>
<p>In addition to these sections, which portray his acknowledged fear of
forgetting, there is another where he describes the experience of facing
his fear of phantom memories poignantly as the “most harrowing of [his]
life,” The catalyst was that, as an apprentice, he has assumed that many
of the upper-class prisoners given into the guild’s custody were
supporters of Vodalus. Upon reading some of the clients’ court dossiers,
however, he has found that none of them were, but only minutes later
thinks that he has heard Vodalus’s name in a conversation though no one
else seems to have heard it. He extrapolates that his whole memory of
meeting Vodalus and Thea may have been a hallucination or phantom
memory, that only his killing of the volunteer may have been real.</p>
<blockquote>
<p>It was in this instant of confusion that I realized for the first time
that I am in some degree insane. It could be argued that it was the
most harrowing of my life. I had lied often to Master Gurloes and
Master Palaemon, to Master Malrubius while he still lived, to Drotte
because he was captain, to Roche because he was older and stronger
than I, and to Eata and the other smaller apprentices because I hoped
to make them respect me. Now I could no longer be sure my own mind was
not lying to me; all my falsehoods were recoiling on me, and I who
remembered everything could not be certain those memories were more
than my own dreams. I recalled the moonlit face of Vodalus; but then,
I had wanted to see it. I recalled his voice as he spoke to me, but I
had desired to hear it, and the woman’s voice too. (<a href="#wolfe1980shadow"><em>Shadow</em> 27; ch. 3</a>)</p>
</blockquote>
<p>These limitations, however, are marginal compared to the feats of memory
he performs, for example, when retelling the same dialogue verbatim and
identically several times throughout the narrative.</p>
<p>Fear of forgetting is one side of the chrisos, but the more optimistic
side becomes evident once Severian becomes more aware of the possibility
of time travel.</p>
<blockquote>
<p>The chiliarch said, “We’ll stay here and die with you, Conciliator, if
you desire it.”</p>
<p>“I don’t,” I told them. “And I won’t die.” I tried to reveal the
workings of Time to them, though I do not understand them myself.
“Everyone who has lived is still alive, somewhen. But you are in great
danger. Go!” (<a href="#wolfe1987urth"><em>Urth</em> 279; ch. 39</a>)</p>
</blockquote>
<p>This consolation, surely unhelpful for the chiliarch, may be one that
has great meaning for Severian, for whom time was gradually becoming as
freely navigable as space. Hence his fear may be less of him losing his
memories than of the corresponding events getting irretrievably lost in
the past. A fear that only abates as he grows accustomed to time travel.</p>
<p>His motivation for writing his <em>Book of the New Sun</em> may spring from the
same source. In <em>Return to the Whorl</em>, it is revealed that Horn alias Incanto, the narrator
of <em>The Book of the Short Sun</em>, introduced Severian to the writing of
autobiographies or memoirs, but that alone would probably not have
sufficed to drive him to invest endless hours into the recording of his memories.</p>
<p>In his own autobiography, he laments, “But what a disease this writing
business is!” (<a href="#wolfe1987urth"><em>Urth</em> 1; ch. 1</a>) This statement suggests that he
is not merely motivated but compelled to write. Moreover, he repeatedly
claims not to be writing for anyone. The same volume starts with the words:</p>
<blockquote>
<p>Having cast one manuscript into the seas of time, I now begin again.
Surely it is absurd; but I am not—I will not be—so absurd myself as to
suppose that this will ever find a reader, even in me. Let me describe
then, to no one and nothing, just who I am and what it is that I have
done to Urth.</p>
</blockquote>
<p>His manner of publication is also not optimized for mass dissemination,
distinguishing the book from the autobiographies of other famous rulers.
Of the first manuscript, he gave one copy into the care of the library
of Nessus, a giant, dusty archive underneath the city that is not open
to the public (<a href="#wolfe1980shadow"><em>Shadow</em> 42; ch. 6</a>), then wrote a second copy
from memory (surely with some unconscious alterations), sealed it in a
trunk, and tossed it into space. If it had not been for G.W.’s knack for
time travel and xenolinguistics, at most a librarian or two would have
read it. (<a href="#johnson2007phantom">Johnson</a>)</p>
<p>While some part of him may be unable to acknowledge that his memory, due
to its silent mutability, is unreliable, he may yet be aware of it on
another level, one that does not allow him to express this insight.<sup id="fnref:14"><a class="footnote-ref" href="#fn:14">14</a></sup>
His drive to write then may stem from a latent fear that without his
help the events he witnessed would fade to oblivion, either upon his
death or even much sooner. The reason he has never, to our knowledge,
continued his autobiography further than he had, may also be due to the
waning of the same fear as time travel becomes natural for him.</p>
<p>From this evidence it becomes clear that while Severian’s recall is
exceptional, it is imperfect and subject to slight alterations, which
his strong imagination disguises from him. It is also likely, though the
evidence appears more tenuous, that Severian’s fear of these
imperfections is the driving force behind his writing.</p>
<h4 id="lurias-s"><a class="toclink" href="#lurias-s">Luria’s S.</a></h4>
<p>From the 1920s to 1950s, the Soviet neuropsychologist Alexander Luria
studied the journalist and mnemonist Solomon Shereshevsky and summarized
his findings in his book <em>The Mind of a Mnemonist</em>.<sup id="fnref:15"><a class="footnote-ref" href="#fn:15">15</a></sup> Solomon Shereshevsky, called S. in the
book,<sup id="fnref:16"><a class="footnote-ref" href="#fn:16">16</a></sup> evinces several parallels to Severian, as
<a href="#wright2003attending">Wright</a> already observed, but also a few marked differences.</p>
<p>Jerome Bruner’s foreword already contains a good summary of the parallels:</p>
<blockquote>
<p>For the mnemonist, S., whose case is studied in such exquisite detail
in these pages, is a man whose memory is a memory of particulars,
particulars that are rich in imagery, thematic elaboation [<em>sic</em>], and
affect. But it is a memory that is peculiarly lacking in one important
feature: the capacity to convert encounters with the particular into
instances of the general, enabling one to form general concepts even
though the particulars are lost. It is this latter type of “memory
without record” that seems so poorly developed in this man.</p>
<p>Several notable things about the disorders of this mnemonist are
especially fascinating from a psychological point of view. For one
thing, the sheer persistence of ikonic [<em>sic</em>] memory is so great that
one wonders whether there is some failure in the swift metabolism of
short-term memory. His “immediate” images haunt him for hours, types
of images that in much recent work on short-term memory are found to
fade to a point where information retrieval from them is not possible
after a second or so. Along with this trait there is also a
non-selectivity about his memory, such that what remains behind is a
kind of junk heap of impressions. …</p>
<p>So powerful is his imagery that this man can easily drive his pulse up
by imagining running. He is flooded and disturbed by the images and
impressions of childhood, and, when he was a child, his imagery of
school would become so “real” that he would lie abed rather than get
out from under the quilt and get ready. It is interesting that, given
his mode of remembering, there seems to be no childhood amnesia, and
his memories from the earliest period can cause him acute malaise and
chagrin. Throughout, there is a childlike quality in the protocols,
protocols that are rich beyond anything I have ever encountered in the
psychological literature on memory disorders. S.’s life in some deeply
touching way is a failure. He waited for something to happen to him,
some great thing. In the conduct of his life, too, there was a
passive-receptive attitude, almost precluding organized striving. In
place of the more abstract and constructional attitude of planning,
there was waiting. (<a href="#lurija1968mind">Luria 5–7</a>)</p>
</blockquote>
<p>Certainly Severian’s writing is brimful with particulars and rich in
imagery. More irreverently, it could be called a “junk heap of
impressions.” Similarly, that Shereshevsky could “easily drive his pulse
up by imagining running” is reminiscent of Severian’s description of
“moments that … recreate themselves … in the throbbing of [his] heart
and the prickling of [his] hair” (<a href="#wolfe1980shadow"><em>Shadow</em> 9; ch. 1</a>). But
Severian seems to also know about some of the side effects of his condition:</p>
<blockquote>
<p>Some say [perfect recall] is linked to weak judgment—of that I am no
judge. But it has another danger, one I have encountered many times.
When I cast my mind into the past, as I am doing now and as I did then
when I sought to recall my dream, I remember it so well that I seem to
move again in the bygone day, a day old—new, and unchanged each time I
draw it to the surface of my mind, its eidolons as real as I.
(<a href="#wolfe1981claw"><em>Claw</em> 261; ch. 8</a>)</p>
</blockquote>
<p>Shereshevsky found it difficult to recognize faces (<a href="#lurija1968mind">Luria 64</a>)
and voices (<a href="#lurija1968mind">Luria 25</a>) because of how changeable they were. He
could not extract and recognize whatever commonality most people find in
the same face or voice what allows them to agree that it is indeed the
same. He compared the attempt to trying to tell apart waves on an ocean.
Similarly, Severian has problems recognizing faces, for example, when he
repeatedly sees the old autarch but the recognition keeps being
one-sided. There are also cases where he considers that people may not
be the same people anymore, having changed or matured so much. In the
case of his own person (<a href="#wolfe1983citadel"><em>Citadel</em> 401; ch. 37</a>), the observation
is debatable on several levels, yet his readiness to accept this
unintuitive nonidentity is telling.<sup id="fnref:17"><a class="footnote-ref" href="#fn:17">17</a></sup></p>
<p>The same extraneous data that distracts Shereshevsky and Severian also
distracts the readers from whatever patterns they would otherwise
recognize. That is one of the ways in which the book forces
inquisitive readers to sympathize, rather than bargaining for their
empathy. One could imagine that Shereshevsky was not describing the
influence of <span class="caps">S.M.</span> Eisenstein’s acoustic voice in this paragraph but that
of Severian’s literary voice:</p>
<blockquote>
<p>You know there are people who seem to have many voices, whose voices
seem to be an entire composition, a bouquet. The late <span class="caps">S.M.</span> Eisenstein
had just such a voice: listening to him, it was as though a flame with
fibers protruding from it was advancing right toward me. I got so
interested in his voice, I couldn’t follow what he was saying.
(<a href="#lurija1968mind">Luria 24</a>)</p>
</blockquote>
<p>The Urth we see is thus an eclecticist world somewhat like the House
Azure, “in which the accumulation and interconnection of what were
originally separate buildings produce a confusion of jutting wings and
architectural styles, with peaks and turrets where the first builders
had intended nothing more than rooftops,” (<a href="#wolfe1980shadow"><em>Shadow</em> 62; ch. 9</a>)
except that Severian’s uncomprehending, defamiliarizing lens (aided by
the translation) fuses the heterodox assemblage of numerous religious
mythologies, weapons and ships from various cultures and times of Urth’s
past, animals from at least nine geological epochs, and authors of
various genres into a fairly homogeneous whole. Put differently, it
sounds like a color blind person describing in great detail a most
artful Ishihara Color Test to the reader, all the while completely
missing the number depicted in the center. If the readers could look at
the image directly, they would see it clearly—although different readers
might see a small set of different numbers.</p>
<p>There are also several instances (e.g., <a href="#wolfe1981claw"><em>Claw</em> 261; ch. 8</a>) in
<em>The Book of the New Sun</em> where Severian gets lost in his memories
forgetting the reality around him just as the young Shereshevsky mixed
up the real getting up and getting to school with his imagined
(third-person) version of it (<a href="#lurija1968mind">Luria 151</a>).<sup id="fnref:18"><a class="footnote-ref" href="#fn:18">18</a></sup></p>
<p>Unusual stability is also common to Severian and Shereshevsky.<sup id="fnref:19"><a class="footnote-ref" href="#fn:19">19</a></sup> The
latter’s synesthetic associations remain very stable throughout his life
and enable him to recall memories from decades past (<a href="#lurija1968mind">Luria 12</a>), even into his early childhood (<a href="#lurija1968mind">Luria 76</a>). Similarly,
Severian’s character seems to undergo curiously little development while
his position in society and his political stance change radically and
repeatedly. What there is in terms of development can often be explained
by him gaining knowledge and applying it in a very immediate fashion.
The instances where Severian comments on his earliest memories, however,
are few and fragmentary, and due to this sparsity it is not clear
whether Shereshevsky’s were similar. In the first book Severian claims
“From my earliest memory I remember all. That first recollection is of
piling pebbles in the Old Yard” (<a href="#wolfe1980shadow"><em>Shadow</em> 16; ch. 2</a>). In the
second, however, he manages to penetrate further into his past; the
neonatal blur in the paragraph is very reminiscent of Shereshevsky
protocols of his earliest memories:</p>
<blockquote>
<p>I sought to recall that celebration of Holy Katharine’s day that fell
the year after I became captain of apprentices; but the preparations
for the feast were hardly begun before other memories came crowding
unbidden around it. In our kitchen I lifted a cup of stolen wine to my
lips—and found it had become a breast running with warm milk. It was
my mother’s breast then, and I could hardly contain my elation (which
might have wiped the memory away) at having reached back at last to
her, after so many fruitless attempts. My arms sought to clasp her,
and I would, if only I could, have lifted my eyes to look into her
face. My mother certainly, for the children the torturers take know no
breasts. The grayness at the edge of my field of vision, then, was the
metal of her cell wall. Soon she would be led away to scream in the
Apparatus or gasp in Allowin’s Necklace. I sought to hold her back, to
mark the moment so I might return to it when I chose; she faded even
as I tried to bind her to me, dissolving as mist does when the wind
rises. (<a href="#wolfe1981claw"><em>Claw</em> 384; ch. 27</a>)</p>
</blockquote>
<p>While serendipitous here, this paragraph also shows the difficulties
Severian and Shereshevsky encounter in trying to concentrate on one
train of thought and not being swept away by circumstantial
associations. <a href="#lurija1968mind">Luria (155)</a> writes:</p>
<blockquote>
<p>There were many instances too in which images that came to the surface
in S.’s mind steered him away from the subject of a conversation. At
such moments his remarks would be cluttered with details and
irrelevancies; he would become verbose, digress endlessly, and finally
have to strain to get back to the subject of the conversation.</p>
</blockquote>
<p><a href="#wright2003attending">Wright (110)</a> gives a lengthy example from
<a href="#wolfe1982sword"><em>Sword</em> (148; ch. 27)</a> where a red cape triggers a chain of
association so tenuous and arbitrary as to appear non sequiturs, but
which Severian presents as an argument for various fantastic empyrean
machinations behind the piece of cloth. Of these fantastic ruminations
there are several throughout the Urth Cycle.</p>
<p>Both mnemonists also show a surprising aptitude for seemingly unrelated
cognitive tasks. Severian solves several criminal mysteries through
surprising feats of ratiocination (e.g., Gunnie’s and Purn’s involvement
in the killing of his steward in <em>Urth</em>) and Shereshevsky is able to recognize
minute inconsistencies in stories and perform impressive calculations
through idiosyncratic processes of visualization (<a href="#lurija1968mind">Luria 102</a>).</p>
<p>Apart from these many striking parallels, there are also a number of
differences. With Shereshevsky, for example, the basis for his
exceptional memory is in his five-fold synesthesia; no synesthesia is
discernible in Severian’s case. Shereshevsky’s memory is perfectly
static even over decades; Severian’s memory shows curious alterations.</p>
<p>Before one can assemble these into a coherent whole, more parallels
between the two S.s have to be explored.</p>
<h3 id="the-marionette"><a class="toclink" href="#the-marionette">The Marionette</a></h3>
<p>The first sentence of the first chapter of the first book reads “It is
possible I already had some presentiment of my future”; the last
sentence of the same chapter reads “It was in this fashion that I began
the long journey by which I have backed into the throne.” Together they
summarize well Severian’s expectations, hindsight bias, and passivity.</p>
<p>One of Shereshevsky’s protocols from 1937 could be that of a Severian
lost in our time and forgotten by the hierogrammates.</p>
<blockquote>
<p>I read a great deal and always identified myself with one of the
heroes. For I saw them, you know. Even at eighteen I couldn’t
understand how one friend of mine was content to train to become an
accountant, another a commercial traveler. For what’s important in
life isn’t a profession but something fine, something grand that is to
happen to me… If at eighteen or twenty I’d thought I was ready to
marry and a countess or princess had agreed to marry me—even that
wouldn’t have impressed me. Perhaps I was destined for something
greater? … Whatever I did, whether writing articles, becoming a film
star—it was just a temporary thing.</p>
<p>At one point I studied the stock market, and when I showed that I had
a good memory for prices on the exchange, I became a broker. But it
was just something I did for a while to make a living. As for real
life—that’s something else again. But it all took place in dreams, not
in reality…</p>
<p>I was passive for the most part, didn’t understand that time was
moving on. All the jobs I had were simply work I was doing “in the
meantime.” The feeling I had was: “I’m only twenty-five, only
thirty—I’ve got my whole life ahead of me.” In 1917 I was content to
go off to the provinces. I decided to get in with the movement. So I
was in the Proletcult, ran a printing shop, became a reporter, lived a
special sort of life for a time. But even now I realize time’s passing
and that I might have accomplished a great deal—but I don’t work.
That’s the way I’ve always been. (<a href="#lurija1968mind">Luria 157–158</a>)</p>
</blockquote>
<p>Had Shereshevsky become the monarch of South America and later the
Jesus-like cynosure of that continent’s predominant religion, these
feelings would, in retrospect, have seemed presentient.</p>
<p>Having always, on some less rational level at least,<sup id="fnref:20"><a class="footnote-ref" href="#fn:20">20</a></sup> expected his
ascent, Severian goes along with it docilely and unquestioningly—and
that is the essence of what he does, too. There are few major decisions
by Severian that are not forced by circumstances or dictated by
authorities, and even these decisions, like allowing Thecla to die,
serve to mark the passing of the authoritative leash on Severian from
one authority to another (in this case from the masters of his guild to
Vodalus). When the decrees (or maybe “rescripts” (<a href="#wolfe1987urth"><em>Urth</em> 138; ch. 19</a>)) of the authorities diverge, Severian seems to simply follow the
one he perceives as greater rather than to decide for himself.</p>
<p>His upbringing has of course reinforced this docile nature, for the
torturers “carry out the sentences that are delivered to [them], doing
no more than [they] are told, and no less, and making no changes”
(<a href="#wolfe1980shadow"><em>Shadow</em> 81; ch. 12</a>).</p>
<p>Furthermore, he places great importance on symbols, which may either
render him susceptible to magical or mythical thought that relies on
symbols and to those who employ symbols as means of control, or which
may be a result or symptom of a pre-existing disposition to magical
thought. Maybe it is a mutually reinforcing cycle.</p>
<blockquote>
<p>Certain mysteries aver that the real world has been constructed by the
human mind, since our ways are governed by the artificial categories
into which we place essentially undifferentiated things, things weaker
than our words for them. I understood the principle intuitively that
night as I heard the last volunteer swing the gate closed behind
us.<sup id="fnref:21"><a class="footnote-ref" href="#fn:21">21</a></sup> (<a href="#wolfe1980shadow"><em>Shadow</em> 11; ch. 1</a>)</p>
</blockquote>
<p>It is certainly true that there are many specific “things weaker than
our words for them,” but that does not mean that all things are an
undifferentiated hodgepodge; Severian is probably underestimating
“things” here. Such constructivism has been criticized on various
grounds, crucially that of its self-refutation and related problems
(<a href="#boghossian2006fear">boghossian2006fear e.g., 66</a>).<sup id="fnref:22"><a class="footnote-ref" href="#fn:22">22</a></sup></p>
<p>However if a coin or a gem can alter one’s loyalties and attitudes so
dramatically, it is probably a reassuring illusion to believe that
nature is inherently so malleable or ambiguous so that it is not just
one’s conception of it that is being manipulated. He comments on these
tokens, or Vodalus’s coin in particular, a few pages later:</p>
<blockquote>
<p>We believe that we invent symbols. The truth is that they invent us;
we are their creatures, shaped by their hard, defining edges.
(<a href="#wolfe1980shadow"><em>Shadow</em> 14; ch. 1</a>)</p>
</blockquote>
<p>More concretely again, he later recites a verse that his older friend
Roche told him would keep hidden items from being discovered by others.
Severian uses it to protect the coin he had received from Vodalus and
observes that he “was somewhat astonished to discover that [he] was now
old enough not to be ashamed of it” (<a href="#wolfe1980shadow"><em>Shadow</em> 25; ch. 3</a>). His
concern here is with having overcome his adolescent fear of appearing
childish, which veils the age-independent superstition in trying to use
such a charm. He does, however, dismiss more inconvenient aspects of the
spell, showing a certain divide between what he feels and what he knows.
Then again, he shows less reflection when commenting on the “sensation
of being watched” (<a href="#wolfe1980shadow"><em>Shadow</em> 33; ch. 4</a>). The chapters are
riddled with similar episodes.</p>
<p>Another section highlights his internal conflict between the rational
appraisal and the distractive magical intuitions. In the Garden of
Endless Sleep, he considers a hyacinth<sup id="fnref:23"><a class="footnote-ref" href="#fn:23">23</a></sup> while the others who are
with him think, guessing from his circumstances and countenance, that he
is considering his own death.</p>
<blockquote>
<p>Is it possible the flower came into being only because Dorcas reached
for it? In daylight moments, I know as well as the next that such
things are impossible; but I am writing by night, ….
(<a href="#wolfe1980shadow"><em>Shadow</em> 147; ch. 24</a>)</p>
</blockquote>
<p>This propensity is not unknown to Shereshevsky:</p>
<blockquote>
<p>One time when I was planning to go to Samara, Misha [his son]
developed stomach pains. We called in a doctor, but he couldn’t figure
out what was wrong with him …. Yet it was so simple. I had given him
something that was cooked with lard. I could see the pieces of lard in
his stomach …. I thought to myself I’d help him. I wanted him to
digest them …. I pictured it in my mind and saw the lard dissolving in
his stomach. And Misha got better. Of course, I know this isn’t the
way it happened … yet I did see it all. (<a href="#lurija1968mind">Luria 144</a>)</p>
</blockquote>
<p>Luria summarizes this aspect as follows:</p>
<blockquote>
<p>With each individual there is a dividing line between imagination and
reality; for most of us whose imaginations have distinct limits, this
is fairly clear-cut. In S.’s case the borderline between the two had
broken down, for the images his imagination conjured up took on the
feel of reality. (<a href="#lurija1968mind">Luria 144</a>)</p>
</blockquote>
<p>The difference is that Shereshevsky was not, according to Luria’s
account, exposed to any manipulation that he would have to disguise from
himself. Severian is, and <a href="#wright2003attending">Wright</a> points particularly to
the Claw’s psilocybin-like psychedelic influence described at the very
end of the third volume:</p>
<blockquote>
<p>Seeing it thus without its case of sapphire, I felt profoundly an
effect I had never noticed at all during the days before it had been
taken from me in the hetman’s house. Whenever I looked at it, it
seemed to erase thought. Not as wine and certain drugs do, by
rendering the mind unfit for it, but by replacing it with a higher
state for which I know no name. Again and again I felt myself enter
this state, rising always higher until I feared I should never return
to the mode of consciousness I call normality; and again and again I
tore myself from it. Each time I emerged, I felt I had gained some
inexpressible insight into immense realities.</p>
<p>At last, after a long series of these bold advances and fearful
retreats, I came to understand that I should never reach any real
knowledge of the tiny thing I held, and with that thought (for it was
a thought) came a third state, one of happy obedience to I knew not
what, an obedience without reflection because there was no longer
anything to reflect upon, and without the least tincture of rebellion.
(<a href="#wolfe1982sword"><em>Sword</em> 200; ch. 38</a>)</p>
</blockquote>
<p>If you are tasked with destroying most life on Urth so that it can
re-emerge to evolve into something better (by the standards of the
hierogrammates), then “obedience without reflection” is surely
inevitable lest the magnitude of the risks and uncertainties become
apparent. The claw lends Severian the comfort of the illusion that he is
integral in something overwhelmingly grand and good, and as the old
autarch told him, “You came for pleasure, did you not? If a dream adds
to your enjoyment, why dispute it?”<sup id="fnref:24"><a class="footnote-ref" href="#fn:24">24</a></sup></p>
<p>Another way to dismiss the global near-omnicide and all considerations
of necessity, proportionality, and effectuality is to dismiss
consequentialism in favor of some brand of moral philosophy that
concerns itself with intentions only—which of course are good so long as
obedience to a perceived god is considered incontrovertibly good: “Until
we reach the end of time, we don’t know whether something’s been good or
bad; we can only judge the intentions of those who acted.”
(<a href="#wolfe1987urth"><em>Urth</em> 237; ch. 33</a>)</p>
<h2 id="reconciliation"><a class="toclink" href="#reconciliation">Reconciliation</a></h2>
<p>Given that in the Urth Cycle “everything … has three meanings”
(<a href="#wolfe1980shadow"><em>Shadow</em> 190; ch. 32</a>) and seeing the ambiguity that
facilitates such versatility, great caution is necessary when forming
sweeping theories of the text. In its vast repository of facts it is too
easy to let ideology blind oneself to gaps and contradictions. As a
preemptive countermeasure, this section will try to introduce its
hypotheses with all their qualities and shortcomings in so far as they
were apparent to the author.</p>
<p>These hypotheses will concern the parallels and differences between
Shereshevsky and Severian. Though the true relationships of cause and
reaction may be more convoluted, it appears that for Shereshevsky his
synesthesia was integral to his mnemonic abilities, which in turn caused
his cognitive limitations due to the noise he had to consciously sift
through in order to recognize any patterns. So how can Severian perform
such similar feats and be subject to such similar limitations with no
noticeable synesthesia? And how come that some of his memories change
when Shereshevsky was so notable for his prefect recall even after decades?</p>
<h3 id="the-man-your-mother-bore"><a class="toclink" href="#the-man-your-mother-bore">The Man Your Mother Bore</a></h3>
<p><em>Urth</em> gave away many a secret that people like <a href="#Kincaid2004">Kincaid</a> would
have liked to figure out or had figured out themselves. One of them is
Severian’s nature as aquastor. In a particularly revealing conversation with the
hierodules<sup id="fnref:25"><a class="footnote-ref" href="#fn:25">25</a></sup> Barbatus, Famulimus, and Ossipago, this is made explicit.</p>
<blockquote>
<p>“But if Apu-Punchau is myself, what was the body I found on Tzadkiel’s ship?”</p>
<p>Nearly whispering, Famulimus sang, “The man whom you saw dead your
mother bore. Or so it seems to me from what’s been said. Now I would
weep for her if I had tears, though not—perhaps—for you still living
here. What we did here for you, Severian, the mighty Tzadkiel
accomplished there, remembrance taking from your dead mind to build
your mind and you anew.” (<a href="#wolfe1987urth"><em>Urth</em> 359; ch. 50</a>)</p>
</blockquote>
<p>She is of course wrong in the sense that Severian had died and was
rebuilt in much the same fashion several times<sup id="fnref:26"><a class="footnote-ref" href="#fn:26">26</a></sup> before that
incident, but had they been talking about Severian’s death by drowning
in Gyoll that starts the narrative, then she might have been right.</p>
<p>Much earlier, the aquastor of Master Malrubius, created from Severian’s
childhood memories of him, explains his nature thus: “Once you met a
woman named Cyriaca, who told you tales of the great thinking machines
of the past. There is such a machine on the ship in which we sailed. …
But we are maintained in the physical world by the energies of the
machine, and its range is but a few thousand years.”</p>
<p>This immediately suggests that Severian, as aquastor, may be able to
draw upon the storage capacities of a computer, maybe the ship computer
of the Matachin Tower, in order to augment his memory, and the
interaction of these memories with his biological body may introduce
occasional inaccuracies.</p>
<p>Furthermore if Severian was created as aquastor not just after his first
death but from his birth, created according to the designs of the
hierogrammates or their subordinates, then he could have been born
without the help of a father, though some gruesome scenes in
Baldanders’s castle suggest that a female host was necessary to create
Dr. Talos (<a href="#wolfe1982sword"><em>Sword</em> 176; ch. 33</a>). This would also underline the
Christian mythological imagery with Severian being born to a virgin
thanks to advanced alien technology.</p>
<p>An alternative would be that he only became an aquastor a few years
later in his life, an event that would, in retrospect, be marked by his
first memory (“That first recollection is of piling pebbles in the Old
Yard” (<a href="#wolfe1980shadow"><em>Shadow</em> 16; ch. 2</a>)).</p>
<p>The text does not seem to provide much evidence for the theory of
Severian’s ontogenesis by design to the degree Dr. Talos was designed;
Ouen is rather assuredly Severian’s biological father; and furthermore
Severian’s mother Catherine may have taken the vows of abstinence of the
Pelerines, thus giving her a motive to lie about Ouen’s fatherhood and
outfitting the appropriately subverted allegory in any case.</p>
<p>However, the second theory does hold some appeal as the Old Yard, either
by name or by implication, is repeatedly invoked in dying visions or
dreams of Severian like an echo of an early, unremembered trauma.<sup id="fnref:27"><a class="footnote-ref" href="#fn:27">27</a></sup>
Furthermore, it is where Severian is shot and probably killed once again
in ancient times resulting in the breach of the curtain wall
(<a href="#wolfe1987urth"><em>Urth</em> 255; ch. 36</a>). Finally, the Old Yard has been the site
of executions and probably excruciations,<sup id="fnref:28"><a class="footnote-ref" href="#fn:28">28</a></sup> and is located close to
the Bell Tower with its Bell Keep the apprentices are forbidden to enter
for unknown reasons, so it may be a rather hazardous place.</p>
<p>But the evidence for a first death when Severian was barely old enough
to stack pebbles is thin and as a means of explaining his extraordinary
memory fails to account for one important phrase in the cycle.</p>
<h3 id="our-own-shadows-race-into-the-past"><a class="toclink" href="#our-own-shadows-race-into-the-past">Our Own Shadows Race Into the Past</a></h3>
<p>“I don’t forget much,” (<a href="#wolfe1983citadel"><em>Citadel</em> 402; ch. 37</a>)
is how Ouen explains that he has learned reading and
writing without formal education. Ouen being Severian’s biological
father, this suggests that Severian has something akin to an inherited
predisposition for great memory.</p>
<p>But that does not yet explain the memory itself. Absent any solid cause
and grounding of his gift, such as highly developed synesthesia or the
hard drives of a ship computer, it seems necessary to assume that
Severian’s memory is no better than average.</p>
<p>In the same conversation with the familiar trio of hierodules, Severian
learns that his ability to time-travel—or to walk the Corridors of
Time—is dependent on the energy of the star he identifies with and that
he cannot draw on it in the day of Apu-Punchau (until after the death of
that version of him) because the light of the star had not yet reached
Urth. It should be noted that—with the exception of his decade-long
sojourn as Apu-Punchau among the autochthons of the stone town—he can
draw on this source of energy throughout his whole life.</p>
<p>There are also many indications that his awareness of this energy source
maybe aids his conscious control over it but that it can work quite
independent of this awareness. In a different context Severian explains:</p>
<blockquote>
<p>I did not know that then, but it is a profound mistake to believe that
we must know of such things to be influenced by them, and in fact to
believe so is to believe in the most debased and superstitious kind of
magic. The would-be sorcerer alone has faith in the efficacy of pure
knowledge; rational people know that things act of themselves or not
at all.<sup id="fnref:29"><a class="footnote-ref" href="#fn:29">29</a></sup> (<a href="#wolfe1980shadow"><em>Shadow</em> 14; ch. 1</a>)</p>
</blockquote>
<p>Examples are the countless cases where he resurrects beings—starting
with Triskele before he receives the Claw—but continually misattributes
the causes.</p>
<p>Dorcas, also attributing the phenomena to the Claw, gives a parsimonious
explanation of the workings behind the events even though she does not
even consciously know that Severian would attain the ability to time-travel:</p>
<blockquote>
<p>“Severian, when you brought the uhlan back to life it was because the
Claw twisted time for him to the point at which he still lived. When
you half healed your friend’s wounds, it was because it bent the
moment to one when they would be nearly healed. And when you fell into
the fen in the Garden of Endless Sleep, it must have touched me or
nearly touched me, and for me it became the time in which I had lived,
so that I lived again. …” (<a href="#wolfe1982sword"><em>Sword</em> 60; ch. 11</a>)</p>
</blockquote>
<p>Severian later lends extra emphasis to this statement by recounting
himself retelling the episode verbatim to Miles (<a href="#wolfe1983citadel"><em>Citadel</em> 217; ch. 3</a>).</p>
<p>With Severian’s unusual ability to resurrect reduced to a manipulation
of time, it becomes even more parsimonious to explain his unusual memory
in the same terms. Not physically transferring himself into the other
time but just bending time for his consciousness now to meet his
consciousness then might require much less skill, focus, and
energy,<sup id="fnref:30"><a class="footnote-ref" href="#fn:30">30</a></sup> so that he could master it at a much younger age and still
several resurrections away from the Severian who would go on to become
the Conciliator.</p>
<p>The inaccuracies of his memory may then be grouped into three categories
in order to explain them in terms of this system. First, it is possible
that from a certain present different pasts whose differences have
proved inconsequential for the present are are blurred to a degree and
thus hard to distinguish for the time traveler. Hence Severian’s
different versions of the pistol hand-off in the Necropolis, where, in
both cases, the pistol soon ended up in Thea’s hands. Second, Severian
may be loath to enter into the trance-like state that his casting back
into the past entails for his present body, especially when he had just
done so and still thinks that he remembers (with his ordinary episodic
memory) what he saw in his transtemporally enhanced remembrance, so that
he misremembers facts just as anyone would, for example, who spotted the
pikes first. Third, his ordinary semantic recall may be subject to
production errors. A moment’s reflection would have convinced him that
there is no way Dorcas would have used human skin for his little sack,
but since the sheath of Terminus Est was fashioned from something called
“sable manskin” (<a href="#wolfe1980shadow"><em>Shadow</em> 90; ch. 14</a>) and both were items very
dear to him that he carried with him across the Commonwealth, he might
have momentarily mixed them up.</p>
<p>The fluidity of the dividing line between this travel in his imagination
and actual time travel is exemplified when he dreams on the shores of
Ushas, is told in his dream that he no longer dreamed, and soon emerges,
wholly physical, as Apu-Punchau in spe. When he returns to the
approximate time of his departure, he learns that a wraith-like aspect
of his had the whole time remained where he had slept, one that the
priest of this new religion of Ushas professed to be able to sense.</p>
<p>According to this theory, Ouen’s good memory may be not so much the
genetic precursor and foundation for Severian’s gift but a latent
ability of his to access the energies of the new sun due to his close
relation to Severian. A gift handed back in time. Such notions are not
foreign to the text, where Dr. Talos observes that “just as the
momentous events of the past cast their shadows down the ages, so now,
when the sun is drawing toward the dark, our own shadows race into the
past to trouble mankind’s dreams.”<sup id="fnref:31"><a class="footnote-ref" href="#fn:31">31</a></sup> (<a href="#wolfe1982sword"><em>Sword</em> 186; ch. 35</a>)</p>
<p>Symmetrically, the same counterchronological influence may be the reason
for Valeria’s association with the Atrium of Time. <a href="#borski2004solar">Borski (1–9)</a> presents a convincing argument that two of the women Severian
has a relationship with, Dorcas and Valeria, are both his grandmothers.
This is clear in the case of Dorcas, but there is also ample evidence
(detailed in <em>Solar Labyrinth</em>) that Severian’s mother Catherine is the daughter of
Valeria and her second husband Dux Caesidius, and that it is her who is
sent back in time on the final day of Urth.</p>
<p>Further evidence of this theory is that the period Severian lives as
Apu-Punchau—the only one, as noted earlier, when he cannot draw on the
energy of the new sun—is kept very short despite its many years in story
time and contains no direct speech until right before the end when he
can escape to the future again. Severian has not scrupled to render the
language of the Ascians in his own in his account, so the foreign
language they speak is not likely to be the reason for the stylistic
break, and the content of the two chapters is close to what someone with
ordinary memory would be able to remember of the events.</p>
<p>Conversely, there are several unexplained visions of the past embedded
in the cycle, which <a href="#andre2008lexicon">Andre-Driussi (93)</a> has compiled under the
headword “Corridors of Time,” thereby implying the same conflation of
memory and time travel that may be at the basis of Severian’s mnemonic
ability. At least two of these visions take place before he has obtained
the Claw and one of them clearly reaches back much further than his own lifetime.</p>
<p>What further corroborates the theory is that it repeats the pattern of
annular fusion, the circular or mutual recursion that is found all
throughout the Urth Cycle, be it in the schemes of the hierogrammates
who want to ensure their own creation by ensuring the genesis of the
humans of Ushas who will evolve into the hieros who created the
hierogrammates, or be it in Canog’s <em>Book of the New Sun</em>, which serves
as blueprint for Dr. Talos’s play <em>Eschatology and Genesis</em>, which
Severian uses to guide his account of the future that Canog protocols
and eventually turns into the first book. Severian’s mnemonic ability is
the reason he is enlisted by the hierogrammates and his mission
eventually is the counterchronological cause of the same ability.<sup id="fnref:32"><a class="footnote-ref" href="#fn:32">32</a></sup></p>
<p>Father Inire may have the same gift as Severian. <a href="#borski2004solar">Borski (43–70)</a>
makes a good case for Inire reappearing throughout the cycle
as a father figure in various guises. He is usually distinguished by
this stature, size, abilities, manner of walking, manner of social
interaction, and significantly his impressive eyes, either directly or
in that he tries to hide them.</p>
<p>The ostensible hierodule and machine Ossipago is likely to be one of
these guises. He is the one who enables the other two hierodules of the
trio to step through time (<a href="#wolfe1987urth"><em>Urth</em> 360; ch. 50</a>), and
reduplicating himself throughout time is probably also the way Inire
manages to stay “alive so long beyond the span of his short-lived
kind.”<sup id="fnref:33"><a class="footnote-ref" href="#fn:33">33</a></sup> (<a href="#wolfe1983citadel"><em>Citadel</em> 405; ch. 38</a>) So when Famulimus says that
“Only Ossipago here has memory like yours” (<a href="#wolfe1987urth"><em>Urth</em> 405; ch. 38</a>),
she may have meant memory that has transtemporal access to all
brain states all throughout the persons lifetime.<sup id="fnref:34"><a class="footnote-ref" href="#fn:34">34</a></sup></p>
<h3 id="for-that-were-you-chosen"><a class="toclink" href="#for-that-were-you-chosen">For That Were You Chosen</a></h3>
<p>Some previous sections have already pointed toward the enhanced
manipulability through myths and symbols and scintillating gadgets that
may be a concomitant of a memory that does not filter any noise, but the
hierodules Barbatus, Famulimus, and Ossipago imply that Severian’s
memory capacity itself may have been an even more immediate and forcible
reason,<sup id="fnref:35"><a class="footnote-ref" href="#fn:35">35</a></sup> but not necessarily because of any record he would write,
though <em>lethe</em><sup id="fnref:36"><a class="footnote-ref" href="#fn:36">36</a></sup> suggests that interpretation.</p>
<blockquote>
<p>Barbatus’s pleasant baritone flouted the gloom. “You’re conscious.
What do you remember?”</p>
<p>“Everything,” I said. “I’ve always remembered everything.” Dissolution
was in the air, the fetor of [Severian’s own former body’s] rotting flesh.</p>
<p>Famulimus sang, “For that were you chosen, Severian. You and you alone
from many princes. You alone to save your race from lethe.”
(<a href="#wolfe1987urth"><em>Urth</em> 353; ch. 50</a>)</p>
</blockquote>
<p>This description, as well as previous, identical descriptions of the
nature of aquastors, suggests that the accurate recreation of a person
is aided by good memory. Anticipating<sup id="fnref:37"><a class="footnote-ref" href="#fn:37">37</a></sup> that their marionette would
die repeatedly on his journeys, the hierogrammates hence selected
someone they could recreate almost losslessly from his own memories.</p>
<p>That is, if the hierogrammates are even the directors of the play.
Father Inire, genius architect and adviser to all autarchs over maybe
1,000 years, is a formidable contender for the role, but as hierodule,
he may be just the subcontracted architect of the hierogrammates.
Finally, Sergei Novikov may also be behind it all, seeing how the future
may dictate, through the self-consistency principle, all the less likely
events of the past that caused it. But such speculations would go beyond
the scope of this paper.</p>
<h2 id="conclusion"><a class="toclink" href="#conclusion">Conclusion</a></h2>
<p>First, the reader of <em>The Book of the New Sun</em> gets to experience the
overwhelming deluge of data and associations that cause mnemonists like
Shereshevsky and Severian to perceive much less clearly patterns that
would be obvious to most people, a condition that also seems to entail
passivity and, with Severian, docility, the latter of which is
reinforced by his upbringing. These processes are detailed by reference
to the research of <a href="#lurija1968mind">Luria</a>.</p>
<p>Second, and more fundamentally, there is the search for the basis of
Severian’s condition. Cause and effect cannot be assumed to be clear-cut
in a medium as interconnected (or curiously disconnected) as our brain,
but to the reader of Shereshevsky’s case history it would seem like his
synesthesia is the foundation for his mnemonic feats, while the latter
are the cause of some of his peculiar character traits. The absence of
any such synesthesia in Severian forms a conspicuous lacuna—a clear sign
that it is up to the reader to interpolate.</p>
<p>One argument that reconciles these similar symptoms with their
dissimilar causes draws on Severian’s latent ability to navigate time,
which is only fully revealed in the coda of the cycle. Various fragments
of evidence can be marshaled from the text to substantiate this hypothesis.</p>
<h2 id="endnotes"><a class="toclink" href="#endnotes">Endnotes</a></h2>
<div class="footnote">
<hr>
<ol>
<li id="fn:1">
<p>Somehow it is rarely called a <em>sequel</em>, maybe to express its
finality or because of <em>coda</em>’s more lupine etymology. <a class="footnote-backref" href="#fnref:1" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
<li id="fn:2">
<p>Assuming G.W. is an intradiegetic version of Gene Wolfe. <a class="footnote-backref" href="#fnref:2" title="Jump back to footnote 2 in the text">↩</a></p>
</li>
<li id="fn:3">
<p>It could be argued that just as Severian writes his autobiography,
so the hierogrammates have written the story of his life, all of
which, in turn, is the work of Gene Wolfe. <a class="footnote-backref" href="#fnref:3" title="Jump back to footnote 3 in the text">↩</a></p>
</li>
<li id="fn:4">
<p>For a much more expansive summary, see the first appendix to <a href="#andre2008lexicon"><em>Lexicon Urthus</em> (389)</a>.
That summary, however, is focalized on Severian at the respective
times. <a class="footnote-backref" href="#fnref:4" title="Jump back to footnote 4 in the text">↩</a></p>
</li>
<li id="fn:5">
<p>A “hidden house … everywhere coextensive with the public one”
(<a href="#wolfe1981claw"><em>Claw</em> 333; ch. 20</a>) and surely an allusion to the Urth
Cycle itself. The secret house is a product of the architectural
genius of the vizier to all autarchs for the past millennium, Father
Inire. <a class="footnote-backref" href="#fnref:5" title="Jump back to footnote 5 in the text">↩</a></p>
</li>
<li id="fn:6">
<p>According to the Kabbalah, Yesod is a sefiroth of Yetzirah, a
lower created world than Briah, but maybe it is lower along one
dimension and higher along another. <a href="#UrthJordan1998">Jordan</a> <a class="footnote-backref" href="#fnref:6" title="Jump back to footnote 6 in the text">↩</a></p>
</li>
<li id="fn:7">
<p>There are indications that time in Yesod runs counter to time in
Briah, so that the hierogrammates may not so much be trying to
ensure their procreation in the next Briahtic manvantara but their
own birth. <a class="footnote-backref" href="#fnref:7" title="Jump back to footnote 7 in the text">↩</a></p>
</li>
<li id="fn:8">
<p>Which has led reviewers to miss the fact that the narrator of <em>Peace</em> is
dead although it is said outright on page 17. <a href="#hall1981interview">Hall</a> <a class="footnote-backref" href="#fnref:8" title="Jump back to footnote 8 in the text">↩</a></p>
</li>
<li id="fn:9">
<p><a href="#wood2008fiction">Wood (5)</a> makes a distinction between reliably
unreliable and unreliably unreliable narration. The Urth Cycle may
well fall into the latter category for there are many themes, such
as roses or gold, that indicate important relationships in the text
that Severian has probably remained unaware of. In some cases, these
same themes are then used to seemingly lead the critics astray
throughout most of the text and mock them subtly in the last volume. <a class="footnote-backref" href="#fnref:9" title="Jump back to footnote 9 in the text">↩</a></p>
</li>
<li id="fn:10">
<p>The same evidence that <a href="#wright2003attending">Wright</a> enlists to show how
closely all of Severian’s life has been stage-managed by the
hierogrammates, centrally many episodes where parts of the story are
staged in real or dreamed theater plays, could also be read to
indicate that Severian’s career itself was very literally a play to
entertain the hierogrammates without all the existential
significance that Apheta imbues it with. Yesod certainly feels like
a stage with its house specifically constructed for Severian’s trial
and its trapdoors that lead to backstage areas with giant fly lofts
and backstage exits from the planet (<a href="#wolfe1987urth"><em>Urth</em> 164–169; ch. 23</a>).
Apheta reveals in the same chapter that visiting the surface of
the planet is a rare reward for them who labor inside the planet.
Gunnie then references Dante’s <em>Inferno</em> just as they are about to
exit it again, maybe a hunch she has about the pleasure-seeking
disregard the hierogrammates might have for the lives of their
actors. Such an interpretation would also turn much subtextual
“writerliness”—such as etymology and history linking the names of
characters or themes of roses and gold linking families—from an
extradiegetic influence by Wolfe into an intradiegetic influence of
the hierogrammates. <a class="footnote-backref" href="#fnref:10" title="Jump back to footnote 10 in the text">↩</a></p>
</li>
<li id="fn:11">
<p>Flaying of the leg is described as inducing a “slow, generalized
welling of blood” (<a href="#wolfe1980shadow"><em>Shadow</em> 23; ch. 3</a>), but any form of
torture would require a judicial decree, and in the ostensible story
at least, the guild tries to cover up their internal issue so not to
lose the trust of the Autarch’s courts. Then Severian’s wound might
be self-inflicted. <a class="footnote-backref" href="#fnref:11" title="Jump back to footnote 11 in the text">↩</a></p>
</li>
<li id="fn:12">
<p>The curious order of the words in the title may indicate the
perspective of the hierogrammates and hierodules for whom time runs
in the opposite direction, or <em>Death</em> actually refers to Severian’s
first homicide in the chapter, rendering him in effect Death for the
first time, a name and role he would subsequently assume with
frequency and varying degrees of willingness. There is another
related, collocationally surprising but apposite inversion in the
title of the play that would serve as blueprint for Canog’s <em>Book of
the New Sun</em>, <em>Eschatology and Genesis</em>. <a class="footnote-backref" href="#fnref:12" title="Jump back to footnote 12 in the text">↩</a></p>
</li>
<li id="fn:13">
<p>It is never called <em>eidetic</em> but maybe just because the term
might have been anachronistic for Severian. It was coined by Erich
Jaensch in the 1920s. (<a href="#OEtyDEidetic">OEtyDEidetic</a>) <a class="footnote-backref" href="#fnref:13" title="Jump back to footnote 13 in the text">↩</a></p>
</li>
<li id="fn:14">
<p>Compare, for example, <a href="#vargha1994agnosia">Vargha-Khadem, Isaacs, and Mishkin</a>. <a class="footnote-backref" href="#fnref:14" title="Jump back to footnote 14 in the text">↩</a></p>
</li>
<li id="fn:15">
<p><a href="#wright2003attending">Wright (108)</a> notes that “as Wolfe took ‘beginning
and advanced courses in Abnormal Psychology’ at Miami University in
Ohio in the late 1960s, and an introductory course at the University
of Houston, it is probable that he was familiar with the clinical
studies of the psychology of mnemonists when he began writing <em>The
Book of the New Sun</em>.” <a class="footnote-backref" href="#fnref:15" title="Jump back to footnote 15 in the text">↩</a></p>
</li>
<li id="fn:16">
<p>An ineffectual pseudonymization not only because Shereshevsky was
a well-known mnemonist but also because both his first names are
mentioned in the text (<a href="#lurija1968mind">Luria 50</a>). <a class="footnote-backref" href="#fnref:16" title="Jump back to footnote 16 in the text">↩</a></p>
</li>
<li id="fn:17">
<p>He is yet unaware of having died and having been recreated as
aquastor. <a class="footnote-backref" href="#fnref:17" title="Jump back to footnote 17 in the text">↩</a></p>
</li>
<li id="fn:18">
<p>The parallels between the mechanisms behind Shereshevsky’s alter
ego—who does onerous things Shereshevsky himself is loath to do
(<a href="#lurija1968mind">Luria 153</a>) or behaves in socially awkward manners so
Shereshevsky gets spared the embarrassment (<a href="#lurija1968mind">Luria 156</a>)—and
Severian’s alzabo-induced legion of them seem merely nominal. <a class="footnote-backref" href="#fnref:18" title="Jump back to footnote 18 in the text">↩</a></p>
</li>
<li id="fn:19">
<p>It should be fair to call Severian’s memory <em>unusually</em> stable
despite its slight mutability. <a class="footnote-backref" href="#fnref:19" title="Jump back to footnote 19 in the text">↩</a></p>
</li>
<li id="fn:20">
<p>Maybe it can be imagined as a feeling akin to Romain Rolland and
Sigmund Freud’s so-called oceanic feeling, only with the additional
feature that it puts the subject in some key position. Such a
similarity would also provide a basis to explain the Romantic themes
that Severian perceives. <a class="footnote-backref" href="#fnref:20" title="Jump back to footnote 20 in the text">↩</a></p>
</li>
<li id="fn:21">
<p>Another example may be the way Severian often conflates
ocean-going and space-going ships while to us they are starkly
different. <a class="footnote-backref" href="#fnref:21" title="Jump back to footnote 21 in the text">↩</a></p>
</li>
<li id="fn:22">
<p>Severian’s musings are also reminiscent of the debate over
whether language determines thought or vice versa, which, on account
of the intertwined, often inseparable nature of the two concepts,
probably has to be answered differently. <a class="footnote-backref" href="#fnref:22" title="Jump back to footnote 22 in the text">↩</a></p>
</li>
<li id="fn:23">
<p>However, this thought sequence may have several more meanings. In
“The fiacre drew up to her with the skittish animals dancing to one
side as though she were a thyacine” (<a href="#wolfe1980shadow"><em>Shadow</em> 112; ch. 18</a>),
Severian indirectly and counterfactually likens Agia to a thyacine.
He, however, stands close behind her in the scene, so it is possible
that they shied away from him as much as from her. This
interpretation is corroborated by the fact that <em>thyacine</em> is a
misspelling of <em>thylacine</em>, “the native Tasmanian ‘wolf’ or
‘zebra-wolf’” (<a href="#andre2008lexicon">Andre-Driussi 348</a>), and, obvious extradiegetic
associations aside, the wolf is typically associated with Severian
(<a href="#andre2008lexicon">Andre-Driussi 17</a>). Hence the thought sequence about the
hyacinth may mirror Severian’s possible drowning or near-drowning in
the lake, since Dorcas saved him by plugging him from the water (and
from the other hand that tried to pull him down) just as she plugged
the hyacinth moments later. In an inversion of it, it must have
seemed to amnesiac Dorcas as if <em>she</em> came into existence to save
Severian, when really the saving had been mutual. When Dorcas
finally insists that he had been thinking about dying, Severian does
not again contradict her. Furthermore, thanks to the misspelling
(the only appearance of the animal in the books), the word shares a
sequence of six characters with <em>hyacinth</em>, a common type of hint in
the cycle, which, in this case, may have the additional function of
intimating Agia’s family bonds with Severian (<a href="#borski2004solar">Borski 10–17</a>).
It should also be noted that <em>Hyacinth</em> is the name of the
wife of the protagonist of Wolfe’s <em>Book of the Long Sun</em>, who (the
protagonist) is subtly linked to Severian. <a class="footnote-backref" href="#fnref:23" title="Jump back to footnote 23 in the text">↩</a></p>
</li>
<li id="fn:24">
<p>The quotation is of course taken out of context, but then again
an earlier section has already drawn attention to the similarity
between the House Azure and the reality of Urth as we might see were
it not for Severian’s consciousness and the translation filtering
our perception. Then the quotation might not be out of context after
all and might even be directed not only at Severian but at the
reader, reminding the reader to consider dismissing the very thought
they have at that instant. <a class="footnote-backref" href="#fnref:24" title="Jump back to footnote 24 in the text">↩</a></p>
</li>
<li id="fn:25">
<p>Though Ossipago may be a machine or Father Inire in disguise. <a class="footnote-backref" href="#fnref:25" title="Jump back to footnote 25 in the text">↩</a></p>
</li>
<li id="fn:26">
<p>Assuredly once but likely more often. <a class="footnote-backref" href="#fnref:26" title="Jump back to footnote 26 in the text">↩</a></p>
</li>
<li id="fn:27">
<p>“I knew he was looking for me in the Old Yard below.”
(<a href="#wolfe1980shadow"><em>Shadow</em> 12; ch. 2</a>) “The deepest bell in the Bell Tower
was ringing.” (<a href="#wolfe1987urth"><em>Urth</em> 65; ch. 9</a>) <a class="footnote-backref" href="#fnref:27" title="Jump back to footnote 27 in the text">↩</a></p>
</li>
<li id="fn:28">
<p>“By ancient custom, we must not use the steps (although I have
seen Master Gurloes assist his vault to the scaffold with his sword,
in the court before the Bell Tower).” (<a href="#wolfe1981claw"><em>Claw</em> 234; ch. 4</a>)
“It’s no more than it seems, just a stake to immobilize the hands,
and a thirteen-thonged scourge for correction. It used to stand in
the Old Yard, but the witches complained, and the castellan made us
move it down here.” (<a href="#wolfe1980shadow"><em>Shadow</em> 81; ch. 12</a>) <a class="footnote-backref" href="#fnref:28" title="Jump back to footnote 28 in the text">↩</a></p>
</li>
<li id="fn:29">
<p>While his conclusion is surely correct in most cases, the
argument is invalid, since the proposition that things can have an
effect without our knowledge of them does not imply that our
knowledge of them is necessarily without effect, but this is
unrelated to the argument of this section. <a class="footnote-backref" href="#fnref:29" title="Jump back to footnote 29 in the text">↩</a></p>
</li>
<li id="fn:30">
<p>Barbatus suggests that the amount of energy at one’s disposal is
crucial for determining the number of people one can ferry through
time. (<a href="#wolfe1987urth"><em>Urth</em> 361; ch. 50</a>) <a class="footnote-backref" href="#fnref:30" title="Jump back to footnote 30 in the text">↩</a></p>
</li>
<li id="fn:31">
<p>“Dreams,” here, can be taken quite literally since the fight in
the ensuing chapters had been foreshadowed in one of Severian’s
dreams in the first volume. <a class="footnote-backref" href="#fnref:31" title="Jump back to footnote 31 in the text">↩</a></p>
</li>
<li id="fn:32">
<p>Or at least why he ended up as the successful candidate if
Baldanders was a previous one but was found to be too egotistical
and ambitious for the job. <a class="footnote-backref" href="#fnref:32" title="Jump back to footnote 32 in the text">↩</a></p>
</li>
<li id="fn:33">
<p>Hierodules typically “live only a score of years, like dogs.”
(<a href="#wolfe1982sword"><em>Sword</em> 178; ch. 33</a>) <a class="footnote-backref" href="#fnref:33" title="Jump back to footnote 33 in the text">↩</a></p>
</li>
<li id="fn:34">
<p>If instead Ossipago is really only a machine rather than a
machine with Father Inire inside, then this statement would seem to
corroborate the theory that Severian’s memory is owed to
cloud-sourced storage augmentations. <a class="footnote-backref" href="#fnref:34" title="Jump back to footnote 34 in the text">↩</a></p>
</li>
<li id="fn:35">
<p>And at least in dismissing Severian’s faith in the Claw, they
have been remarkably (surprisingly) honest with him
(<a href="#wolfe1982sword"><em>Sword</em> 184; ch. 34</a>). <a class="footnote-backref" href="#fnref:35" title="Jump back to footnote 35 in the text">↩</a></p>
</li>
<li id="fn:36">
<p>“The river of oblivion, one of the streams of Hades, the waters
of which possessed the quality of causing those who drank of them to
forget their former existence.” <a href="#wordnikLethe">The Century Dictionary and Cyclopedia</a> <a class="footnote-backref" href="#fnref:36" title="Jump back to footnote 36 in the text">↩</a></p>
</li>
<li id="fn:37">
<p>Trivial for beings who can contravene time or whose “clocks run
widdershins round both … suns” (<a href="#wolfe1987urth"><em>Urth</em> 360; ch. 50</a>) by
design. <a class="footnote-backref" href="#fnref:37" title="Jump back to footnote 37 in the text">↩</a></p>
</li>
</ol>
</div>
<h2 id="works-cited"><a class="toclink" href="#works-cited">Works Cited</a></h2>
<p><a name="andre2008lexicon"></a>
Andre-Driussi, Michael. Lexicon Urthus. Second Edition. Sirius Fiction, 2008. Print.</p>
<p><a name="boghossian2006fear"></a>
Boghossian, Paul. Fear of Knowledge: Against Relativism and Constructivism. Clarendon Press, 2006. Print.</p>
<p><a name="borski2004solar"></a>
Borski, Robert. Solar Labyrinth: Exploring Gene Wolfe’s Book of the New Sun. iUniverse, 2004. Print.</p>
<p><a name="bridge2012neural"></a>
Bridge, Donna J. and Ken A. Paller. “Neural Correlates of Reactivation and Retrieval-Induced Distortion.” The Journal of Neuroscience 32.35 (2012): 12144–12151. Print.</p>
<p><a name="budrys2012benchmarks"></a>
Budrys, Algis. Benchmarks Continued 1975–1982. Ansible Editions, 2012. Print. Benchmarks Continued: The F <span class="amp">&</span> <span class="caps">SF</span> “Books” Columns.</p>
<p><a name="gordon1986gene"></a>
Gordon, Joan. Gene Wolfe. Borgo Press, 1986. Web. Starmont Reader’s Guide.</p>
<p><a name="Greenland1982"></a>
Greenland, Colin. “Review of The Shadow of the Torturer and The Claw of the Conciliator.” Foundation: The International Review of Science Fiction 24 (1982). Print.</p>
<p><a name="hall1981interview"></a>
Hall, Melissa Mia. “An Interview with Gene Wolfe.” 1981. Web. 10 Sept. 2014.</p>
<p><a name="OEtyDEidetic"></a>
Harper, Douglas. “Online Etymology Dictionary: Eidetic.” 2014. Web. 22 Aug. 2014.</p>
<p><a name="johnson2007phantom"></a>
Johnson, Carroll B. “Phantom Pre-texts and Fictional Authors: Sidi Hamid Benengeli, Don Quijote and the Metafictional Conventions of Chivalric Romances.” Cervantes: Bulletin of the Cervantes Society of America 27.1 (2007): 179–200. Print.</p>
<p><a name="UrthJordan1998"></a>
Jordan, Jim. “Three-Tiered Universes.” 1998. Web. 07 Sept. 2014.</p>
<p><a name="Kincaid2004"></a>
Kincaid, Paul. “Review of ‘Attending Daedalus: Gene Wolfe, Artifice and the Reader by Peter Wright’.” Foundation: The International Review of Science Fiction 92 (2004). Print.</p>
<p><a name="lurija1968mind"></a>
Luria, Alexander Romanovich. The Mind of a Mnemonist: A Little Book About a Vast Memory. Harvard University Press, 1968. Print.</p>
<p><a name="wordnikLethe"></a>
The Century Dictionary and Cyclopedia. “Lethe – Definition and Meaning – Wordnik.” 2014. Web. 12 Sept. 2014.</p>
<p><a name="vargha1994agnosia"></a>
Vargha-Khadem, Faraneh, Elizabeth Isaacs, and Mortimer Mishkin. “Agnosia, Alexia and a Remarkable Form of Amnesia in an Adolescent Boy.” Brain 117.4 (1994): 683–703. Print.</p>
<p><a name="wolfe1983solar"></a>
Wolfe, Gene. “A Solar Labyrinth.” The Magazine of Fantasy <span class="amp">&</span> Science Fiction April (1983). Print.</p>
<p><a name="wolfe1995peace"></a>
–––. Peace. Tom Doherty Associates, 1995. Print.</p>
<p><a name="wolfe1984plan"></a>
–––. Plan(e)t Engineering. Nesfa Press, 1984. Print. Boskone Bks.</p>
<p><a name="wolfe2007return"></a>
–––. Return to the Whorl: The Final Volume of “The Book of the Short Sun”. Tom Doherty Associates, 2007. Print. Book of the Short Sun.</p>
<p><a name="wolfe1983citadel"></a>
–––. The Citadel of the Autarch. Sword <span class="amp">&</span> Citadel: The Second Half of The Book of the New Sun. New Sun. Tom Doherty Associates, 1994. Print. New Sun.</p>
<p><a name="wolfe1981claw"></a>
–––. The Claw of the Conciliator. Shadow <span class="amp">&</span> Claw: The First Half of The Book of the New Sun. An Orb Book. Tom Doherty Associates, 1994. Print. An Orb Book.</p>
<p><a name="wolfe1980shadow"></a>
–––. The Shadow of the Torturer. Shadow <span class="amp">&</span> Claw: The First Half of The Book of the New Sun. An Orb Book. Tom Doherty Associates, 1994. Print. An Orb Book.</p>
<p><a name="wolfe1982sword"></a>
–––. The Sword of the Lictor. Sword <span class="amp">&</span> Citadel: The Second Half of The Book of the New Sun. New Sun. Tom Doherty Associates, 1994. Print. New Sun.</p>
<p><a name="wolfe1987urth"></a>
–––. The Urth of the New Sun. Tom Doherty Associates, 1987. Print.</p>
<p><a name="wolfe2001there"></a>
–––. There Are Doors. Tom Doherty Associates, 2001. Print.</p>
<p><a name="wood2008fiction"></a>
Wood, James. How Fiction Works. Farrar, Straus and Giroux, 2008. Print.</p>
<p><a name="wright2003attending"></a>
Wright, Peter. Attending Daedalus: Gene Wolfe, Artifice and the Reader. Liverpool University Press, 2003. Print. Liverpool University Press - Liverpool Science Fiction Texts and Studies.</p>Introduction to Effective Altruism2014-06-22T18:30:00+00:002020-09-07T11:35:00+00:00Dawn Dreschertag:impartial-priorities.org,2014-06-22:/introduction-effective-altruism.html<p>Effective altruism employs rational, evidence-based methods to optimize how effectively we spend our various limited resources on improving the world. (Here’s a recent <a href="https://forum.effectivealtruism.org/posts/oko3LekWn95Dh6ANW/introduction-to-effective-altruism-2">“Introduction to Effective Altruism”</a> that will hopefully continue to be updated in light of new insights.)</p><div id="pec-encrypted-content" style="display:none">MdxG+HuNNXrRpscK5UYoxg==;dmskgQmBOegjcrWLk/27qYUS6gCBI02RgtznG/YPJQNFB/91u/87wOGIgCIUG+C+sPuo8rK8uAeYuSguMthMxujDrXga8a4FTOpicyaksAMj0D3DfNCyMcQrUdJurWE7XvtwpgAR3eF8r57cggd0qDkyBNo3O9KGlh4uUVTcF8YUF/EMHJOnp5QUFLiwDoX6hCFU8uh2qo2+6Nl3wzztWS99YYzsS6YLuBXknHezs1hbQfjm2+93UzKAtGrkztJ1BSbINMMAMnYaYBF+05MZy1s6SIRW6QWTD3VEuEAI3buU2QYzJ8L/i67PBXl/4jDM50sZZ9qlFYNl7SlsB1Xo2vgCt1e+9ka/HtMgNKaHSLpf5ug3hyKq0DA0LfCx/DOfrxahHA1Vx/eAZ/4ole8SaOVTV7C2/Gb1mtV2TAZ8LXuWDBbj3fTaOZFtj6K4dMqbPnUg6oOMaeHz2EraxvnXI3oo7ype1fK7d29HvkzHQg6tLdl0czZLZHTEadzGHMySiNkZhvthVd6gj7RCHfhuAy9vT08vgZFCUZVvXQ0Ofh4hgInePHrXdOszGDc+pqDG1Jv9IcjDOWQ/driFBEMHA0dNeBvOCE9SJxIrRCAp/xHxj3/zvYU0yTstDD7mVhhjz26/Zd1Xdg7lflpsyu7UxNmU2COEDZeQG1CX7M8Dzm18GeJ42h5FA+1Ms2SKKAFAbzcK5WojYpFDWKB9suW/JIi/BwFDKiYopskHaPWgUcj1T9hHGK9mKsgmWpe8T7UrOpnYSCsm2dQFmfWuuuonD7KrvWrU1kpwQAVGsr6yexOl2nUhLTT9GohbHtFgD3qMTbsJjf64XhHky90iezsN3oeK8xr3ndsgiPbTghIs+ZacqOgfAA1aetyOXX2aExqx40xtEEPAQXub6pZP9LdtuR0pddrcLItqARgq6VKWhjQrFL+3EY1H8kY0fGFTZkStiBock9EaGIWAJvgFgt5gyeXdZa6ZvovqEumhOxgT7FWVXJkrcbkxoIvi7iMaJQTiOXUXU0isC37QW8gnuSnvn1tZSFgUlgY6CsDhU8/5xbFO6r1plQUqwGGtFKpb98LMNvkGxGbjGR9tbWvNr0AWScg4+rYLxoQxMHOQAQQGpM2AAQQ5BUbV/iOBLQlQy1ORfZoUJURhHsYZN8mA3nQDD4jgl18OM/jSGcctpzLpew4i/3ly60mRbZJ8V1sSKlZ5xdob5YvzJnVvh/vdHh8YJ+BsuZ1SnxLm+7maisbtX+6s5ecNdky2+ffkRBhY2xSBrs7g/LEsdISP1d3qrozLK2DEhGi3LbKfzUOKNCLzDvXeOfUGicZpat1u1Uv016aGUjAGE207tL38VQ99DToJFadRukxeFKH0SaZfFpaD1IV0U0RYjD9UlaywOhqa7UmkXHMP4s+VUBIn3GOng9td5Y3EN7movM1O4UMTuWAsW6EHsiAPaH+T4yXhH8hBHgjmJNt4wPXKQxPQf2iv9TLAWbULwd7jioDkZ+BRqg7YkF0XYYYBcSw3pA2vRS4nOPAaowlHqYQagP9q8NuSYpMCojpFFn9bVSON/yNXmz8MRfZA2nGjAobH8pX/iPqYrTkLMxFt6ZuUtMJaBeHhsJWuYxvSBlRrGAx5sMJoxgSwMFI9zfC3K6VyoYsjVFh+gMCexaGWtzut0QVpAC9ERSPZtPgRDsco5MvNGz3RFCrVyTOCf2jiUetXWBTzshvnRwShCoI2fHoGpus4VMbH8BqZAIQNd1bCf6F6EZOzIIFAE7CPtmxI8LAA+mPLtWqo1vm/RAoh4LAlkK+S2Lf/km3YybWoMmbcqCD64X5M5kyXfPYBxbFnjK5iKimVwZHh5L1lZnRwwgSVMGjOnlEt9jQfajXhzzPPS/Rupt+9mtpNLugsYVy5ql00XesP2nnkWKJFN8nYiSIvQKCit6mE4747nyjMVdWgsuUxNup85BE/fndITGBK7JReFKaPQbLeSp0cR0mYpd2V2Amhx1v7hK2/ZmNFuc26mn/MGR6n3nTfzE41pw63pImkjSMi5Rp4tMvSMOvqTsYbgTgiaGLFWOiPG1PEO4UMM0qaQw6ndCeAImEIsSsjdoqsrwZh/bjq8Olu20V2CUKeOr7+ZwDmDT3dR/5Q599PgrPnc1VxhUawjbu7w9iPqgKyZKIcGbDYSp5EKDRBu7xf61E66mAtatLVcr3V7R6VF88vEk/s04wy5BFTVBaM5s6hH09umyVNCDoaFypFBLGL1UGGUyHdWi+pMxZ/Wyb5MLxWrqAMFfCOCvET8FWe+bsb0+NtnUNYW1URiXNlJqMpl+cGen/5aInkPWoj3pp2qo+ZF7GQD7Ojyc/ySgtO4K1wTlyE/T1RVv6nUuCx5rqJEDdTIPKwOYEY2gvaaM3sM4h4EZp7qJ0L1EUHCL7PjGTOrYnh0Yj96NA4o4DCCC3/MqXLZjKdZqU94sFdzk3kyYUn/gXrpGnsAu5nsWY5+heTlBmPE1/MGGptdk09ZDGJUitv0781Y8xk52ufQKDQsscNVjWqajjlOLiJ00kDaxFwRELgGrNsRrVuDXfvWvvhzZlEW3Vn0ZF6GtuPtwGVSSKvMIfwT0vkT7uwWTXPo2mr+md1NqPPbHEACO2t4TbafArmvgSuTmOa+0pU3yNsNAJc09JNCXZsrcrQg1e7pfQbKzVL3Bkcvl1ZBydmkGkcULrBVrB8H5S+WEd++g2kFJMsF4c+n1OVa01w+z5seftYz9NQdXJclaznX1VEiNGDcEHR9u9YIEFeQDO3/e5aqMhP8m/eCu2u5ezp6mpxqdRa1/skVWCHEka/pc6y3Tjw01RCIMkKnCwSMxeXkUtvXGm5xmGWYPMTz/TPoSvpgQdSv5NExMIkcATZ3xgF/iqMMKIZfqAJ0FA+GeBrqJCdsro89sSKdk5FSxThIh6eoe180PLeavAlrCVRKPtAKo628qJePoXQa1xC2o/u90nYwSOwGCnbgmtygp51Slp+Xk0fo2msm8+9tskO0SjtZpWtlwBK0U48jk1lMCPm/T8AmGX6tG2Xnsyl/kGkHmEwI1aixrAjbDw6kzFmoKw2j8SQmWdceSIed8VsHDe3hmd6Zq9z5MDQJFIYe94sxKJec3xK6ISAkojsasL0RKy0X1c1wq24tdXoH4BLTb572Cv48UCAcH5auKeXDR26neFfNPWg9cIKYFulB+arE5mLzhfwCdymupD1RvNRRgQly/CE206fXdc7XJgu01yOVKH1/xELV0sVNB9FckE28u2mB1veh8vzBzJE9WLXGoKNkMLH29eguvjatqfOsgDlpvbkvsL7zHSrv2AkkKk/V+A60qItHuXj0t6lrBPn05x78CEBQyS16EJROepW6kqAVOdR3Vav5qj1yhHg0iaoyhGM+vsOgZgnY6wLDds3pi80ikk6uRs9U/IhDZfzboppVCYh6d03kIBzn2FhkyKUlQqkrlZ9KqxFFvaDcpEUoTvUJaQ1+zESlvzVIRS8O9Q4YLTIohnWjyhERvS8JDmwN0lfqLlCUIqM5wMzXcQbQ9gG6KDfDVqS9TNeZ2kUvJxZF2cKzXBJsIbCMbY5GzV6/HQ96E+i//tXuiYAmZaaPusQgmBlaZWOn5K94kW4vS9qGOVw7YtsoPGTzGS7Q7JIKPqMxT6lr11vwC4kbuMaTD8TDJvdxoWQ3PI3UunCsfa8hruv3iySXjOq7TO8CW/KjyDxmf4RZCyqH/MFvOPrPltXrflYfEhoB99futbbOGRyllwFNPSAKEnafZO3jQYEiajqbK+QX9aPEbujG62pqZRwWNR3qCx77NB5/00512MBFpnMIRgSwfx2kFUxgKGgBM5+HNC3XtPN9PdVO6Tyunck7exRdZtCmLkbpg4kJdcVKjafkuvvfhUVcQI6urMy45z7C9eLSHEdUydpbfFTPCLwM7wNrHJmBQFBn51zN1o+VbCGQk+2Q6ii50UOJ2Hnag5d+b9VtaGhHdgzkAN/neA4f2bud3krQ8BhBiXYqdxAVmkjJiej8arE2pOY3wwjkrk3UUUf4jhNjLmfG1jgGnMHL96+jDqv6r4P9YEjZF1hf3mu6zQ5rpc4GdMefB6gniZLKvmOK2B6+edKzyVtHa3AN1nU7mSD1MnrzZf0auH3ic10jez7OHwTOraBBPTQRMnVMezRHHVGG7aE8izFfCWrBwGSytb1B0xxZF9KjC+9nfW0VZHjWchVfmF6hk1H10souJ9Yc1LN+rcMfposj1F0iMT4bzRNgMa/c47ZGot1qATjtbyqZ9ymjuFSBS2GfEYjWD6dpjPIorsb3g68h4TtnsXsb8SyYfRlV6IXcfeTPJnVtG07hN1uliZ5eYzJIBSeFmFT0j+L7wI3ZAuJD0LmEHwUGRDoY8XjEglfQ8ubA5q4kj3ryym4/inSNqjJmN7LQk7iJc+4WHfBR8wF3Nj7FbDxWQhYLKgJquMAhdHtIR7fGaYbN1983XYzk0JoowvW7mMBg6R44Oi2Gj/i4YOnRk1QLWYvi8tJ85BW/gN4hdcXMHjg8NspuvQJKwCnk3yspFg9Pz6DQJgv1A7oYucsDDqXnfgXaDnhDM213qxIXoQLSlHZeyytCzUUiFD4EYJpGfZaItvXFH2VCiezqE7Tc+4NSj/zS33wxb/UVkRvR3l02GejrlW0qF6+wZUJv4laLn9xTTN851GoO1vU5YJjMLrTYapUXy3tZPUTJJ+Gg7ASqB/a/gAa3WLucDgp8hRwxMFxzNEUyeW74M4KEoUd6nb6c+SnX7aW02C/KYizQTQvC/u4eNAcutQEQ5yCb9Y5p9zMagj7gzMdeFXvBH2NxrryqBUsLFuyv8ExiwN0r05ULnvohgFf+DEIrL/wad+OKb0EduMENqby6D3ifh7V1tlK1UD3qVVZRSr6EP2a0Y1XojzZwd/YJoyFnGM5stvYVtth0CPd53b4Fi38qXMBlrOjOiHn+eBxFmctCIG/VOfBq5kE5k1Ie4wHu5N0dOyVyJQY3X7TpdsHPh8UySJRSAjh4n1V++8bSWQdt9vR4YRWkUorhaUZQtoslpB2t/5jtR4DjjUqAsrkzapt2I+2nnsqaQpxwAgh0CXkhbAi59k2++q3S21nQXleWp7OYeYzzAyp4Nfz++S7ji4esiCVudsjneXIHuSDIRC5HKxdnF4cL02wNo/N78Wf1vggaOuWsyiUhi571TuouRzyqprfIE7mHq4K06mHDobe30v2LHAYe+v4/nLREfhzfsYknAKoGwIv3mB+zQmXUuqobBi/4qoLaFKfE1Nox4fGds/3kO3famoeK+bKKy82SN1lGHdcer0NAOCCXBolMjqV95PINeLOizTqiiu5lbNzlIpgVETcL4yWVRAj9LU4MDskOK7PxhLyD/lKstOZ8MdzJ+LyGXThzrWIEtT6OPVbrqA+FRxWrctgoOIpRUa41P7niX12AaLTwp5xkITuYuhtbKpGlZw0UvyRkyxzpTYofroN9Mi5P0kluS53Xf/8LAYjVlr3eArLi+WNwgUIEblgPyEksJHqZhWQc8H0H4KDahRWwYM2q1qGNTOQbooaF9qKGcLwWI9ozPvFzv+fv8LlFi/gSJ1BA85P/B+ujtcyzgJwZahrJsA/cgiO1qJcWCSovXP8/oCaBVXuQd85R/wqehUK9cRKyQVMg9VGhOKeyfIhsEh08EQRtNPDutjY51U53Qg04IpeIElgyF0z0/VA4MO67FyEEe2jdV9PJ+YaH3U0bfjUhICDEmtJtZYyWKzt7E1VtkcCZqotq3iBnNUvTTrJTP1ydkDXuvF4eKBJHLQNpAJTkCLF5s2BrFVAKVeowvOqyJfaR2l7iWbn8uV8slguNQz2fPZYkXKq4sFjq76DO4N35dD6ZBfvBw2JPtMMrbTxW2BIX6Ae5U3wgY7wsBUdUeLPJP+0Qdf7NqCTqpA2YT0OTOPT3l9TnWbzqHBbrwbGm3+2kQg+361Y2hgwszft7xLLcntCLiM1dCaff0k1cT/tpYY/mpt8oeNNuPkzPXa/19/D30TAx7UDDSthps+U/V/zYi/qoAPidIWCW5A09hlToFwPZ2FlXile1uzfLNVFg1UT3Z0/3XK87KHwmQcOgXDaEK+DNVI3QvEX2Fzc7srpUMyOBeQX/5mub1qsrGnmWu7RX8H69BV99cQ9fK6fCBni6vHcNMOrtkxbFhCXdqykaPxc/OXewWqWyJINpLVMbw3PY0ozpdfwta2RBikOw3xNhlFaVr7ZNulHG4WD1qs68gWh89Wdr9ZUTVPN72QApOkoEC/OJe1q/xp04UWWfIvKwjsoI/U1hOYySutghcDdJcKjnTxbW76gmVw0m1186KaVfpIIX/yp6LnDmfD/dvvutvZnLhG634EPvNRiK2/BMoXv+vtPcw3ykQB2r/jjFfzI45AL16LmSfd2MALf6jNS1041PBCcECT7tN5RcGfyJFreXNPgeLYBmkWBJ8GDZ7Yr3eG3Ca/uo15Jlp0QqwerOmbnmmH0LThzPOw5kYmXMl8TC4qp6a3UtWF41CbUr/2sGrlq2iDbquiip9IBrX6cj3Bw96w7C1oT13ZsjJkhckxRrSHx4qQkF2B0OvrpveGHq9EAE44ZIGFSup2ztsuGg213Nim4ILjbrbz02NA4w0FRDgjWbCgfF9fwMWQCF4x6WmLeVv9BasXOj/TAPqQuIq+jaNU+rkXlgLc0g2YTErnQTJk1NgAAJbseUBU8uaDw5Zw2OMRMmh73CJNeyOntBP7qYpG9yAWvFREHqq3ECOi47aIlrkRBYGBXjJ99HVidRLT/dRfD4uGiT/lHDnOprD8Gz/HALMsh6Ue6BWv03jtSmY1Vy+23qs8CLhaoTJ7WDeVdKLZADoGkL7KvM7fmRRAbnz+LI8MAP70k7uhULV55iS/cb/9XKGVFNNuZZcILxau/2OXQ7oXdcUW5XQ/xzlo2QPzZFQJlZIdgstcnU54N7xynKa0vV4N4RP8/2nypXQtFNXzdYOc0YOxfqsqWXx0adQa8wm61oYefW0YCUCHWOrpuZRQ4HXsPB+/z62iLjXyelNWCgqdL5DOhxXkNHRgtFEhl50UJQFRb4m/QRk3qPgi654b6LqZbLOJ5ozszlHhuxQt4CfnFqZnNdT5JgifdQUQzYVX39ju6N60F8w7Ed7MhjNXaQ4JLgg3FUy9Kf80GJk5tNO5EqD59xGx60CWa2vVUQsh81p1n7TO0jQ72E2FgqhiFNbUKwDCDH14HaLFaNX/vTx/W81meg/PyZyz8R5bfpdBmuXq/7VqYJNDB6ucIlAMoIewKpcmpH/+VkZnc1CWy0bI8ZDPoCFehTV/Kf1ZyDGB62tqtx3UAj2pbK9dj3DEsOfCWVqA/4aMopmAUz0eoqyzH+d4FtaWmYnaKFEXU+LFWMaESAD5Z1Lo6GDKyrT85flkG4UTTdDd7igColentVmmV+KhTN7p2F1h0WLYJhQSVmV0kkHYUggfnBYao5XGerqe0HuM//OciOmh6c5hG+HxCbmA7jaISyX1nci3xI3Fx+HBAKCbLKkwzUHcQVkoDTsggYEQDvoGWetifMoI6IBQy0vkaLVQloJQud665UMUJbt6//iD7pL/6lKPQUtEpe0ZGFe3WX8/e/3mKFSEDn4WjRQNG5hpXKOa8YBzE0cxYPM2aNswMRJZ2kIk9FHZrbUuqeuxvzjhYf8fM7OXhtIb7j/0UDm3zWrxv0f70ZdCdyWgFojgzAC3yywV37f58UF605758r8mIxnngiGRch+M4KIeSo18T9r05a+K73mA1r74y1JpH4KgRPS4DBkoN9sJXhNW5Lm36+EsWnWDM5vxeWByVE9nN4WNRmuV+dEaJCMbBH01/kwIXIAmBUYAwTHl/KdbYJVa1oAe9XDdr+qgD2dRjxTfCVJ6sZVHkJt9yi5I8m1WDz+miys6qRKJuNWAIT3a94ZoCx+1LMoDESYxRbMjaXVvnY+ePuYraUxftYtKvhyFsyPZxbqH0Ilt5vDk2GjcmwbuzRCmq5gJgH1HcJ24Szv3bx/32G621UgdEh3iuG1z9WGCkNOnoGawNAj1hgpkv9l62xR7yVcWOHitMQcryHRtf4buLUo5pQsuZUEAoAODY3UZAiPLC556vFCc+vt+Jj8ZzLNVblup0TIKNH7A4KTCL09urcOmr0C0YFh1PQdTVWbJS/GQROUI0le7uw0etlHegcYwn4wStfIwnV+xm8DA/cnq95e5ZNRNCOGukFcQKa7YcOyBt0G1/MP+dEvuSI3k41gHXnQOtTRGyHGrjp7yaWnBt0WdgpDZRRqCHGOLThWox+TuBGH4/YTJ+dTMfRzskRTzees5u4qmaXYlCnbLNB/lT3eISYEM9liNvHMn45OfC1txCfBfgZD9I+Uawr61JR2SEmuB8Y0a22i6HhIxCNAF/wdwQP/wDXgiU1xcwnNpIfZvsyPqE+tVffKH4JvS5l1EdueoTzPLi4N29OmzATBeW8MmQAn+hcyIR7zID8lAhaMMLWLedjQLHqp4ExkvmBrA6KRVlSFNMGYSMA2QPkhzw0hLdwEd7TnlYr1k6iq3TJtfCJufHyf4gCbF9YrbRAERbwLmCF6BMR0T7fpIKXNmT62FrKbuR2JnIoF7IiUX/eSKn3FKrpdDaB22XyI8t8fdrxwUZmcy/d2xiZe0Nh7fJLKMd7Gtcvu6WAMXtdWZkEv6rf3Ne/54CaCiZ8IzullQ3fXRVa3u8n/kOtZI10qGizxcvceuaN4JSUDnzwoztLJIA3xBjVQeX55FSiL1ic/S/eR5i2/5DmysDlnlIXt3qlYbYWzqJtd7vbkBY81C5aUGo0cRiYm17A7xDgFhfF+0yYm2Ss3tGdpC/qncB9oW3vQdWBflRe3lMxOu38mZhdt5gR3orlr4m42vcSouUJ7pWgEamp2a/68BvA0wz/3b+sgYdRExARpKIk21kBm7ujTfe6wRw/usr24Typ5zLi4AMCTkXU/lr6MNkeuUfHVlNohEckL9AfxZ2f2sN9+oDqBFUelm87s1V8UtzAVXvMeJgpGWcM7jyT4RgUPwX38z505vpoeLvGi3SpDSJiWpsbSvmesC2zVjmMjmzjZuU2PAfKgDsWNMaAND2DtJ3vdQUuTEW2jkCzutxRS66BayjWbzL4HTHVTU8RmCsh0DU3rMIyIF0wxY69lnng7TtvVvWx7Hg6jBL0Gq4bGgLAVMqPD/82G5zWf+TBcWEsxNSaeccUam9pN0nbxTHEA0WvPb9cf7Zbg6kh8qPih2k87j7dLsWkKXLw8fRJl2YwIRrQ4ots+nOCPwM3exfXlLq+O6l5ilREngx1ue1y8PCgyupznvNAwNy341r7H9rXdqRRWb+YSkCTf7Ymfzn9SedV/jtPkmfklbNTNDELJ1yj6lQN2dTAdcAHMTe1RoiBWTlB7wLpaLdeAXYVu5C2ksaauaBojhhJM7ftGbbj1wWMat2sdfj8FSC2hRGfBWOZbPB38bJg2jljKNF+B48smTb/a3uFaqrRuhw9d3RVsVN3GUXcpqqDUDYtCgpHaEfEp0cqZMLo8Pro4pv0/77KlSvLrmplNn0HcVB1LHuenA9uUvr/FcROmeMh+qrhjE/u9WiX1MhVIejoL5m54aCuUzsjaoHenEFdPP9u97Aq+b0j+ryPh8xCc/yqWOaZqXbqJ2I2TcIzoWv3pg6OBQiOXrARIjBuQDmfJxSY2hIsokazivFEseBYre3P9JzkvBzwI9cnHNYl+rYXZhRJoaVSdWqTSG2T1NutYzMaxZdGzv6kQ+n2fhWQeoWnHzQCWLyUiRqtQFGr+ypZMRQaBjDIrKKpZFmajfvaKFvaGWB/YO06tZp8rvqDB+wuz1FeUBNmQZ58K27t/7Tu938kTG+m7V1vmjkmh7bTxxzsc/YY2E+AQgZqpkptIrToEAmy8Zc/LjcAbP9EM8E13XL3It65MXg1VFizwEGdMXbJj5VKMlZhuqvyKOnPbF8l4ja76kg0O2ylKyJvrWzfcm8icLdAesl5Zfkedc20xdocT+/4xyIciaUQ+v7ZOJjQNQ5sswkCXN0xeoahrmpe82L+fklzuDXbGPLdGGIxsmNsHLCVZebQ70W4inKMTU/s2RpttzXV1Mn0oZUgtVFC7VTbAPFKgWKmJl1Q5x6jO0XwkszRPShjQT0UypZbw9/zZXkUJVYWs+3WiRRwWXY045BM1ObbBTJ7jhu6mU+QX8vupUY6J152PjzPw58TbLP5FFyTpSU+qz3MYbDGWRnBmYDPYwkyZ0NiCI7ETGr4xKDt2MBrbuGs1S64rtGXWFUSvIkVsfT3fGTX0Xs9+9L9pZehp11XDIvXXGlaW/Z84l9HYqqy03gHr2LEWF3djj3xbMgfx8UnvOJHUg/ZY/MNng9mnDItXZE9dWinKwgOrzFNHRx4MF133pqqhXmKZ6EQMkHR3UYQgQuqFh5SAhRmg3Bv4YcCClP3lA6IoCFbI1JHsMbCWEG3tsa5oY4qTgAy7d1Vs7D++WTRB6ms556x5Damub5QRUZfiH89hhHhnue4x28zKTv6c5+31/UkomIYH74WbR5sgV3lO2AjvzGHCZc96Wdqz/h5kje1PH6vavf29+LK/qnPLZQJTfzmxoU2ASGoumG9gTPrHyIjrtyv3uGGectBOAcDtrqcwiuy/VDZk3+4P3sGXVZmVUfKI2lmwyz2lj3mKSRLF7bPuDaQ+M01P9wyZz02sBrKXg01z1ApE/+/gTCAr8S96OnbbwPnzWJZPq4auTFxPmRRYpTGHU3K2mUk9QftxFeKVvdS8fHO5K33WsSOFhM70WTAlZlDjk7bJaHtqleXR4syjyrD9RvdYLsXvwnX5KszLqx/uHgYBZNtBxaxrVuQGd0jIEdcyux0Tn5cJbxKnHL4xTZtJnXJrWfDQ3EmihuYb+bOTlvvEsJcThAvyNRah0AfI3bWPu5nwi8YIgD9AQO+IqqIUEHwN1DGj+dO+fnzCCOevWCsrJbvgHG8ba8Rl5m/Jd+wxzBzEjTFXso/e5hb1fgJM4ZZmbQtTP3NdUrsdTKdZDlccSEpu9Tgzj4gY9ojpifwEyyzaDnvgbWfyR8MCy3m9EuR0B6JqTg3K1kWNIIxBIBmAEw7M1nysuyaVRLP5IsReks2PPNZ3xGGMFR1VnCU1RcrOmVfT3mgkfckLdWjWZENllqJEvSeG4YI8JuRxW3AAxJrUkEr//BF8/O/fRVFd85Ffeo+4+0k21hBTqAL98Q5D83Dy3AjaOgYuk3XfJn2Emok7FUWzKSU0T3oBpWNAGFhQqqeSPTkW/Z3FJfjPElCQxlPITuB2mhqbzf41MM9XqewfdZCmF3H6yBsfXeSc9af8gw7ysR27CIfOe365WLbwgecviHUOImPPpnvvRNFezdm9mtlQzytE06Pt/112pTBAo7w9gr7oxtL+hIzKDQluo8knu2d4zyhGDN9dhKl4uOjVk705kcdN8MrZVm6bf5ILDUBKuEInhU13Is4oyIxrVX0k+7GeMMzEqzSHgLgHGxSVn5SsOLr6JoSKOdakLWhSBZFKYRUjiBL0y5QQcrwUWPGbS55iGwOO+FMyhoAFTxrTN/4CUB5oIexNXoJdIGqBfJx5OC81jQqYj+raVUanbIYJpYWeJUWMOzZWvc/YSh1+3tQ+WMe6U6QxLQZHgRMoEk3IN1IfIO5jU22F6SeZrCFKXUzCB7AnMTyWY3fpZobHoREKZuPsrPGP8pn7ISWwKYbNVXJb9DeiIXV1ex60KSpWjeAYYnO5aBRgrtBKFW9+S2jyIrVXrXJ22m+1KYhakdIDmpTCBYIOrY5L5MXwr60PC9w+A/sGYtiPnDcPtBwzsWYanxseeYORiPNvXZ1ClWJj/Fk4KGg9uTc4fOqVP3KlguNpb+CN3H0WQLkgtp6GequtH2xVxfAyPw22AlyQY4hVrVzLqVIQF6LTdUfuLgzFaBAOpFz/gAeuoqirahioF+GWO3wMKamK+xTc3ZuYYKqsTaMvIXoWjvYjuzCA4wfazsYR9DdQ6vIdet13vVIzxngnHUELYNljcYe80zxntxqkLxCNIFp9QJonounQsXJk8KI35S26Dlbxz8w8TOYDHqiO6J9WAS4V2qRgm+kyAT3HIohGxBQfFXUZ47YT1LtxchDdrOXT42ICBiThN9gKmq3naT97T1bFtMU992vO8dXlMzDnHpLK1EvFV1Vf0H4nPWiXABlicaYdLzAMDHkJi7OcAjdt1cjwBQ880yOg62t4+jmYdYWJNbVcy9WUgwlmTrVTQvT7YHZJ+X9AMLTESP5FcUABY9pOxLrPLdDI5t45DK8wOlTffcomjL2MzYAKs6nS1sVYExhka5lWj4U+1KYjli4h9YP9BkskYp4iIP3nP8+3jXqKvcvFrW7fbX/N1+DrmRLRmPdBWm4vIlyqS2fvdxhzS8BNIJ+xSOcRXrmG1WNc7X+kl3C3BhdpcVNK8xU36k34iblFYX331/cgRltiDaciEuc1GT9RXSnebfaO9TPiCtlxmGzwhgkMeriDQ/Gx/v3J693KAFmE1+sAhnNbnmPFQYWaLBHk0QbjIdDxH+dnsvij5GgiYqwnL/VqbE39zfuVHe7cGS6AiDBJyfJMMnzF2I4qf+yRkTXwVpHkpURToxS//d75QSfh48Qj4tyoYB0ICwwuqs27hVAUHsUJB9N+hAbi43mqdl2Np6OsZzA0cL52U/oX1otwpBDpzg7mMQxI8P6/7voHb85jZGJYcw/7cwqeCrpYfC+XiA87cgsO7Y+9wOfFbZhz0cSt+B3/fB3/+E5y5Ia1PTDi/YR9T0SmpRPJxc6ieYHsrKV6aflHNIxzGhL0j/OL4nSKRJSNYuSe+D1IiVapMDKjlRYZTEpSRMRAX3OxuqkIhz86BnLm23rd3sjvYLBo4ktju7c93FWi6CB0L85Yt4Qdb7URxSflr80Be1gef7F84UVTL8zYTewyo6cql6lnndVV28nCywlyhUnuv1suHhMPgPxFzgASgs2NRr/gQ4TzvbJ7EtxmsZhP2113lil+U2UcOqMF/QDOLkt2xQ2N/GQTRGyhz14M9UsjjNG9XA1X+xYC9fftHLFXrxhRn3wtwUP+Yvy5Fp1bnPgMrr9DDWGl23UOe8pTfJ7hpyai46cULXSxqtJ8GcdjuHgf83KHUYBaYWK//rE8kY8MNxRCsWjENAM8qAFUgnRWi7T0I5WJbHTmJGbeANuvAdVZ8LcIRv1OX3BQpHrNUwlpo/cCZXOpcnZr/HriRMZgecNx0DnbMaJU8t0Z4tDgxPZcjzk1S8ZkPJp+VgYn+yCaGJu2HuHLG4sZjrAFPyc0Ydcjs2YLz4/eITOLozu8EE3zzWbRKsxNXfNTXoVgilFGLOrRUrztiGgmkQBlADRpyHydaGuvRi8mLhqT9HLl89GXbCJTc+r5sC19rFnbSeUVhw12PAPDYg/mMIhrZ11MJv99v1KBFLmEuUWT0bK9289o+4isDlyhpj5LYSlC88e2IfMhHGBgQiy8go98pefgnuysFLrmdZU2tvrFCYywY8V6rpE4rU4dRC+LjeNJYVL3Q5JUrn5CTi270Dpv6DFbGrSftuZ7bDM5nXG4bEffuWhH2cB3EbU2r3M0izwKnzwj92lXj608b3v7Jvq/cNKLLfdTay390RcuRfzORvOoqulfgppKedldlM8Qu+n8RzqgeIReK6DZmvOI/mA56r+IIE+6qmE7UcvBnR4QXnrF8srIdKOko2CFaMAERiLXdVsCsmn6HjBdAmgeKRuKLF0FrHiyFKCQTS8xoFujqJjDxI4AGm5mC2s5F78SWJz//F65EyFOO0MvUX3L6iNb6gOLLIbZOP+WkyCqspeCelsHXzzeeLkiVVojctM4qKMQvToDoPeXnJLXkTF8hlO1oPZfV9xXU+cxtjnUYClyTelbRXxyf0ch3saCVADxThFCs+Q45AeheU0eC1/XYVT8iWeDHdetF0Gp/3R0urU3Gj5ySWcSy4Cq0DBUrISCzVpYiMS4XsrmLEG2jI43OanpPxBVl4g4IrAL+Gb23jV8lRZoXHpflHIoE3Md9LlBXeDnEhM7Q2IfTSmqwO+u/GHDOmUmLf7OL2Hx6nh4YMjaStDqONupvIHox5h8154+scOaXV/gaXO1Oqq40a8riNRnZQ4NF2zUXgj9VTYdC4xTg7U8BIhRZkMBnB66zl33ypVbnTqdaqJ4RkLKo0zrcVPEF/Uh9EsAHhBivK504bYRcFapYOUcM2Iu80C/3TWxb2UH0yDpxSgjg2wE0vI4Zqg+a3prvT62sdinuInzVtalbskYApHQIMzSrYkcoCQ1nqBpkpr18Tlk/Zl2+0I4CLghAJwfO1cqN2UOIYBsaWY6BOENs+dSlcIYQCmB+4uuJnK7MgOiSJ04eRwpSn+qOWJPKd4uit/656A6RjuyZce5eDBSHrnQITAWSxoGt1CYYx+AmRepDPsA+V0jVJQLyWhBT9O4aAVodJP3i7XIyW9mTcx4WJuKzA7LtLzg5i07OS3ZLQLwmgD3lRYRg853J2vMxRQE6i/7Ku2DvNBSXyai70v8j2TryV/BdYmFXoX1CdE9yU/dmjzBT9ErcSerpJBKy72MCb1teA53GgNSEJ7FbNWsh3m/hrQqqEv0GuvqFj3b+/d1ZbjZ0MABBEvhgIBLP4XiNKExKltJuUI9WOjiXNJZC8sD8mItDLIx8CcRmh4UfsqyCii0F2qDAosH8a3ic8TJKFf/tBIP0RcLFlBRlso30ntmoinBKWyk7d9onZRLLKPjwOnd3FIcPb5STiVIhpcy03bYf2JkDaYrnyPnM1EmuhU2NK/zyVRUV8kM5Z/mEBv7OcyJbP9QphylZnJmVuyFe7KYhQm2xdiWwFVnH47Z/SrJqX0Cgj6rPjD/yKWgkaowaLfHcV2UmZts07RFbyDsu8rwbwbu4plqWDohAOJWlJRmI0c0kltPO1nvuTw0JxDOC7s2k0pOLpp8pR6whoSWip8i3LCvBubMIlDiJmnQon/TUQYwlKnX/CQqCRa5iLYy9SDK9WAVFS0sj4mNVB0mzVwmpAYZKFdVXdvSr6wEhoVV66Vq7NbW9KVUUsOJ4yVXqs6kA8jfbW0WAEf+Anl1x+1jL7K/dIqHq++EXHDw/Xo4RiyPVmsadcptjMm12RKs3MmGSguX5URcWhmuPrW40yYifKa59G2LxZ33SUFxbrNlabayYNmUKqYgC3R0yfGYBJJBFjSNfJekz+fFGHnAm42eoy8FHke1TD3BgZQraDsnq62qTVpiDDUx8ZjCmsaaQygH6DYS/HvyfEk3iKdCIRZI3+ribdsoHCAUwgP/LXCuIQDE9aP4CtRVV08Wewk9Y7uoj+fIAhiI+M32fy3JlAGI4xPBGcr14hb/fNRi3gTZYwTL8ZAbmkm0TOcRGjAI+dJN8hdK+mZTdXEo4uoWJvgRABvMJVDJc57nWQY1cgryD9jYLnaAS5UfD3191yQQ83CdR4PwGs//4hAbZNEr02N0y0b1vafry36P8KAisUrNqauZNxrvcYBjK58XJbQiM2hjanm9P3VBNAeGzMFRhx4PB0l71uTaFk9fXw3wjZPsP6P/eBirHXxTv5PstXJPixrqwj8UuHmlWDhmd+YbY8hi3CL+LRSOfTxzoJHOjrdonC89OBw0nkFosSXSwoL+wesYOOGuUZT2KLQO3nBLLlhN9vVg/gyjPKzRLCuB+5JLC3brLKbVuHYKSeRiP8JHKPpEOLs8fsT96ZQ/A86dEmpD861xXMZQNA+yJEefZLKMpUsyjbFLHe9dOHKFhG4228KndjYzVKER7BPXFiTlskmuTRAAVSZg7fzU6p3nLkp3nOwZdRAbi1re1A2IjwVA58TRJuXgJuqZjH+b775Els1rO4IKrjyeBR2Uy6F6POVDGugqSUokuC/4FFTVdlyo7ML7/7vBQAO2k3Z71mUJmYovo6l02NE3jzRfrzzTHlFJ2PY5mnOqWVvxIOZIt9SMZU5OuBW87SOg/ab65b4oYN4Ms+TfVCl405xG2arR6qN6a+r6Wvvzs6MEuz4NAZj25Z1WXW3EDm/7iZyL8p4ZgOjQ6QtPITCBsdQ/llqwY2+4YWOR0VuX0TKgv7q5y/mhbDvGo1bgim6Pe/QAysZbsz80FGSRiL1DX5o/GIEkgjsXYq+zuxgtXA03VIrfXB6vxdYNVXKtwsUmtROj+BtItIrth5kf1cmpI250vJ0unM04dA8xHf0a3SjiK9kfIMfvFKPfZeT+mYB3QF7HZP6jpKUj4iEgfJ5IRY/76yt68NoxSvPuoHxAvb8bVjNhQvNIBZ6Y3oQ5uA17NNg+OY85U23Ot3GYWWa0QnkXBRwBpBNKljv1L0pl5lPxuOReKdcYQglSWNx/h3JwWh7E2JY7UGg6WIS+6PEpXB4CVXSTT20BsQsCI75ZOeZsdHWF6ueED4h95hx5IVmUsMGPtt3v43bzzHOC1yiW58piADz5QhkC5YV38jGWo40k15LBXS4Ci2WvW+NjoRHfo5S9ngcbQra16W+VIM8hMwTbCM+jOUJg4ssNymOm7u9O+F2rmeTIvU6JGclkS0Y+sH6RerzmQBTqhy4Oi1IgOrPHUV4FuIh85xh9/+7c61QnRPXGUyiEgkvvgtO+y4QiNCYA8dx7NWu2qYqcPL+J/YutL+bynLwJYomAuVR5WL1llB3JtvWcSYREyJev4LMbpS0hiPLsBjw/7VRhkLnkbQ/C8vbt/nWBTuMLPrYGMMa63eIpDrCf78ew31lz5OvD/12ETDOzY+D+VRTObW0ugrbE1nGB5ETdI/sgAyrmK9XUdzRHbE2f4I2ez3rozzey+Rb/xqQEOnwHNfJaFvN46OXwHFz3KgUIhKn66IY8h7yHwTTjLsfxlhQ7AOuzBYuM8Gpp06RY+jIJz+hu+3ij4HhXjzo6fzNVObmkBkzJXAy0belrxku51p0rQ4mheRw3KDkmwCOW1wNPC5AHiJnrJVmqBXfYuqYwCcyhcH/M2RA5xC03dzQoavNvGPwyGcJ2KqFld3Y/wibUhY2WmIhIzDCFl9XzTvXOx2NtNTBziSZ0ke3UFamvKCjy7GZC7eJVrq9XZX9XxjxQcoN7NRWxZ4aUblIBQ30qrQyAm5ydIpHOclzyykjSuTRGGxbUzL81GqKvPLImy9izRp6nUALuF8FpnWv46ma+bpJ/Dxwk5IwiQh0ezTWgPIvPfy6Y4QJh9wmokFvh9L70k5os3nZduwfQHUPvnsCTUXPsx2AsHz21i+MfXrPjWR+JFXljOu//RqLuo77hcK6c4Qm9hYwH31SfIPHqzDkAN93ZMJTx6PTyI+ZTiZOxPY2fqQ6ntbRWbvK9sggfcxMNCHkqieVMma8v6aQ6udJyv0z6jAdjFvaD4R0tNO6j424QtOHLLECpd1oKfvYv50OcotZP7CMIa7YwIecNfvqDHrpbmsjPjLihyukKB3mBBD07sNYRQVHod7ArVn1cyFZVVsc3dooS1FHusBQu8BDBWVqE+rZkNfLVA6YneI7C5oNuFJFjCn1L4nSb6VCDLiqrA2a4DWW39yNkI1sotOrIuk462dSWQS3NOzr+e5ch8NEySGGCJIC5bmNT9s/2VKO1tsUCWFTH+TYDXul4GkO+gWeFYrgZ7ZdQ737dmLayV6w1pldhsDuyq5Ydfn/VNGbVoBesS/1ed/97pz44dHXyR+CzZT0d3du4O9MyljD5GN2u7udMoXf5vSQr4CfLJSHXp38doQFjJsjMSebEJ5ohEA4Fuv6Su4cKQ+Tk2gcDHE1hfkLb7lgFfVERcy8lX1BKFwGwCnhI/KrAjc4tNSC2c6t+Vmtq3fvJOxjAOJFn0hJkG62OaNjA+0WmR3T4cLWiRieJjBgZsNYrBE5amsSlrNflShw6diDFg+8OX9T8/a2CPrn2atSAO+EWEFheBevNvXq1y0GAumVrjXHP0kHY64/DTOQJRcGQpgR6flfrelcPbNtXI3Zp7feGmZZPyvrhQtK4PyNsdz5jXQGEqqZpNis8iQWaEHTmx0osFMPYvIaIZpukePFtHmMBZPjFcyKtB8MKZF2TG6JBMOf1Waq7Q70W9ZNy/+eIURjk06+MaWtjZnkIOrup9Fc+0xAayvuEE0aGMigE5DmMBSin5zfAzqCBY9phB4nL8MWTOlhurQLUFZTDvc/2cZorZYjlp+V8PR92PMDuDatBHQJNqyGTewTgYeuMiHRWrInTLBApZNznGyhZ9DLXu+MlgOYn0yDnPevFthCx+S/j+eZqKplZxfjX0GYMe2lM5WgDg5aMq/xcmeZ4+4hMoqM5Y6LJING+CmXbLvBnwl7h69AQekaXSe8Bfx87+VlpwAQPN9kuGXoCLHtyrtmJQVujdcKwlnJvpeo7BaIqb5qP5h6S0/omdVDtOOtG7PPOtkcEcHNxGKoZRCMujEAMoTBpWhzCnwv6lTcBtp73p0AVVNTbv98xX3N0xe2zryEp59D7EHLe6uNWiMNLfEHu44gzQFBT6SGxMCRfrA9E42Gbj6IiX9/tIUsvW3d5/4J2g3ZcguN3fIZQHHGmTJbPG1UWJQ74OrJSC7Ry7FN1PBlhCBKn4P54u0ky2x5KJt0NfoOt9oi2PDcpZKoPU2HOUyxMdzcWkJw/tdS52FEqEwhwq2l3toiLmAuw+HZ5xcnXaD8eKAAtawN+gx+prSQy6SMX9xOMTu2uzniq19Nk2S4Ne/wutKGDro9DOAYA4pJVqTzxZS8gUVUZa9huD1nyFp9ENCytkV1MDfZgyT7+Elw6J79OOXUFw3iFpZzc1ZF/JySTFvX81HuuZofkWE5UMC6rKSbiI8v1htVVEwyLORRBXQ5blrdMukLQsbMoBig1Em4ITddKxaivmGdAd38iqi37lFkb6gGhWcpX6UXB3zc+7ZTgetEoVYuTC5NY9gcm/490tccjEt1QT78Cg1Kp9OWZIx6oHQGkR5YNSiDiteTcnGDadzqQI7V9ecoGgiGwcYAVjAaxLoJ+w+5XvDMGhvUxrug/lOmeQVZ3y42y+RK6yk/yM6i15yW2kdf2btZ5WgFInohwM1fQ6b44ro+2iWMNNMvhvRSbKVLb8BUtOEJd4XlOtsjYBBpkb8RoNkbMaXj6HzAmK+dtndwOXIgwl4GQelAeHyU96ETXLVBWnx8clD2P5EBRisf7nwLwDUK5fdqjkIrk9FQCq6PkrtoeSd7XfnggrGY1tsr3SaEVdk1NepDL6YMIBtDgtkG8nGFPbjulvRvE3TRdCi3Hu75hvbGL/BFW7aGZXNAdmVzaWOsm3Zq7GhTtfYqy8aQDDop7/Z9LN32yYfCK6UmJumPJH9EJJ0Oe4zH0j7JD7u3OPKDhojwdqbFgIOhkijhIGzqm+5fKMhWKamhQEcbu0KUFYvtjvfwzPUNHKIlFqbMSDNUPlidkyU1kIKRaRmLGZy9QDGhgpN/T9r7N3VkLWCIl44OYPip0rPDDf+61bZVX/WQrZq7HSuZL7pDMh5ZsD/cJQZOuiT5SgYtpQiKBoYnVMBGqraHiqooyCqpdPh1+y84dUIvQ5pyRin2jEysvf/m4Ezv5PKjrpzXj7dqkPoIQ08SZovBoJRI41tr/j9yVZUXNjvgyZobrDYj9xWAEHlAQzoS1oYgNg3Nyoj/FLqKMGC3edTyjlZMmY/tNlxtGe95qNYJaGMsD+5xvHVjYtyl6fKun61jrTy8txo96hhv7o6pdBslYkHrRST8GdSs5xMxnJJcqe12Mczqs4IzVqAPqqqob6i/MaxN0gE9rR+jvbkqIMsArsnPvFWuVc2JWEXW/2trX06ISl5g1LKMsM0c8gDJY/wtcSiH51vMPsJGVC3RzkVSG57BDH3eUag1SU9mNA02CBZR3R0tidgE+89xa9Rxndxhx+JydNo6PIUIi3j8HcXR4tnwhCNqmTRIszuhTREW+GjN+bAHPyV/uvPJAK6s2GbFUh9CmO6kqZlgGS1VcYtk90bjifq8kFOHtqtEVoW6NIn8NtGHWqiRbmfX3fiYlk/er2e7QESNfxVdtjt0K5algvLSXVCP6DsVoU9lQg49RPTgdXUXmqh6tfepPJb0xIgwjzDB5QhsPx7ipHfVvzIeo9zFq3PwL2PGJPH7jMwNLpTudUX18m61GiaH9EPCmrRfzHagAbXb4EY/oI7zQ3goyUCE8H/FA==;^</div>
<div id="pec-decrypted-content">
<h4><i></i></h4>
</div>
<form id="pec-decrypt-form">
<p>
Old articles can be embarrassing. If you would like to access them anyway, you may
<a href="https://bit.ly/3jPN8tT" target="_blank">request access here</a>. Please indicate
who you are in case I don’t know you or don’t recognize you from your email address.
</p>
<input type="password" id="pec-content-password" placeholder="Password" />
<button type="submit" id="pec-decrypt-content">Decrypt</button>
</form>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/core.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/enc-base64.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/cipher-core.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/pad-nopadding.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/md5.js"></script>
<script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/components/aes.js"></script>
<script type="text/javascript">
(function () {
var strip_padding = function (padded_content, padding_char) {
/*
* Strips the padding character from decrypted content.
*/
for (var i = padded_content.length; i > 0; i--) {
if (padded_content[i - 1] !== padding_char) {
return padded_content.slice(0, i);
}
}
};
var decrypt_content = function (password, iv_b64, ciphertext_b64, padding_char) {
/*
* Decrypts the content from the ciphertext bundle.
*/
var key = CryptoJS.MD5(password),
iv = CryptoJS.enc.Base64.parse(iv_b64),
ciphertext = CryptoJS.enc.Base64.parse(ciphertext_b64),
bundle = {
key: key,
iv: iv,
ciphertext: ciphertext
};
var plaintext = CryptoJS.AES.decrypt(bundle, key, { iv: iv, padding: CryptoJS.pad.NoPadding });
try {
return strip_padding(plaintext.toString(CryptoJS.enc.Utf8), padding_char);
}
catch (err) {
// encoding failed; wrong password
console.log(err);
return false;
}
};
var init_decryptor = function () {
var decrypt_btn = document.getElementById('pec-decrypt-content'),
password_input = document.getElementById('pec-content-password'),
encrypted_content = document.getElementById('pec-encrypted-content'),
decrypted_content = document.getElementById('pec-decrypted-content'),
decrypt_form = document.getElementById('pec-decrypt-form');
decrypt_btn.addEventListener('click', function () {
// grab the ciphertext bundle
var parts = encrypted_content.innerHTML.split(';');
// decrypt it
var content = decrypt_content(
password_input.value,
parts[0],
parts[1],
parts[2]
);
if (content) {
// success; display the decrypted content
decrypted_content.innerHTML = content;
decrypt_form.parentNode.removeChild(decrypt_form);
encrypted_content.parentNode.removeChild(encrypted_content);
// any post processing on the decrypted content should be done here
}
else {
// ¯\_(ツ)_/¯
password_input.value = '';
}
});
};
document.addEventListener('DOMContentLoaded', init_decryptor);
})();
</script>Numbers Are Ponies Too2014-03-06T12:00:00+00:002014-03-06T12:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2014-03-06:/numbers-are-ponies-too.html<p>Amber Rose has freshly moved to Canterlot and is eager to start her journalism degree. Little does she know that she is expected by an old acquaintance who has spent years on a time travel spell to eradicate a formative experience from Amber’s past. Amber succumbs and finds herself battling the ghosts of her own adolescence—but life lets her choose again.</p><p><img alt="Numbers Are Ponies Too" title="Numbers Are Ponies Too" src="/images/numbers-are-ponies-too/numbers-are-ponies-too.png" /></p>
<h2 id="the-ascension"><a class="toclink" href="#the-ascension">The Ascension</a></h2>
<p>Amber Rose had no presentiment of her future—not this time—although her life and that of countless others was to depend powerfully on this day.</p>
<p>She walked right through the center of the gate into the Great Hall of the Library of Canterlot, the filly seemingly lost like a red and lightly yellow mote of dust against a firmament lined with thousands of books. She was not lost.</p>
<p>Stained-glass windows that reached all the way to the ceiling bathed the rows of innumerable shelves and cabinets in iridescent light. The colors played on her coat, her blank flank, and her mane, seemingly setting it ablaze when she passed under one of the ruby rays. It was a colorful spectacle that, despite its light, would strike her years later as a portent of her darkest hours. She turned left and headed straight for the librarian at the reception desk.</p>
<p>The main vessel of the library stretched into a distance at least twice the length of the main square in Ponyville. Even though it was narrower than the square, it yet held rows upon rows of richly ornamented shelves that ran from its central aisle to both sides all the way to the piers that separated the main vessel from the outer aisles. These aisles were each wide enough for two royal chariots to pass them abreast and stretched upward to the height of Ponyville’s town hall until their pointed arches terminated in the triforium. These aisles extended to the far end of the library but were cut short on the near end by the reception area to the left and the cloakroom to the right. Amber reached the reception.</p>
<p>“We need a print of an old-looking map of Equestria as prop for a theater play. It has to be rather large. Do you have something like that here?” Amber asked. She kept her voice low as etiquette demanded.</p>
<p>“Hi there! What’s your name? I’m Fleur.” The librarian smiled at her. She was a tall unicorn, and her physique reminded Amber of Princess Luna, but that was where all semblance ended. Her coat was all white, but she had streaks of the lightest pink in her mane, a mane that looked like it had not seen a brush in days and was carelessly clipped back, probably so it would not bother her when reading. Her cutie mark showed three fleurs-de-lis.</p>
<p>Amber frowned. “I’m Amber. It’s rather urgent.” <em>Can she please cut out the singsong voice? I’m not a toddler.</em></p>
<p>Then she noticed several maps and posters hung on the wall that separated them from the left aisle. “The map over there looks old, the top-left one. It’s just a reproduction, right?”</p>
<p>“Yeah, I can get you that one.” She furrowed her narrow eyebrows. Passion flamed in her eyes. “Actually, no. It virtually breathes impurity! Lots of little mistakes: You see that island there in the southeast, for example? Well, it’s too small to be seen from here, which is all for the better because neither does it exist! Hah! There is no Apple Cider Island! They just made it up! I eclectically hate it when ponies sully maps like that, just to mark them as their own. Like dogs pissing on, uh, maps.” She started to nod repeatedly.</p>
<p>Was it the light or had her coat reddened? How was that even possible? Amber would have been amused by the odd outburst had she not been in such a hurry. Some other guests threw her compassionate glances before they returned to their books.</p>
<p>“I’ll get you a better one.” With that, Fleur vanished downstairs, off limits to mere visitors. Amber envisioned extensive, dusty archives. Minutes passed. Only thirty more to the performance, and she still had to get back to the castle.</p>
<p>She had worked so hard and risked so much to get the lead role. Her understudy might not even know her lines, one of the risks Amber had accepted. All her friends depended on her. <em>No</em>, Amber decided, <em>I’ll just take this map. No one will notice or care about its imperfections.</em></p>
<p>The poster was high up, even by adult pony standards, and clipped into a rail along with the other prints. Amber tugged at it with her magic, but it wouldn’t come loose. Ripping it off was not an option. She would even insist, she decided, that they only fasten it with magnets for the play, not tape. She had to return it unscathed after their final performance in three days.</p>
<p>When the tugging was of no use, she tried to press down on the hidden mechanism that was holding it in the rail. She wished she had feeling in her magic like she had in her hooves so she might be able to understand how the clip worked. As it was, all her pushing and pulling was to no avail.</p>
<p>All the ladders she could see hung in rails fastened to their shelves. She remembered Juniper Berry’s lessons in carpet flight, and looked around. It was of no use. The only carpets were the ones along the aisles, narrow and much too long to be useful for the purpose. In any case, Juniper had often warned her that indoor flight required considerable skill lest you launch yourself right into the ceiling.</p>
<p>There was, however, a single wooden shelf at that wall, filled with many copies of the same dictionaries and encyclopedias, which must be in high demand. Even though it didn’t have its own ladder and looked much newer than all the other shelves, Amber surmised that she could climb it and use it to reach the map that hung to its right.</p>
<p>She pressed her head against the wall to peek behind it. The wall was not as cold as she had expected but she paid no heed to it. At the top of the shelf, she could see two angle brackets pinning it to the wall. That was all the reassurance she needed. With a few dexterous moves, she hung high up on the face of the shelf. Her hindhooves were securely propped on a lower shelf while several rows up her forehooves were always scrambling to find enough friction.</p>
<p>Then she reached the top. She ignored all the dust and the pieces of trash that had accumulated there and leaned onto it to balance her weight away from the drop. She had lost all fear of heights when she trained with Juniper a mile above Ponyville with only a thin carpet protecting her from the pull of gravity. Now she was up only four or five times her own height and looked down without alarm. With her right hoof she reached over to the rail that held the precious map. From this perspective she could see the mechanism. She had to slide a nob against a spring and the map would come loose.</p>
<p>The moment she touched the clip, a deep throbbing sound filled the library, interspersed with loud cracks. For a moment she could not see where it was coming from, then she felt something akin to vertigo. Gravity seemed to be going out of its way to pull her off her precarious perch. It was a puzzling sensation.</p>
<p>When the angle brackets were sheared out of the wall with a puff of gypsum, she understood what was happening. The wall that filled her vision and that she had assumed to be as solid as the rest of the colossal building was a thin sheet of drywall, visual cover for the reception desk. Some fastings had broken away, and now it was tumbling toward her. The heavier shelf fell even faster.</p>
<p>Time seemed to slow down for Amber because she had enough of it to decide that she would bolt. She would help clean up this mess after the performance. Her friends depended on her! Maybe she would even have time to retrieve the map from the detritus. Books slid out and fell toward the ground far below.</p>
<p>Amber envisioned the trajectory of the shelf and the wall. The shelf with its tons of books would hit the ground first, then the wall would fall on top of it, breaking, probably. If she managed to fall right next to the shelf, it would shield her from the impact of the wall.</p>
<p>But time did not slow down enough. She could no longer kick off. She was already in midair, shelf and wall crashing toward her.</p>
<h2 id="the-status-quo"><a class="toclink" href="#the-status-quo">The Status Quo</a></h2>
<p>“Hey! You must be, uh, Rose Bud? Did you know that roses have all sorts of funny meanings? Like, y’know, love and happiness and innocence and all that junk! So cool, y’know? My boyfriend keeps talking about that stuff? Total flower nerd if you ask me.” The pinkish unicorn planted herself next to Amber in the crowded lecture hall and started to rummage in her bag with her magic. Her cutie mark was stylized bird, probably a magpie, Amber mused. “Oh, hold that quick, will you?” With that she hovered a paper cup of steaming coffee right in front of Amber and let go to better concentrate on her rummaging activity.</p>
<p>Amber tried to grasp the cup in her magic. Her horn cracked and sparked, an auburn aura flickered around the cup, then it fell through and spilled hot coffee over Amber’s legs before it hit the floor. With some effort, she managed to right the much lighter cup on the ground. She bit down the searing pain along her legs. First she would correct the other pony’s misapprehension of her identity.</p>
<p>“You guessed that from my cutie mark.” It was no question. “And no, my name is Amber Rose.” She broke eye contact only for a second to indicate the puddle of coffee with a glance. “I would appreciate a towel or tissue.”</p>
<p>The deadpanned remarks caught the magpie pony off guard. She stared a moment too long, then started rummaging with greater urgency until she found a pack of tissues. “So sorry about that. What’s wrong with your horn?”</p>
<p>“I had an accident some eight years ago. My magic has never been what it was,” said Amber.</p>
<p>The magpie pony dabbed at some of the spilled coffee. “Oh, so sorry to hear that, Rose Bud,” she said. Her face was a caricature of concern.</p>
<p>“It’s <em>Amber Rose</em>.” She took another tissue. Some ponies made her wonder whether she had chosen the right degree. How could she make a difference in the world when she was surrounded with colleagues who could not even get her name right?</p>
<p>“Hmm, so there <em>are</em> limits to these prophetic abilities of parents? Y’know, when they, like, totally predict their children’s cutie marks when they pick the names?” the magpie pony who had still not introduced herself said. “Now, if your name is Rose, you know what that means, right? Once that bud there blossoms, it’ll totally be like a beautiful yellow rose, y’know?” A smile, then it faded. “Except, y’know, cutie marks never change, now do they? But don’t you let it get you down that you’ll never blossom, you hear me!” She tried to make her voice sound belligerent.</p>
<p>The pain from the hot coffee lessened. “There’s nothing prophetic about it.” Amber looked her in the eyes while she quietly enjoyed to again ignore the latter half of the magpie pony’s verbal effluvium. “Ponies’ physiology allows them only once in their lives to form a cutie mark, and this usually happens as they enter adolescence. At that time, their interests are still very much inspired or influenced by their families and their immediate environments, the same factors that often inform their parents’ choice of their names. Their own names probably even influence them. Many ponies switch occupations later in their lives and find that they are just as or more gifted in their new vocation.”</p>
<p>The magpie pony looked at her blankly.</p>
<p>Moments before the lecture was about to start she turned toward Amber again. “I’m Mag Pie, by the way.” They shook hooves.</p>
<p>It was the first lecture in Mass Communication 101 in the first semester of Amber’s journalism degree. The professor introduced herself as Ms. Butterfly Peak, almost like Amber’s mother’s name, and stressed that she would prefer to be addressed with her first name, keeping the “Ms.” but that she did not insist on the academic title.</p>
<p>The stale air made Amber drowsy, but she looked up when Ms. Butterfly concluded the introductory section of her lecture with a first question.</p>
<p>“Here’s a case in point. Who knows what the great astrophysicist Tenderhoof is known for?”</p>
<p>When no one dared to speak, Amber answered without so much as raising her hoof. “He was a fraud. Till his death he made millions selling his stupid cutie mark readings to gullible parents, companies, even academic scholarship programs. Countless students were barred from their dream jobs when proper vocational counseling would’ve revealed their real aptitudes and interests.” She kept her voice calm but her passion shone through.</p>
<p>Ms. Butterfly looked askance. “No, that must be a different pony. The Tenderhoof I’m referring to is astrophysicist as I said. He’s still alive. He made an important observation of Starswirlian physics, the Tenderhoof self-consistency principle, which bars us from using Starswirlian time travel to change the past. But it’s not so much the principle that I wanted to mention but the way he …”</p>
<p>“You’re right,” Amber interrupted. But before the lecture concluded, she would actually get to correct the professor a few times on issues outside her field.</p>
<p>There were no bat ponies in the audience, so when the professor touched on the founding of Equestria in one of her examples, she felt she had to stand up for them. “You’re forgetting the bat ponies. Equestria would not have flourished without their high technology. They already had their cities inside the mountains when the three tribes settled this land.”</p>
<p>Then the lecture was over, and Amber wanted to leave. She waited for Mag Pie to clear her seat at the aisle. Mag Pie, however, turned around and looked at something right below Amber’s chin. “Uhm, Amber Rose? I’m sorry about what I said earlier, y’know? That you would never blossom? I’m sure you will.”</p>
<p>Amber quietly decided that she was all right after all.</p>
<p>“Can you tell me how you got your cutie mark?” Mag Pie asked.</p>
<p>“I was in a coma when it happened,” Amber said, “due to this accident.”</p>
<p>Mag Pie seemed uncertain how to react. Amber wanted to reassure her that there was no reason to feel sorry for her, that she had won at least as much as she had lost that day, but then Mag Pie just nodded a goodbye and dove into the stream of students along the aisle.</p>
<p>Amber did not mind much. She expected to see her again the next day.</p>
<div> </div>
<p>Amber stepped outside of the department wing and was greeted by the smile of her older brother, Damask Rose.</p>
<p>“Hey bubs!” she exclaimed.</p>
<p>“Hei sisko!”</p>
<p>He had called her <em>sisko</em> for as long as she could remember. When she had just started school, he was abroad on a student exchange where he must have picked up the foreign term—and he was not known to forget things.</p>
<p>Amber cleared the stairs with one jump and embraced him.</p>
<p>“How was your first day, Amber?”</p>
<p>“It was fun!” She slumped against him as they walked and felt the irregular bobbing of his limp. Thankfully, it did not cause him pain anymore. “I think I’ll have some fun debates with the professor. And someone spilled coffee over me.”</p>
<p>“That was fun too?”</p>
<p>“Naw.”</p>
<p>“Did they survive?” he quipped.</p>
<p>She laughed. “The professor or the filly with the coffee? Both survived. She was alright.”</p>
<p>“The lakelet with the spouting fountain is nearby. You remember it, right? The one where we inaugurated your first model ship, the shallop, when we visited Canterlot the first time? We can take a swim. The coffee stain will wash off.”</p>
<p>“Oh, bubs, that reminds me. You won’t believe what I saw the other day!” Amber did not wait for a reply. “You remember that model ship I loved so much back then?” It had been eight years since that vacation, but she knew her brother would remember it. She had entreated her parents for days to buy it, but it was much too expensive.</p>
<p>“Of course. A three-masted, lateen-rigged xebec.”</p>
<p>“Yes! Yes! With the weird figurehead. I saw it again in a store the other day, it was even on sale,” Amber said. “The bowsprit was as long as I remembered, almost as long as her foremast, but I had forgotten that it was raked forward just as her mizzenmast was raked back. But I immediately recognized her when I saw her there.” She enjoyed reminiscing about the stay even though it had ended in disaster for her—and her brother also had a painful encounter with hotel furniture, she recalled.</p>
<p>“Did you buy her?” Damask asked.</p>
<p>“Naw. It would’ve been just another piece of memorabilia.” They walked in silence for a moment. Then Amber remembered her brother’s suggestion to go swimming in the lake and said, “I’ll go straight home, shower, and unpack some more of my moving boxes.” Hastily she added, “But please tag along.”</p>
<p>“Aw! What’s the difference between a shower and a bath in the lakelet? Soap is overrated.”</p>
<p>“Not at all,” Amber said. “Soap is lyotropic liquid crystal, it’s like bathing in gems!”</p>
<p>“But do be careful that you don’t end up like our Ponyville soap shop, burned to a crisp because <em>somepony</em> forgot about the no magic rule.”</p>
<p>“C’mon, I was just a toddler!” She knew he was just teasing her with the old story but quickly changed the topic anyway. “Say, what courses do you have this semester?”</p>
<p>“We have mandatory math and physics courses, but I’ve also signed up for a course in plant engineering. Dad recommended that I specialize in that direction, and it does sound interesting.”</p>
<p>At one point on their walk home, when they were passing a hypnotherapist’s practice, Damask noticed a tail—they giggled about the pun. He thought they were being followed when he repeatedly saw a pink tail vanish around corners when he turned. But why would anyone want to follow them?</p>
<p>At Amber’s apartment they hugged goodbye.</p>
<div> </div>
<p>The next morning, Amber awoke to the soft rays of the morning sun playing on her face. Or maybe it was just the flap of the newspaper falling through the mail slot.</p>
<p>Sips from her hot cup of espresso washed new life into her limbs while she read the latest news. After years of advocacy, Princess Twilight Sparkle’s campaign for a greater role of bat pony history in school curricula had reached its tipping point and was finally gaining broad support from teachers and parents. Amber still remembered the librarian fondly from the time they both lived in Ponyville so many years ago.</p>
<p>But there were also bad news. Crop failures were wreaking havoc in the Griffon Kingdom for the third successive year. Their reserves had long run out, and they depended on donations from Equestria, Prance, and other nearby nations. According to the latest estimates, over eight million griffons suffered malnutrition. The article concluded that if only they had been able to predict the catastrophe a few years in advance, they could have held back greater reserves.</p>
<p>When Amber turned the page again, a flyer sailed to the floor. She bowed down and picked it up. It was much too unprofessionally printed to be from the newspaper press she decided, but it did promise a hundred bits for just a quick questionnaire. That was more than usual but not so much as to reek of scam. The room would be crowded.</p>
<p>Thanks to the comprehensive social security in Equestria, Amber never thought of herself as poor. She had everything she needed and could even engage in many of the cultural activities Canterlot provided. A few extra bits were still always welcome.</p>
<p>Amber was amazed to notice that it was perfectly timed, too, as if the organizers of the study had predicted her course schedule. She would stop by the department of post-Starswirlian magic, take the questionnaire, and most likely still have enough time to get back to the journalism department at a leisurely trot.</p>
<p>She recognized that the address was on the premises of the university, so she found it easily. She had expected to find a Rainy Day–style building mirroring in its architecture the outlandish spirit of the department it housed. Instead she found that the address just lead to another part of the same plain and functional building that also held the department of Starswirlian magic as well as several other schools of magic. She stopped herself from double-checking the flyer. <em>Room 023. That must be on the ground floor.</em></p>
<p>Although Amber was barely a minute early, only a single pony waited in front of the room. She was a tall, white unicorn with bluish mane and tail, and heavy glasses. Her cutie mark were three fleurs-de-lis. <em>A hundred bits for a simple questionnaire? Scores of ponies should be milling around that door. Or are they already inside?</em></p>
<p>The door was ajar. Amber decided to just walk past for now as if she was headed for the lecture hall, but peek inside the room from the corner of her eye.</p>
<p>“Hi! Are you here for the questionnaire?” the other unicorn asked.</p>
<p>Amber’s heart raced. <em>Darn, I used to be better at this.</em> “Uh, yeah.”</p>
<p>“Wonderful! The room is taken by another group unfortunately, but we can go to my office.”</p>
<p>“Shouldn’t we wait for others, Ms. …?” Amber wondered whether it was wise to ask two questions at once if you wanted answers.</p>
<p>“Uhm, yeah, we can wait a few more minutes. You can call me Liz.”</p>
<p>“What are you investigating with the questionnaire? Empirical studies, or any interest in the real world, is rather atypical for post-Starswirlian research, I understand.” She had not intended to make the question sound disrespectful. Neither did she care. <em>White fur, fleurs-de-lis—who does she remind me of?</em></p>
<p>“Oh, don’t tell me about the ‘real world’!” the researcher said much too loudly. “I’ve been a model, a nurse, and a cartographer; I know everything about this so-called ‘real world.’ Change the projection and you got a completely different ‘real world.’ And all of them are eclectically awful.” She wheezed. “Ponies go hungry, are hurt, or fall sick, the causes so complex that no pony can make sound predictions. We sully our models with our arbitrary simplifying assumptions then measure the awful precision that remains, giddy like the foal who drew a watch on her wrist and saw that it gave the right time once a day. There is just no escaping the eclectic impurity once you put units in your magic.”</p>
<p>“Twice.” Amber was already starting to enjoy this batty conversation. <em>Was she really using forms of </em>eclectic<em> as general-purpose swear words?</em> “So your conclusion is to cloister yourself in your imaginary realms of pure, theoretical magic and forget about the real world?”</p>
<p>“Once this is all over.” The unicorn nodded slowly and repeatedly, her eyes suddenly downcast.</p>
<p>Amber was ready to either chide her for abandoning her fellow ponies or encourage her to let actual problems give focus to her abstract pursuits, but this cryptic statement caught her off guard. “What do you mean?”</p>
<p>“Oh, I mean just this particular, uh, project of mine. It has been occupying my mind for some years now.” She seemed to notice her nodding and stopped it. “For which I’m glad, don’t you worry. It has given my life purpose. I often picture myself on my deathbed evaluating my life’s accomplishments and deciding whether it has been worth its while. I don’t want to breathe my last breath in the knowledge that I let the impurity prevail, Amber, that my life had been in vain.”</p>
<p>“Yes, that’s why I chose to study journalism—well, apart from the stuff about impurity.”</p>
<p>“But you are aimless! You’ve been searching the whole ‘real world’ for any kind of purpose you can appropriate for your life, and yet here you are: aimless!” She took a deep breath and forced a calmer register. “And it’s all my fault. I’m so sorry.”</p>
<p>Amber considered that. As crazy as it all sounded, it was not without merit. “Why is any of that your fault?”</p>
<p>The researcher started to nod again for a few seconds. “I think we’ve waited long enough. Let’s go now. My office is over in the castle.”</p>
<p><em>Way to change the topic, but seriously, she has an office in the castle? The royals don’t let just anyone have an office in the castle!</em> Amber wondered again where she might have seen her before. Maybe at a Summer Sun Celebration? Or at Twilight Sparkle’s coronation years earlier? She wished she had her brother’s memory. But something else was still on her mind as well. “So why do you conduct empirical research?”</p>
<p>“Oh, believe me, it has taken me some years of soul-searching, and I don’t think I’ll do it again. Appreciate it!”</p>
<p><em>Evasive answer number </em><em>two.</em> It was exhausting to try to get straight answers from that pony. Amber decided to just get the hundred bits and forget about it.</p>
<p>They walked in silence until they reached the castle. The researcher produced a key and lead the way through a side entrance and then through a maze of hallways, staircases, and a few more locked doors. Amber noted with relief that some of the guards even greeted her, so she was not just some grifter who somehow got her hooves on the right keys.</p>
<p>They were walking past many simple, functional offices now, each with a window in the door that allowed Amber to inspect the interior. Not much was there for her to see since most were empty, but the wide windows overlooking a courtyard admitted much of the morning sunshine and even bathed the hallway in inviting hues.</p>
<p>She was again levitating her key in front of her when she turned to Amber. “Pardon the mess. I used to collect newspapers back in the day. I stopped when I missed an issue, but I couldn’t bring myself to trash them yet, so I dumped them all in here.”</p>
<p>The researcher opened her office and asked Amber to take a seat in front of her desk. Amber noted at once that there was only one chair there. Between the stacks of old newspapers, books, scrolls, maps, and desultory miscellanea, hardly another chair could have fit into the place. Most striking, however, was the magical equipment that lined the walls and even the ceiling, devices completely inscrutable to Amber.</p>
<p>“Let’s get right down to business,” said the researcher. She took a notebook from a drawer and wrote down the date. “You haven’t told me your name. Don’t worry, I will not publish it in any fashion, but I would like to be able to get back in touch with you should the circumstances require it.”</p>
<p>“Amber Rose.” While she spoke, she looked her interviewer in the eyes but her attention was absorbed by a strangely pulsing apparatus that hovered above the researcher’s head.</p>
<p>“Where did you grow up, Amber?”</p>
<p>“In Ponyville.”</p>
<p>“How long have been living in Canterlot?”</p>
<p>“For about a month.”</p>
<p>She scribbled into her notebook. “Hmm. Okay. That would be all, but please wait a little moment longer. I have something for you.”</p>
<p>Amber cocked one eyebrow. <em>That’s it? Already?</em></p>
<p>The researcher removed a hundred bit bill from her wallet and levitated a large tube from the wall over to Amber.</p>
<p>Amber noticed with a little surprise that she made no indication for Amber to take the tube into her own magic. Surely, it would have been much too heavy for her. Instead, the researcher leaned it against her chair and set the bill in her lap. In that one apprehensive moment, however, when Amber thought the tube would crash to the floor next to her, she noticed a framed certificate on the wall where it had leaned. It was a PhD certificate in cartography and geovisualization for a certain Fleur de Lis.</p>
<p><em>Fleur! Of course!</em> Now she remembered who that pony reminded her of—apart from the mane, which she may have dyed.</p>
<p>“Amber, do you sometimes wish you could go back in time to undo your past mistakes?”</p>
<p>Amber had not considered it in a long time. She was still reeling from her epiphany and trying to disentangle its possible implications. She could not reminisce about past mistakes now. She looked at Fleur quizzically.</p>
<p>“Well, today is your lucky day,” Fleur continued as if Amber had answered. Then, not wasting a second, she shot a ray of magic energy from her horn into the contraption above her. A cone of light cut outward from it and through air, desk, books, newspapers right toward Amber. When she wanted to jump away, it had already enclosed her on all sides.</p>
<p>Amber groaned under a tremendous pressure, as if she were diving much too deep. Then a throbbing headache set in that knocked her unconscious within seconds.</p>
<h2 id="the-play"><a class="toclink" href="#the-play">The Play</a></h2>
<p>“Did you feel that, too?” her sister asked from the bed.</p>
<p>She nodded. It had felt like a surge of strongly amplified magic somewhere nearby, probably within their castle. She had never felt anything like it before. Two ponies had been involved in it. She would recognize their magic.</p>
<p>Soon guards would arrive and report to them what had happened, she thought.</p>
<div> </div>
<p>She heard a rustling of sheets. So her sister had felt something as well.</p>
<p>It had felt like a surge of strongly amplified magic somewhere nearby, probably within their castle. She had never felt anything like it before. One pony had been involved in it. She would recognize her magic.</p>
<p>Soon guards would arrive and report to her what had happened, she thought. They must not disturb her sister. She was still so fragile. “Sweet dreams,” she whispered and left the room to intercept them outside.</p>
<p>None came.</p>
<div> </div>
<p>Moist. Sticky. Her head lay in a puddle.</p>
<p><em>I’m bleeding!</em></p>
<p>Amber woke up with a start to a bare, twilit room. The sudden motion let the migraine explode and knocked her right out again.</p>
<p>This time she dreamed.</p>
<p>“Amber, could you help me with something?” Cheerilee asked.</p>
<p>Amber was back at school in her old classroom. The other students were filing out while she was still packing her books and quills. Cheerilee walked toward her against the flow of the students.</p>
<p>“Some of the students are falling behind on Starswirlian magical theory. I don’t have the time to give them extra lessons, but you seem to be intimately familiar with the material.”</p>
<p>She did not have to think this over. “I’d love to!”</p>
<p>Suddenly, Amber felt that she was dreaming, and she also remembered the episode the dream was showing her.</p>
<p>Back when she still attended Juniper Berry’s carpet flying lessons, Amber sometimes envied her the position of authority that her skills afforded her. Now Amber would become like a second teacher of Starswirlian theory to the other students. The discipline of magic would be taught by an earth pony and a unicorn who was barely able to lift her quill. Amber loved it.</p>
<p>A few years earlier, her accident had seemed like a curse, but it had opened her so many doors she would otherwise not even have noticed that she now thought of it as nothing short of a blessing.</p>
<p>Then she remembered that she had also betrayed her friends that day. Her classmates all blamed Little Cedar for it. She was no longer at Amber’s school and did not know that her name had become a curse, so Amber let them. She was too afraid that their wrath would turn upon her.</p>
<p>The classroom faded.</p>
<p>“Morning sisko! I bought you some quills. The tiniest and lightest Davenport could find,” her brother greeted her.</p>
<p>“Morning bubs.” She yawned involuntarily although she did not feel tired at all.</p>
<p>“Have you been reading all night again?”</p>
<p>She lay on her bed, one hoof holding open a paperback book. “I’ve been thinking that the weight of the quill might be only part of my problem. I’ve always tried to draw single letters, and then I was exhausted after two or three of them. But really I can envision whole words and even fragments of sentences, so I tried to write in greater chunks.” She closed the book to reveal a page full of nonsensical sentences underneath. “It works like a charm. No pun intended.”</p>
<p>“That is amazing!” Her brother gaped at the page. “You’ve never written that much or such long sequences.”</p>
<p>“I know!” It was exhilarating to be able to write again. “It’s all in this book. Discoveries that go back to Starswirl the Bearded, although the authors apply them to traditional unicorn martial arts.”</p>
<p>Amber again realized that she was dreaming. This one was a real memory too. Amber could hardly remember the dream sequence at school but she felt that she had traveled years further into her past.</p>
<p>Damask and the page of scribbling faded. Her journey continued.</p>
<p>“But Amber, our parents aren’t unicorns, and neither would they want to be,” her brother said.</p>
<p>Walking was still painful for him, but he limped over to her anyway. It had been their first night back at home. The Ponyville hospital was severely damaged, so she had to stay in a hospital in Canterlot until her injuries had healed sufficiently, and the doctors released her.</p>
<p>She no longer had attacks of panic and disbelieve whenever she remembered that her horn was useless now. The attacks had burned her out and left her in an ash cloud of despair. It did not help that she had singlehoofedly ruined her school’s theater play and received an inauspicious cutie mark in the process. She lay on her bed flaccidly, weighted down by a blanket of down and a blanket of regret.</p>
<p>But this was not the Amber she wanted to be.</p>
<p>“You’re right, I don’t need magic, but still,” she brushed her metaphorical ashes away and threw off the blanket, “I will regain it if only to prove to myself that I can do it.”</p>
<p>Damask’s face lit up. She saw that he recognized his dear old sister again, the one both had thought lost after the accident—and Amber recognized herself as well for the first time in weeks.</p>
<p>Then Amber gained distance again and saw the dream for what it was, another memory. Within a year she would relearn writing the unicorn way but she would also find joy in mastering her daily chores the earth pony way.</p>
<p>When she was a child, her dad never thought it necessary to teach her all the little tricks and mnemonics that earth ponies use to paint, clean dishes, and cook magnificent meals; nor did she notice them. A whole new aspect of her culture revealed itself to her, just like the curtain that parted for the first performance of their theater play, still in Ponyville.</p>
<p>The bedroom faded.</p>
<p>Town Hall was crowded with parents, siblings, teachers, and a few guests.</p>
<p>“Today’s performance,” Cheerilee addressed the audience, “is also a dress rehearsal for the two performances in Canterlot in two weeks. I have received permission to extend the school holidays by another week, so we can fully concentrate on our play, and show Princess Celestia what the students of our little town are capable of.”</p>
<p>After a few more words of introduction, Cheerilee left the stage. Amber gave Juniper a sign, then she levitated the black backdrop of the stage to one side to reveal a large aerial picture of Ponyville, Canterlot, and vicinity that was painted on the wall behind it. At the same time, Juniper activated the spotlights and bathed the stage in light.</p>
<p>All six of the main actors, Amber among them, and a few in supporting actors filed out into the light. Some of them were the understudies of the students who would play in Canterlot. They took turns, so everypony could play at some point. Amber had personally overseen the casting and made sure that everypony felt at home in their role—except for her own understudy. Amber had picked Little Cedar because she had stage fright and was happy that Amber generously volunteered to play in all performances and rehearsals.</p>
<p>Pangs of guilt cut the dream short and shook her awake.</p>
<p>Amber’s second attempt at waking was a little more successful than her first, but the headache ran all the way down through her spine and engulfed her body and mind. There was no room for thoughts. She lay in complete darkness, but she had long not attained the level of consciousness that would allow her to wonder about her whereabouts.</p>
<p>Hours passed.</p>
<p><em>Light good.</em> A first observation pierced her mental cocoon. It was dawn.</p>
<p>The throbbing of her head seemed to lock out any sense of time, but it abated gradually and soon she remembered the darkness and her minimalistic greeting of the new day. Memories of the meeting with Fleur returned as well.</p>
<p>She lifted her head and was relieved to find that she lay in a puddle of her spittle, not blood. Her head hurt doubly, once from the receding headache and once from lying on concrete for the better part of a day and a full night.</p>
<p>She brushed off her spittle as best as she could and looked around. The room looked exactly like Fleur’s office but may just as well be one of the other empty offices on that level. An assortment of slashed and singed detritus that she recognized suggested that it was the same room: stacks of newspapers and books, some cut in half by the cone of light, the section of Fleur’s desk that had been cut out as well, even a piece of the concrete floor, and right next to her the chair, the tube, and one hundred bits.</p>
<p>So far she had only raised herself enough to lean her head on a hoof. Now she sat up. Another surge of headache made her queasy. When it lessened, she tried standing.</p>
<p>Something was weird.</p>
<p>The ceiling was higher, the chair bigger. The disorientation lasted only a second until she looked at her hooves and legs, then her flank. Her legs where shorter, her cutie mark gone. <em>Bummer.</em></p>
<p>The the realization set in. <em>My body! Fleur stole a decade of life from me!</em></p>
<p>The adrenaline blocked out the ache in her head. A vignette of red shrouded her vision. “Give me my life back!” The cry echoed from the bare walls and stone floor.</p>
<p>She bucked hard and hit the chair at random. It skidded a few inches and tumbled slightly. <em>All that I’ve learned, all that I’ve been through, it can’t be all lost! All my memories!</em> She started to cry.</p>
<p>Then she realized that she did remember her life. <em>If I had lost my memories I wouldn’t notice a thing right now, would I?</em></p>
<p>The realization calmed her down. Building model ships with Damask, performing in Ponyville, rediscovering the joy of writing, teaching her class in Starswirlian magic, moving into her own home, all the pleasant memories of her past were still there. So were the unpleasant ones. She did not linger on those.</p>
<p>She wiped away her tears. She often cried of joy or sadness, but it had been years since she had felt such panic. Surely some gifted magician could restore her adult body to her, or if not she may get to live a few years longer. She had not lost her personality, her identity. That was what mattered most to her.</p>
<p>She looked at her short legs and shook her head. Her eyes still burned, and the headache was slowly encroaching on her senses again. She did not want to think anymore. She dragged herself to the wall and slumped against it.</p>
<p><em>What if Fleur actually sent me back in time?</em> The implications seemed overwhelming in their complexity. Amber did not feel like thinking at all. <em>No, there are no time travel spells that would’ve lasted so long.</em></p>
<p>She felt like dropping off to sleep again and forgetting about all this nonsense. She wanted to just close her eyes and wait for it all to go away by itself. The thought even struck her as logical for a moment, which gave her pause. <em>Was my brain affected after all or am I just confused?</em></p>
<p>A few minutes later she felt strong enough to stand again. The mysterious tube still lay next to her. It gave her an idea. An auburn aura enveloped it tenderly—then smashed it against the ceiling. <em>Whoops.</em></p>
<p>Magic. She had full access to it again, or at least that of the average unicorn. She should be overjoyed. She felt something akin to it, but it was not joy. It was something vicarious. It was just as if someone had gifted her that model building set of the ship with the strange figurehead, a kind of mediated joy for her filly self who had still pined for her lost magic, joy translated for her adult self only through nostalgia.</p>
<p>But magic was useful, especially since she had no backpack to carry this weird tube in—whatever it may be good for—so she was glad to have it back.</p>
<p>More so she was curious. She lowered it to eye level and peeked inside. It was empty.</p>
<p>No, there was something tightly wrapped along its inner surface. Of course, Fleur PhD would give her a map, she should have known that at once. She pulled it out and unrolled it in the air.</p>
<p>It was as if meeting a long lost friend or old enemy again. Both maybe. Most of Equestria’s towns and villages were pictured disproportionately large, the whole thing was a little skewed, and now she could see the phantom Apple Cider Island too. She rolled it back into the tube together with her hundred bits. It was time to find out what else Fleur’s spell had changed besides Amber’s body.</p>
<p>Still swaying, she walked to the door. The headache was getting more bearable and the vertigo it caused faded, but the suddenly so different proportions of her body still made it difficult to walk. The door was not locked and the hallway was empty. On the one hoof, she would have welcomed the sight of guards there, or of anypony, so she could ask them what year it was, but on the other hoof, she was not quite sure how common it was for a lone filly to cross through these parts of the castle, which were most likely off limits to visitors.</p>
<p>She had memorized part of Fleur’s zigzag route through the building, but tracing it back proved more difficult. Her memories of the past days seemed almost years distant while memories of her foalhood shoved their way back into her consciousness.</p>
<p>After a few turns that only lead her into less and less familiar passageways and galleries, she felt a flash of fear at being lost, something she had not felt in years. She shook away the childish emotion and changed her strategy. If she managed to go in the same direction long enough, she had to reach an outside walls at some point. Even if it should be one overlooking nothing but the great drop along the cliffside, she could follow it until she reached an exit toward the city.</p>
<p>With rooms to both sides of the hallways, she could never be sure whether she was in the midst of the structure or in fact already close to an outside wall, but she felt confirmed when she came upon a door like the ones Fleur had to unlock for them. From this direction, it had a handle and opened readily for her, so most likely her route was outward. She had to strain a little to reach up to the handle with her forehooves to depress it. Only when she watched the door swing shut with finality behind her did she realize that she could have used a certain freshly restored function of her horn for the task.</p>
<p>Another door and she left this more plain and functional wing of the huge building and entered one of the many towers. Her hoofsteps echoed more loudly on the marble tiles than in the carpeted hallways. A path of lighter tiles circled a staircase in the center of the tower and lead out onto a wide balcony that boasted an ornate balustrade much too high for Amber to peek over. She walked out into the fresh mountain air and stuck her head through a gap between two mythological creatures chiseled out of the sandstone of the balustrade.</p>
<p>She could not see the windows in the tower itself, but judging from nearby buildings, she must be on the height of a fourth floor. Beneath her lay a plaza she recognized. The morning sunshine played on a squat structure almost like a temple in its center, which lead down to what her brother had called the Hypogeum Abscititious, where busts of many great ponies of history were on display. If she could get down there, she would find her way home easily.</p>
<p>But what if she had not just transformed into her younger self but actually traveled back in time? Surely that was impossible. No known time travel spell could have lasted this long. But what if? Where would “home” be for her? Miles and miles away in Ponyville and already occupied by her doppelganger. She dismissed the thought. The paradoxical implications of time travel were something she would consider if and only if it should turn out that she had actually traveled back to her foalhood.</p>
<p>In her reverie, she noticed the hoofsteps only when they left the winding staircase and reached the platform. She turned.</p>
<p>“Young lady,” a guard said with mild surprise. She hesitated while she looked from Amber to the tube hovering next to her and back to Amber. “Unless you have a letter of safe conduct on you, I don’t think you are allowed in this part of the …” She checked her watch. “Eh, scratch that, quitting time. You looking for someone?” She took off her helmet and tucked it under her wing.</p>
<p>“I am a little lost,” Amber said. It might be true in a comprehensive, four-dimensional sort of way, she realized. “Say, what year is it?”</p>
<p>“Hah! Everypony keeps asking that. What do I know!”</p>
<p>Amber stared. “Really?”</p>
<p>“Just kidding.” The guard told her the date. “So you time-traveled here or something?” She did not even seem surprised.</p>
<p>She shuddered. <em>So I did go back in time. I don’t want to have to relive, what, eight years!</em> Aloud she replied, “Naw, a ten-minute walk away through the corridors over there.” Amber was indicating the direction with the tube when she realized that the <em>here</em> was more likely meant temporally. “I’ve been trying to find a way out of this labyrinth.”</p>
<p>“It’s simple enough when you get used to it. Come, tag along, I’m going home anyway,” said the guard. She lead the way back where she had come from.</p>
<p>This last part of the way was simple enough indeed. Amber thought she could have found it easily on her own, but with the guard at her side, she did not have to worry about running into any other less jovial guard ponies. What she did worry about were her chances of getting back to her present, or what she still considered <em>her</em> present. Could Fleur help her? The current Fleur—even assuming she was willing to help her—might not know the first thing about time travel spells. Amber dismissed the idea.</p>
<p>“You’re not stealing some sort of painting there, are you?” the guard asked when they reached the exit of the tower.</p>
<p><em>Why painting? It’s a map.</em> It took her a second to realize that the guard could not have known that. It felt as if her adult brain were at war with her filly brain.</p>
<p>“No, it’s a map someone gave to me.” <em>You could not have been more vague, Amber, could you? Why are you thinking in the second person anyway? What if she asks me who gifted me the painting? A librarian from the future?</em> She shook away the thoughts. The burden of proof was not upon her. She had been caught trespassing, not stealing.</p>
<p>“Cool. By the way, I’m Praetoria.” The guard clapped her on the shoulder. “You got someplace to go?”</p>
<p>Amber hesitated. The headache was fading, but she could not possibly walk all the way to Ponyville in her condition. The fading pain in her head also admitted feelings of hunger and thirst for the first time.</p>
<p>“I can invite you to the guards’ commons,” Praetoria continued after a moment as if reading her thoughts.</p>
<p>“I’d love that!” She sounded more euphorical to herself then she had intended. “I’m Amber.”</p>
<div> </div>
<p>The castle used an interleaving system of many different shifts so that no collective changes of guards could open windows to intruders, however unlikely those were. As a side effect, the great hall of the commons did not have any set peak hours. Apart from random fluctuations, its long rows of antique tables and chairs were always lined with a comfortable number of off-duty guards, all of them elite troopers ready to fend off changeling hordes at a five- or ten-minute notice—because who could waste such yummy food?</p>
<p>Anything would have tasted yummy to her in her state, Amber thought, and she had to exercise great restraint not to eat too hastily. Praetoria also brought several bottles of water to the table and finally mango lassis for both of them.</p>
<p>They took a few sips, then Praetoria indicated the entrance with one wing. “Oh, look who’s coming!” She winked. Three ponies walked into the hall, escorted by two guards in armor.</p>
<p>“Bubs! Mom! Dad!” Amber was delighted to see them, but her delight was dwarfed by theirs. They were all so much younger, especially her brother!</p>
<p>“Hei sisko!” A moment later they collided halfway in a big ball of fur and feathers and hugs.</p>
<p>Praetoria walked up to them a moment later. “Your parents had asked the royal guard for help when you didn’t come home overnight. Another day and we would’ve launched a full-scale search. I took the liberty of calling them here.” For a moment, she observed at the four ponies with tears of joy in their eyes. “No family issues then I assume.” She looked relieved.</p>
<p>Amber was surprised how deeply this reunion touched her. She had not believed them lost for a long, wakeful night as they had, in fact she had not expected to see her parents again for several weeks.</p>
<p>Praetoria waved away everypony’s thanks and guided the family out. The four of them walked in silence for a minute, very close to each other, until they were out of earshot of the castle. The tube was bobbing in the air between them.</p>
<p>Her mother folded a wing over Amber, who wondered how much she could tell them without worrying them even more. She herself hardly knew what had happened or how she was supposed to get back.</p>
<p>“Amber, where were you?” Her mother asked the obvious question. “You went into the hotel bathroom, and ten minutes later you were just gone.”</p>
<p>“It’s a long story.” <em>Eight years and a few months.</em></p>
<p>“All right?” her mother encouraged her.</p>
<p>“Hmm, I’m not precisely who you think I am.” At this, her parents looked at her. Not her brother, however, but surely he also mentally reviewed the last few minutes. “I am Amber, but, as crazy as it sounds, someone sent me back here from the future.” <em>Hopefully they’ll think I’m just talking about a few days or weeks. How would they react to having missed eight years of my foalhood?</em></p>
<p>“You don’t look any older,” her father said.</p>
<p>“I don’t entirely understand how or why either,” Amber said. “I think the spell imprinted my brain, this brain,” she tried to nod upward, “with the knowledge and impressions of my future self.”</p>
<p>“Do you know who did that?” It was her brother who asked. Amber saw that he already had his cutie mark of fountain, ship, and rose.</p>
<p>“A librarian here in Canterlot. But I doubt her current self knows anything about it.”</p>
<p>“We have to find her anyway, but how far in the future did all that happen?” It was her father who asked the question Amber had been afraid of.</p>
<p>She hesitated for a moment. “I’m a little over eight years older than the Amber you knew a few days ago.” No pony replied. “This must come as a shock to you; my character must’ve changed radically; I never meant to take the Amber you knew away from you; I’m sorry this …”</p>
<p>“Hey, hey, easy there. I’m so glad we get to skip your puberty,” said her father flatly but could not suppress a slight smile. “Oh, you don’t happen to remember any lottery numbers?”</p>
<p>“I don’t have bubs’s brains unfortunately. Besides, those are all time travel proof. Each ticket influences the order in which the balls are inserted into the lottery machine.”</p>
<p>“Do you feel up for the theater play today? If not, Little Cedar can always take over for you, you know?” her mother asked.</p>
<p>“That’s today‽” Her mother must have flinched at the sudden outburst because Amber felt a tug on the wing that enveloped her. Both of course remembered how much effort young Amber had put into the play, but it had even deeper significance for Amber than her family could guess.</p>
<p>“No, mom. Little Cedar hardly knows her lines and cues. She never expected that I would… I mean, she completely counts on me showing up.” Amber realized that she could not recall any of her lines and cues either. “And she’ll just disappear back to Cloudsdale or whatever before the final performance. Cheerilee will be forced to cancel it.”</p>
<p>“Why would she do that?” her mother asked, but Amber suspected that she was worried about something else entirely.</p>
<p>“She never wanted to be on stage, and maybe she just felt little responsibility for our play from Ponyville. But then again I never reminded her that even I could be dis… unable to make it, even though it was part of my responsibility.”</p>
<p>“It <em>is</em> part of your responsibility, you mean. All this time travel stuff has also afforded you another chance to change, uh, …” Her mother gulped. “What exactly happened to you in the future? Or is about to?”</p>
<p>“It’s not going to happen again. This time I have a show to put on!” Fleur had been wrong to think that Amber still wanted her magic back, but even a painted watch is right twice a day. What Amber had learned from the previous iteration of this day was that it was not about getting past such an experience. It had to become a part of who she was. The only aspect that she refused to make a part of herself was the missed and ultimately ruined play, the sadness in the eyes of her friends, and that Little Cedar would become the school’s anathema when it had been Amber’s wager that was to blame. Her mother was right. This was Amber’s chance to correct her mistake.</p>
<p>Amber was ready to turn on her heels. “Bubs, do you know where I keep the script?”</p>
<p>“It’s in the hotel room on the Chippendale escritoire partially covered by a ‘do not disturb’ sign.”</p>
<p>“That would take too long. I’ll go back to the castle. Somepony will have a copy for me.” Amber ducked from under the wing, gave her family hugs, and hurried back the way they had come. The event would start at noon and run till the evening. Even if their class was among the last to present their performance, she’d only have a few hours to relearn her role.</p>
<div> </div>
<p>The venue was an oblong hall that could serve any number of purposes but had been set aside for three days for the theatrical spectacle. It was shaped like a nave with a wide space where soon rows upon rows of parents, teachers, and guests would mill, save for a narrow aisle through the center. This nave, however, culminated in a round platform where the chancel would be, the stage. On this stage the groups of students were taking turns rehearsing. The rest of the place was teeming with all the other groups who either waited their turn or had set up impromptu stages between the columns that lined the walls.</p>
<p>It had been hard enough to find this place in the labyrinthian castle. She thought she would remember the way, but it had been too long ago. Now it proved similarly hard for Amber to find her classmates. The maroon carpet that marked the aisle worked like a magic force field in that it sliced through the agitated masses of fillies and colts to allow Amber to pass. She scanned the faces left and right. No pony seemed familiar. Halfway through the hall, she found them. They had just finished a rehearsal and were removing their minimal set and props from the stage.</p>
<p>“Amber! Where’ve you been?” Cheerilee darted toward her.</p>
<p>“There was a contingency. Do you have a script for me? I didn’t have time to fetch mine.”</p>
<p>“You could’ve sent somepony to tell us. We’ve been sweating blood and water! Here, take mine,” she shoved a small stack of pages against Amber’s chest, all filled with scribbled notes. “I think I know it all by heart at this point.” She was about to dart off again. “Oh, and find Juniper. She was looking for you.”</p>
<p>Amber had no time to reply. Cheerilee was already on the other side of the hall comforting Featherweight. The colt claimed Twist had called him an idiot, which she vehemently denied. He was sniffling and sneezing, so that Amber made a mental note to ascertain that his understudy was well-prepared. She stored the tube with the heap of her class’s props, and went looking for Juniper Berry.</p>
<p>As she had expected, Amber found the unicorn backstage in the fly loft, where she acquainted herself with the lines and blocks. It was a small shock each time she saw a good friend reverted into a filly, but they had been close friends throughout most of their time at school.</p>
<p>Amber must have felt, even as filly, that Juniper was more mature than most of the other students, and that she could learn a lot from her. Thus, she had been privileged to take part in Juniper’s inofficial carpet flying lessons for unicorns. Juniper always enjoyed telling the anecdote of her parents’ surprise when she came home with a small image of a flying carpet ringed by a magical aura where they had expected a bunch of berries to pop up.</p>
<p>“Hey, Juniper, did you have time to fly around Canterlot already?”</p>
<p>Juniper looked at her askance. “<em>You</em> insisted that we focus on the play and wait with the flying until after the last performance the day after tomorrow, don’t you remember?”</p>
<p>Amber had no recollection of it. “Yeah, I’ll trust you on that. There’s a bunch of things that I need to tell you, but first, you were looking for me?”</p>
<p>“It’s all rather badly planned here. All the groups want to use the stage, but we only have a few more hours, so we only do the scenes where we need this stuff at all,” Juniper said, pointing at the lines all around her. “Could you go through the play scene by scene and see if there’s anything we forgot, any missing props or missing ponies?” She giggled softly, but Amber knew that behind that shy giggle stood a pony who met her ostensible limitations with pure irreverence, something she taught Amber just as she taught her class of unicorns flying.</p>
<p>Amber had often applied the wisdom this young pony had conferred to her, and it seemed unappreciative to undo the key event that had enabled Amber to glimpse its full significance. Yet she had made the decision hours ago, and now she would carry it through.</p>
<p>“Juniper, I already noticed one missing item. The Ponyville town hall had this picture of our region of Equestria where we indicated our town and the Everfree and the old castle; this stage does not. But I brought a map with me.”</p>
<p>“Oh perfect.”</p>
<p>Nothing changed. She could still remember her accident and the years that followed. Evidently, this was a completely new mode of time travel, one that allowed you to have your cake and eat it too, or more likely they were two cakes on different timelines. But Amber did not mention these observations to Juniper just yet.</p>
<p>“Oh, and, Juniper, you should look into finding backup for yourself too, somepony else who can also work these ropes. I’ll go down where the others are and learn my lines.” Juniper stared questioningly, so Amber added, “I forgot them. I’ll explain later. It should be a breeze to relearn them since I knew them once.” Amber hugged her sideways in a way that she hoped would inspire confidence. “We’ll have a whole lot of fun today!”</p>
<div> </div>
<p>“Does my crown no longer count now that I have been imprisoned for a thousand years? Did you not recall the legend? Did you not see the signs?” Diamond Tiara’s impression of Nightmare Moon was almost as spine-chilling as the original. The cone of a flashlight pointed at Ponyville on the map. Soon it would edge deeper and deeper into the Everfree Forest.</p>
<p>“I saw them, and I know who you are,” Amber took a step toward her and looked her in the eyes, “Nightmare Moon!” After an hour of futile learning of Twilight Sparkle’s lines, Amber had decided to refresh herself on the gist of the play and then just improvise the rest.</p>
<p>From the corners of her eyes, she studied the audience. It was just early afternoon, and the room was not any darker than in the morning, but the stage was now so brightly lit that all the guests were almost hidden in the relative dimness beyond the stage. She could however see the gleams of hundreds of attentive eyes all the way to the end of the hall. The adrenaline rush was exhilarating and she enjoyed every second of it.</p>
<p>Despite the size of the hall, the ponies were packed so tightly that ventilation was becoming a problem. Most of them were friends or relatives of the ponies in the plays, and some had traveled with them from as far as Vanhoover, Las Pegasus, Baltimare, or Manehattan. The festive mood was only disturbed slightly during the intermissions when parents who had shoved to the front for their own child’s performance pushed all the way back out. Amber could not make out her family, but she was sure that they would respectfully stay for the whole event.</p>
<p>“<em>Elements of Harmony</em>, <em>Elements of Harmony</em>, …. How can I stop Nightmare Moon without them?” Amber was pretending to search through rows of encyclopedias. <em>Rainbow Dash, your cue!</em> She threw Noi a furtive glance.</p>
<p>“And just what are the Elements of Harmony? And how did you know about Nightmare Moon, huh? Are you a spy?” Noi rushed at her. There was not only mock anger in her eyes, but Amber ignored it.</p>
<p>Soon their performance neared its end. She had done it. Two cakes, one eaten, the other one still in front of her. All the formative experiences of her past eight years were still with her, her treasured memories of studying theoretical magic with her brother, her better understanding of the perspective of earth ponies, and her realization that it was just as full of possibilities. Now she would undo the effects of her betrayal of her friends.</p>
<p>The curtain fell, everypony grabbed some props, and they scrambled out through a backstage door into an adjacent hallway. Cheerilee congratulated the group on a performance well done, but as the group dispersed, a few remained behind and ringed Amber.</p>
<p>“What the buck was that about, Amber? Because you had to be creative with your lines, we almost missed our cues. The audience will think we’re the ones who messed up,” Diamond Tiara said.</p>
<p>“You got to be a bit flexible, silly filly, think on your hooves,” Amber countered.</p>
<p>“Well, when you didn’t say ‘Elements of Harmony’ the final time, I wanted to enter anyway, but I wasn’t sure. That’s why I hesitated,” Noi said.</p>
<p>“Okay, it won’t happen again. I promise.”</p>
<p>They did not seem entirely satisfied, put their tribunal was cut short when Cheerilee took Amber aside. When they were out of earshot, she put a hoof on Amber’s shoulder.</p>
<p>“You’re good at improv, Amber, but, as you know, the others relied on your lines as cues. They didn’t expect you to ad-lib them. And you knew them well-enough a week ago in Ponyville. Why the changes?” Amber saw that she was angry but allowed herself only to show her concern.</p>
<p>“I mentioned a contingency earlier that caused me to be late today. That was an understatement. I’ve forgotten a lot, including the lines, but I tried to relearn them as best as I could.”</p>
<p>“Oh,” Cheerilee looked her up and down as if trying to find any hint as to what Amber might be referring to. “What, uh, contingency could do that?”</p>
<p>“I would prefer to keep it private.”</p>
<p>“Okay, but if you have any problems you would like to talk about, you know you can always come to me for that.” She waited for Amber to nod her understanding before she proceeded. “As you know, the next performance is the day after tomorrow already, and that’s the final one. The princess will be there to see it.” She gulped. “Amber, I know how hard you have worked for this and how much you love the stage, but you can’t learn that script in one day. Little Cedar should take over your role.”</p>
<p><em>Oh no!</em> “That would not be a good idea,” Amber said, deciding, as she spoke that she would not snitch on Little Cedar, especially not after she had knowingly cast somepony who never wanted to be on stage.</p>
<p>“Many other ponies have made arrangements to share one of the performances with their understudies. It’s only fair.”</p>
<p><em>Maybe Cheerilee had looked through her plans all along. No, she would’ve intervened.</em> “She doesn’t <em>want</em> to be on stage. I will know my lines, I promise.” Was there a chance Little Cedar would leave before the final performance despite the changes to this timeline?</p>
<p>“No, Amber. I’ve made my decision. I’ll brief Little Cedar as soon as I find her.” She looked around. “She must’ve gone home already, but I’ll catch her tomorrow.”</p>
<h2 id="the-bloom"><a class="toclink" href="#the-bloom">The Bloom</a></h2>
<p>There was more rehearing on the next day. After Cheerilee had talked to Little Cedar, the filly suddenly had tremendously important errands to run or fly and vanished for the rest of the day. Amber filled the role again during the rehearsals and impressed Cheerilee when she bungled her lines less and less frequently.</p>
<p>Then the day of the finale was upon them. Amber left the hotel early, proud that she would be on time for once. She was just out the door when she heard crashing and cursing from their room and then her father’s terror-stricken shriek. She ran back upstairs.</p>
<p>Her dad sat propped against the wall and fanned himself fresh air with the hotel restaurant’s menu to keep from fainting. A few steps away from him lay her brother. Her mother was bent over one of his hind legs feeling his ankle. The sunlight from the window played on his black and white body. Damask only looked annoyed with his own clumsiness.</p>
<p>“Bubs! What happened?” Amber asked.</p>
<p>“It’s nothing. I just wanted to look into the upper drawers of the Chippendale escri—grr, argh—toire,” and Amber admired that even in his pain he never considered saying <em>desk</em> although there was precisely one in the room, “so I climbed it. Most drawers were empty, but the uppermost left drawer contained three goose feather quills and one swan feather for titling. The drawer below it contained a jar of lampblack ink.” Amber’s mother was carefully moving his hoof. “The last drawer I inspected—grr, argh—held a letter opener. Even while I jumped down, but still before I landed so clumsily and sprained my ankle, I noted that while the blade was of stainless steel, the quillons of the ornate cross guard had been fashioned from Bakelite, possibly because the material was more readily engraved.”</p>
<p>But Amber did not have to filter his explanation for the relevant bits. She knew exactly how it had happened. Her brother had told her the story many years ago when she asked about his limp. A sprained ankle was a serious injury for a pony.</p>
<p>What was worse was that she had known it would happen. She might not have been able to reconstruct the exact date from memory, also because she had been in coma the first time, but she knew that it had happened shortly after her own accident. It was probably something about the unusual lighting situation with the window drawing a sharp contrast between light and shadow across the desk that led him to misjudge the height.</p>
<p>It would have been easy for her to avert the accident. She could have moved the desk or she could have simply told her brother not to climb it. She could have figured out how to remove the drawers, so she could levitate them down for him to inspect and memorize. There was so much that she could have done to save her brother’s ankle if only she had thought to do so in advance.</p>
<p>Amber took a step back and slumped against the wall next to her father. Her mother said something along the lines of “Not you too,” but Amber hardly heard her.</p>
<p>Was there anything she could still do for her brother? Could she find Fleur to send her back in time again? No, the current Fleur would not have the first clue how to do that. Maybe Fleur will end her back again in eight years? Not without the accident that disturbed her precious purity.</p>
<p><em>It is too late for my brother, for his ankle anyway, but</em>, adrenaline erupted into a prickle throughout her chest, <em>but I remember so many other accidents that have yet to occur. So many other ponies, like my brother, who I can save! Who I must save, because I can.</em> The prickling sensation expanded to her thigh.</p>
<p>Her family was looking at her. Her dad was no longer covering his eyes, her mother had looked up from the wound she had just disinfected, only her brother had been looking at her throughout. <em>What?</em></p>
<p>Her cutie mark was a beautiful yellow rose in bloom. A short stout stem, vicious thorns, and countless petals of a strong, dark yellow that radiated in equal parts determination and joy.</p>
<h2 id="the-betrayal"><a class="toclink" href="#the-betrayal">The Betrayal</a></h2>
<p>“Unless your cutie mark gives you the ability to lay magic healing eggs, there’s nothing you can do for me now. Your classmates, however, they count on you.”</p>
<p>Amber would have stayed with her brother, but he would not allow it. Finally, she conceded that his logic was sound and left the hospital. On the way out she peeked into the room where she had lain for over three weeks while the hospital in Ponyville was rebuilt.</p>
<p>She trotted toward the castle. By this time Little Cedar may have already thrown in the towel—if she was still to do so on this timeline—and Cheerilee may be panicking with no cast for the most central role.</p>
<p>Amber would have liked to enjoy the euphoria that this new and so much more auspicious cutie mark bestowed, but her brother’s accident overshadowed the occasion even though he took it least seriously of them all.</p>
<p>On her way to the castle, new feelings started to gush into the already heterodox potpourri. It was obvious that her cutie mark was connected to her epiphany. She had an ability, her foreknowledge, that distinguished her from most ponies, made her special. Now she understood that it was also an obligation. It allowed her to wield great power, but now she also felt its weight of responsibility.</p>
<p>She should record her memories of accidents in some fashion, record everything that she remembered about them. Her memory was surely above average, but Damask would have been a much better choice for the job. She reminded herself that the universe did not choose, and Fleur had her own unrelated, purity-ridden motives. This was her job and hers alone.</p>
<p>When Amber arrived, lost in thought, her class had the stage and was in the middle of rehearsing the scene at the bridge. She was hours late. Cheerilee scowled at her.</p>
<p>“What’s it this time? Another ‘contingency’?”</p>
<p>Amber squatted next to her. “Yeah.”</p>
<p>Little Cedar was on stage. Light coat and red mane and tail had always made her Amber’s mirror image, only with wings instead of the horn. Her gift for aerial navigation, evident in her compass cutie mark, was another distinguishing feature. She knew her lines well enough, but then she only had three in this scene. Evidently she had not left for Cloudsdale as Amber had expected.</p>
<p>Amber disapproved of her own feelings toward Little Cedar. After the epiphany, the whole question of whether she got her role back, her scheming and conniving for the spot in the limelight, and the play itself seemed even more childish to her than before. Her feelings, however, did not reflect this reality. She still yearned to act in front of the princess and scores of royals. The only thing that held outright envy at bay was the fact that Amber herself had brought this situation upon herself so many years earlier, a situation Little Cedar hated as much as Amber did.</p>
<p>After a little while, Cheerilee broke the silence again. “Congratulations on your cutie mark,” she said in a strikingly even tone. “In retrospect, how was your life before you got it?”</p>
<p>Amber pondered that. “Pointless.”</p>
<p>“Preparatory. You know, your cutie mark is sort of like a black belt in life, or in your particular life that is. A first dan of sorts.”</p>
<p>“You’re reminding me that it’s only the start of whatever journey it signifies?”</p>
<p>“Of course.”</p>
<p>When the scene was over and her class yielded to the next, the students began to notice Amber’s new cutie mark too, but no pony was surprised to find an amber rose, so the inspections were short enough. At least the questions about what she had been doing when it appeared and what she thought it meant kept Amber’s mind off the performance and the role she yearned to play.</p>
<p>Noon came, and Cheerilee remained steadfast in her casting decision. If Amber did not get to play for the princess, she at least wanted to use the opportunity to talk to her.</p>
<p>The atmosphere was subtly different from that the previous afternoon. Only a limited number of ponies had been invited, most of them royalty and the parents or guardians of the actors, so the main aisle remained open and everypony stood in orderly rows. The front rows were reserved for the princess and her attendants, and chairs were brought in for them.</p>
<p>Twelve classes would present their performances and Ponyville’s was among the first. A few minutes before the first play was set to start, Princess Celestia slipped into the hall rather inconspicuously and sat down. Amber walked over to her and bowed.</p>
<p>“Your Highness, I hope our play will not offend. We have taken great care to represent the events that have taken place in Ponyville as accurately and respectfully as our humble skills allowed.”</p>
<p>“Then you must be one of the fillies from Ponyville. Your teacher, Cheerilee I believe, submitted the premise a few months ago for approval,” Celestia replied. “I’m sure it will be delightful.</p>
<p>“As it happens, I will make a casual visit to Ponyville tomorrow. I just hope they’re not making a great spectacle out of it again,” Celestia said and smiled. “Who do you play?”</p>
<p>“I’m Twilight Sparkle. In today’s performance, however, a friend of mine will fill the role.”</p>
<p>“I’m looking forward to seeing the real Twilight tomorrow. You know her, I trust, Ponyville being the small town it is?”</p>
<p>“I do. She has often lent me books.” To Amber, she was mostly the princess now, but she remembered how she had thought of Twilight at the time. “I don’t know her very well beyond that.”</p>
<p>The spotlights bathed the stage in light to indicate that the play was about to start. Amber excused herself and left for the hallway that doubled as backstage area. She somehow remembered this “casual visit” Celestia had mentioned, but she could not quite put her hoof to it.</p>
<p>Backstage she was being expected. Little Cedar stood to the right of two ponies oddly dressed for the season—dressed at all that is. All three were pegasi, and all three beamed at Amber, who responded with an askance scowl. They stood in the order of their shoulder height.</p>
<p>“You must be Amber,” the tallest pegasus said. After a nodded confirmation from Little Cedar she continued, “We are Little Cedar’s parents”—at which Amber mused that their names must be Big Cedar and Average-Sized Cedar—“and we just learned that she has been accepted into a prestigious aerial navigation program of her Cloudsdale school. She has been on the waiting list of three years. Now a student didn’t show up in time, and she gets to take their place. Unfortunately, we’ll have to leave for the Crystal Mountains right away to be there by dawn.” She looked at her partner to continue.</p>
<p>“Cheerilee tells us that you were eager to take the role anyway, so this should not be a problem for you. We apologize, however, for the short notice,” she said.</p>
<p>Now Amber smiled mildly. “Even if there were no second cast for Little Cedar’s role, you would not allow for this chance to pass her by, would you?”</p>
<p>“No. It <em>is</em> her true calling,” said the pegasus who had spoken first. Neither of the others contradicted her.</p>
<p>“What her cutie mark is telling her,” Amber said in a dreamy voice, then more firmly, “I’ll gladly accept.” <em>So that’s what happened last time. Mystery solved.</em></p>
<p>They did not even leave Little Cedar the time to say goodbye to the rest of the class. With a hurried thank you they rose and flew off, even still within the building.</p>
<p>Amber decided to go through her lines a few more times before the final performance.</p>
<div> </div>
<p>These royals were jolly fellows! Amber was surprised to find that some of them knew the lyrics of the little ditty Cheerilee had worked into the play and accompanied at least the slow passages. She thought she even heard some stallions with impressive falsettos.</p>
<p>After that scene, the students quickly plunged the stage into what darkness the slightly filtered daylight allowed. The cone of the flashlight on the map indicated a spot deep in the Everfree Forest. They removed several tripods with paper foliage that had represented trees and rolled a band of blue across the center of the stage toward the audience. A river.</p>
<p>Or in fact two rivers, because they were in the lucky position to be able to use the same setup for two scenes. The cone of light moved slightly to the south, where the first of the rivers cut through the green.</p>
<p>This scene was entirely Wind Turner’s. He was the little colt of the enigmatic Turner family, but his demeanor was everything but enigmatic. He was a born entertainer, outgoing and unself-conscious, and while he was too nearsighted to see the audience, he could hear their roaring laughter over his performance.</p>
<p>The circle of light moved southeast. The second river scene had Amber speaking again. “There it is, the ruin that holds the Elements of Harmony. We made it!” This time, she knew she got it right, verbatim. A moment later she added “We’re almost there. Whoa!” She had to pretend to almost fall into a narrow band of cloth. Nothing easier than that.</p>
<p>Noi had drawn the real Rainbow Dash’s cutie mark over her own. Just as her understudy, she was an earth pony, but the cutie mark made clear whom she represented. The coming minute would belong to her, and Amber had time to wonder about her last few lines. “The ruin that holds The Elements of Harmony.” For the second time that day, she was reminded of the ruined Ponyville she returned to when she woke up from coma. The amulets had probably been in a safe somewhere in Canterlot already, but the six ponies lived and Ponyville and helped rebuild it after… What exactly had happened? It was Amber’s turn again.</p>
<p>“Rainbow, what’s taking so long? Oh no. Rainbow! Don’t listen to them.” She charged her voice with so much urgency that she could hear sharply indrawn breaths throughout the audience.</p>
<p><em>Oh, right, the parasprites.</em> For a moment, she was glad. Another mystery solved.</p>
<p><em>Oh no! That started the day before Celestia’s “casual” visit!</em> Ponyville might already be under attack! They might have already infested a dozen homes, and a day later they will be gobbling up Sweet Apple Acers, Town Hall, and her parents’ place!</p>
<p>But she may be able to forestall the invasion right where it started with the single parasprite Fluttershy had found. That may have happened during the early afternoon of the day.</p>
<p>If it is already too late for that, she may still be able to mediate between Pinkie Pie and the other five who seemed to have had trouble communicating. She knew where the school kept the instruments. There was so much she could do to prevent the disaster! <em>It’s also what my cutie mark is telling me—my true calling.</em></p>
<p>“See? I’d never leave my friends hangin’,” Rainbow Dash told her. It was the end of the scene. <em>But what if you have to choose between your friends on this stage and your friends in Ponyville?</em> Amber remembered a newspaper article she had read about the event. <em>And the friends you have yet to meet in Fillydelphia!</em></p>
<p>One thing was for certain, just as Rainbow Dash just told off the Shadow Bolts out of loyalty to her friends, Amber would not forfeit any of her friends for her own ambitions either. The rest of the equation was simple enough too. A ruined play or two ruined towns? She felt awful about what she had to do.</p>
<p>The stage required some reorganization to represent the interior of the castle. The river had to go, and five boulders had to be arranged in a circle. One jump and Amber was off the stage, a second and she stood in front of Cheerilee.</p>
<p>“You know all my lines, right? You’re the new Twilight!” She did not wait for Cheerilee to reply. Her teacher would be able to tell from Amber’s voice that she was dead serious.</p>
<p>Seconds later, Amber slipped backstage. “Juniper, you got yourself an assistant, right?”</p>
<p>“I wanted to be involved in some fashion at least.” It was Featherweight who answered. He still sounded sick.</p>
<p>“Cool, you take over. Juniper, we’ll need two carpets. Direction: south-west to Ponyville. Time of departure: right now. I’ll explain when we’re airborne.”</p>
<p>Juniper looked shocked for a moment, then scrambled to find two carpets among the heaps of props.</p>
<p>Seconds later the play resumed and they levitated the rugs out to the backstage area. Amber caught a glimpse of Cheerilee on the stage. She could not see how the audience reacted to the unexplained transformation.</p>
<p>“Whatcha doin’?” A colt had followed them out.</p>
<p>“We have to leave for Ponyville, urgently,” Amber said. She tried to recall his name. He was the understudy of Noi and had played Rainbow Dash in their first public performance in Ponyville.</p>
<p>“On carpets, awesome! I saw a large lenticular cloud above the mountain this morning, probably due to moist winds from the eastern sea, so you need to be careful not to get into the leeward turbulences but you can also use the wind to travel faster for a while, then correct your path at the last moment.” He looked from one unicorn to the other. “Actually, can I come with you? I’m preternaturally bored, and flying sounds awesome. I’ve never done that before.” It seemed like a statement of the obvious to Amber until she reflected on her own wingless nature.</p>
<p>“You’ll go with Juniper. She’s better at this. Let’s go!” <em>Oh, right, </em>September<em> was his name!</em></p>
<p>Amber knew that she had not used a flying carpet in over eight years, but she had not expected to be so clumsy. Juniper’s carpet glowed with her golden aura, straightened, and lifted into the air with one swift and controlled motion, Amber’s lift-off equaled it only in swiftness. The carpet would have crashed her into the ceiling had she not released her magic at the last moment. Then carpet and pony were in free fall.</p>
<p>Juniper shot up and headed directly for the falling ball of Amber and carpet. Right before they made contact, she converted her upward momentum into a downward one roughly half of Amber’s and caught her softly. Then they were blind-sighted as Amber’s carpet enveloped them. Juniper remembered their height and speed and landed them blindly. The whole ordeal took only seconds.</p>
<p>September spit out Juniper’s tail and danced on her carpet. “Woohoo! That was fun!”</p>
<p>Amber jumped back on her carpet. “I’ll try again!”</p>
<p>This time she managed to lift the carpet more carefully. The way to the southern gate lead around a few more corners and through narrower hallways. Accordingly, Juniper chose a slow pace and Ambr matched it.</p>
<p>“Why do you have to hover a carpet?” September asked. “Can’t you just hover yourselves?”</p>
<p>Explaining the intricacies of carpet flight was Juniper’s forte. Her inofficial lessons often included a theoretical portion. “It is possible to hover yourself directly, but this sort of recursive application of magic, a levitation applied to your own whole body, is of a level of skill beyond the ken of most unicorns.</p>
<p>“Levitating a carpet beneath your hooves is greatly easier, although it is different from other levitation magic in that you can’t feel the progress with your magical proprioception—you know, the sense for the position of your own magical aura—because it always remains at a constant distance to you.</p>
<p>“I usually recommend that they imagine levitating the earth away from them. That’s not actually the case, but it helps them. Then some unicorns have problems not to get confused about up and down. It takes some practice. Maybe you can learn it too.”</p>
<p>“Earth ponies are not known for their flying skills,” September said.</p>
<p>“Neither are unicorns,” Juniper countered.</p>
<p>“There are a lot of jobs in the weather patrol that don’t involve cloud kicking. I’ve been hoping to join the planning teams when I’m older, but those are traditionally jobs for pegasi too.”</p>
<p>“Buck tradition. The average colt your age wouldn’t know what a lenticular cloud is, even a pegasus. It’s your brains they’ll need in the Weather Patrol, not your wings.”</p>
<p>Amber sensed that her young self must have seemed embarrassingly immature to Juniper. To distract herself, Amber started to experiment with her speed. She liked speed.</p>
<p>With one burst she doubled it, slanting the carpet forward for a second so not to be swept off. Then she whizzed around the two, three, four corners to the entrance. She knew that Juniper could easily catch up to her but prudently chose not to.</p>
<p>Amber was lucky that the gate was open. Braking her carpet down from this speed would have been possible yet highly unhealthy.</p>
<p>Outside, Juniper caught up to her at once. September stared in amazement at their height, their speed, but mostly the cloud formations that seemed to whisper to him.</p>
<p>He pointed. “You see the tapering of the lenticular cloud? That’s the leeward side. We have to stay away from that.” He thought for a moment while they soared over the roofs of Canterlot. “Head slightly north of Ponyville, this direction,” he pointed again, “so the wind can drive us a little or we can match its speed, so we can talk.”</p>
<p>Juniper followed his instructions and accelerated in her uniquely smooth way. Amber followed with one burst of acceleration that knocked her legs out from under her as it pressed her against the slanted carpet. They cleared Canterlot and the plateau, and were gliding above a mile of nothingness.</p>
<p>Amber was still fighting with nausea when Juniper swerved to her side. “You alright? You used to be better at this two weeks ago.” She was worried, not teasing. “Remember to increase your speed in small increments.”</p>
<p>“You didn’t speed up in increments,” September told Juniper.</p>
<p>“That’s true. I accelerated at a constant rate over several seconds, but I haven’t been able to teach that to anyone else yet. There are other carpet flyers who have mastered it, but none in Ponyville unfortunately. I’ve never met them.</p>
<p>“All unicorns learn to levitate things at some point, and moving something at a constant speed comes naturally to them too, but there’s something about acceleration that makes it more abstract to them. It’s one step further removed in a way. It’s hard for them to take that extra step in their imagination.</p>
<p>“What they do is that they tilt their carpet forward and at the same moment command it to a higher speed. It complies within a fraction of a second. If the increase was too great, though, the pony gets slammed into the carpet.”</p>
<p>September was satisfied with that explanation.</p>
<p>“Oh, and what are we doing here anyway?” Juniper added.</p>
<p>Amber had recovered and sat comfortably on her carpet. “We have a few minutes till we reach Ponyville, so I guess I’ll use them to tell you my eight-year-long tale.”</p>
<p>It was closer to an executive summary of her time travel, her faux epiphany that she was sent back to fix Fleur’s or her own perceived mistakes, and her real epiphany that it was up to her alone to choose her purpose in life, starting with using her eight years of foreknowledge to avert every catastrophe she could. She concluded with the bit about the parasprites that were about to gobble up their home town, when it came into view. A tiny splotch at the rim of the Everfree.</p>
<p>September paid attention to wind and clouds and birds, and had not spoken throughout Amber’s explanation. Now he looked at her. “Wouldn’t it be better to send the royal guard rather than trying this alone?”</p>
<p>Amber cocked her head. “I haven’t thought about that. A bit late now. Besides, I have a plan. Let’s just hope we aren’t too late already.”</p>
<p>Landing had always been a challenge for Amber because she had to decelerate in time, so she would neither be crushed by the bursts of deceleration nor by the ultimate deceleration that the ground affords. Now, however, as Amber experimented with her carpet, she noticed that she could in fact gain glimpses of this higher level of abstraction that was home to acceleration and deceleration. It was a feeling just like the one, months after her coma, when she suddenly gained the ability to funnel her remaining magical trickle in just the right way to produce whole sentences again with her quill. Her grasp on it was tenuous and would require much more practice, but then as now, her discovery made her giddy and proud. She had to show it to Juniper at once. The landing was an ideal opportunity.</p>
<p>First they only changed their trajectory southeast and down toward the town, but soon they would have to brake to sail down into one of the streets or squares at a comfortable speed. As they veered left, they were hit by the wind whose speed and direction they had matched, and Amber could feel why the Cedar parents had worn jackets. She shivered.</p>
<p>“Let’s land close to Sugarcube Corner,” Amber shouted over the noise of the wind in her ears.</p>
<p>When Amber could just make out the ponies in the streets, Juniper started to fall behind, slowly at first, then more rapidly. She had begun to gracefully slow down. Now it was Amber’s turn to show what she had learned.</p>
<p>She had to concentrate very consciously to regain the control over her levitational magic that she had grazed minutes before, concentrate on an activity that was completely unconscious to most unicorns. She was already countering a constant acceleration, that of the earth. Now she just had to replicate the same kind of force in a different direction while maintaining the first.</p>
<p>The weight on her legs told her that it was working, but she wanted visual confirmation. The ground was too far away for her to use as reference point, and she dared not look around lest she lose her concentration. <em>Come on, Juniper can move freely on her carpet too. Try it!</em></p>
<p>She turned her head just in time to see Juniper right behind her and closing. Juniper banked hard to her left, Amber reversed her thrust.</p>
<p>They missed each other by a hoof’s width, but Amber accelerated too hard. Her hooves lost friction, and she tumbled to the stern of her rug before she could react and slope it upward.</p>
<p>Her hind legs were already hanging over the abyss and she wished for claws for the first time in her life, when something forced her magical aura off the carpet as well. She was too preoccupied to fight back.</p>
<p>A moment later the bow of the carpet dove back underneath itself and caught Amber from below on its underside. Then she regained her magical hold of it.</p>
<p>Juniper hovered next to her. “I think you just accelerated at a more or less controlled rate. Congrats! Hardly anyone can do that.” She seemed completely collected. Only September’s firm bite to her tail betrayed the dramatic carpet acrobatics moments earlier.</p>
<p>“Did you just fly two carpets at once? How…?” Amber started at her.</p>
<p>Juniper just shrugged and nodded in the direction of her cutie mark.</p>
<p>“Let’s not waste any more time up here!” September had released his bite to speak.</p>
<p>Without another comment, both carpets dove toward Ponyville.</p>
<p>Amber used her new trick again to brake before they set down. This time she took care to keep Juniper and September in front of her and at a safe distance.</p>
<p>Ponyville had not been devoured yet.</p>
<p>Sugarcube Corner was open, and Mr. Cake welcomed them.</p>
<p>“Hey Carrot, is Pinkie Pie in?” Amber asked.</p>
<p>“One moment.” He vanished into their kitchen. Moments later Pinkie Pie bounced toward them.</p>
<p>“Pinkie Pie,” Amber began, “I don’t have enough time to explain the how and why of everything, but you have to trust me. We need your help.”</p>
<p>“Lemme guess! You want to buy prosthetic hands!”</p>
<p>“What, no.”</p>
<p>“Okay! Adorevil little creatures are about to overrun the town, and you need me to assemble my one-pony band and fly with those fillies to Fillydelphia to save the day!”</p>
<p>Amber was a little confused each time this happened. “Precisely. September here has some crazy meteorology knowledge and can guide you. Juniper is probably the best carpet flyer alive. You’ll be in good hooves.”</p>
<p>“Let’s …,” Pinkie Pie started.</p>
<p>“One more thing. I’ll need your party cannon.”</p>
<p>“It’s right here!” Although the cannon was about the same size as the whole pony, Pinkie Pie shook it out of her mane with one practiced motion. “Now let’s go!”</p>
<p>Amber turned to Juniper and September. “Pinkie Pie will have to collect some supplies before you leave—you have my permission as, uh, student to pillage the school’s stash of instruments—and Fillydelphia is to the east, so you’ll have to fight the wind all the way. The town may already be full with the creatures by the time you arrive. I hope that I can prevent that here.”</p>
<p>“Not necessarily,” September said, “I saw a few stratocumulus castellanus that hardly moved. We might escape the wind to a higher altitude, but we’ll need warm clothes.”</p>
<p>“Good. We’ll meet again right here. Take care of each other and good luck!” With that, Amber grabbed the party cannon in her magic and galloped toward the edge of the Everfree Forest.</p>
<div> </div>
<p>“Then suddenly this filly, Amber Rose, jumped out of the bushes where she had lain in hiding motionlessly for half an hour. ‘Fluttershy! Stand back!’ she commanded.</p>
<p>“I jumped back, I fluttered, I scrambled, I almost fell. Then she stood over me—actually, uhm, I think I’m the taller one but, uhm… She stood over me with fiery mane and fiery tail, her eyes burning with determination.</p>
<p>“That’s when the little creature, a voracious parasprite, tried to escape, but the filly caught it in the auburn grip of her magic and shoved it down the barrel of her cannon—uhm, can I maybe say ‘party cannon’? It sounds less, uhm, threatening?</p>
<p>“A second later she shot it far out into the forest, together with all the usual streamers and confetti, never to be seen again.</p>
<p>“Don’t you think it’s a little, uhm, overly dramatic?” Fluttershy asked and put down the slip of paper with the report Amber had asked her to recite for the Foal Free Press.</p>
<p>Amber bit her lip. “Actually, scrap that.” She took the slip out of Fluttershy’s hooves and hovered it into a waste bin. “Old Amber would’ve wanted to be celebrated in the papers, but that’s not what my cutie mark is telling me.” Amber saw that Fluttershy was inspecting the yellow rose, uncomprehending. “It’s telling me to help ponies because I can, nothing more. Just tell it as it happened, including our silly chase of that thing if you like.”</p>
<p>“Okay. I’m much more comfortable with that.” Fluttershy breathed a sigh of relief.</p>
<p>“But you can still mention my name, alright?”</p>
<div> </div>
<p>Amber had been waiting at Sugarcube Corner for a few hours and felt well nourished when she saw the carpet closing. First it was just a dot against the darkening sky, then she could make out Juniper, who stood at the bow, her horn aglow. She stood very rigidly, Amber noticed.</p>
<p>The carpet swooped into the street, landed softly in front of the bakery, and Juniper collapsed at once. She was huddled in a blanket in addition to her jacket, but the whole bundle shivered violently. Pinkie Pie wanted to get her to Nurse Redheart immediately.</p>
<p>“No, just a cold,” Juniper whispered. “Need to get to Canterlot. My parents. Amber, you need to fly us.”</p>
<p>Amber and September assured Pinkie Pie that they would seek medical attention right away when they arrived, and Pinkie volunteered to return the instruments to the school. Then they took off.</p>
<p>Juniper did not feel like speaking, so Amber had no idea whether she had a headache or nausea in addition to obvious chills and fever. Still, she took care to launch as softly as possible. She wanted to ask about Fillydelphia, the parasprites, what happened to Juniper, but first they had to get airborne.</p>
<p>The wind was pushing against them from starboard bow, and the further Amber accelerated, the harder it became for them to hear each other. They were cold too. Juniper was drifting in and out of sleep, hardly conscious of the maelstrom around her. Even September seemed a little drowsy.</p>
<p>“Try putting up a protective field,” he shouted at Amber. She barely understood him. “Juniper did it on the flight to Fillydelphia!”</p>
<p>Control the carpet, which moved so heavily with three ponies on it, and a magical field at the same time? <em>Juniper has done it!</em></p>
<p>It took Amber countless attempts. Always the spell threatened to disrupt her focus on the flight and she had to abort to secure her hold on the carpet. This time Juniper could not catch them if she messed up.</p>
<p>She kept the parameters of the flight as simple as possible. Constant speed, slightly slanted against the wind. Darkness had descended in full when, after many more attempts, she managed to erect a simple shield on just the windward side. It looked as if it might collapse any moment. Once it was in place it took little attention to maintain. For now they were safe from most of the wind and noise, and Amber could ask the questions she had been burning to ask.</p>
<p>“September, what happened in Fillydelphia? Did you get rid of all the parasprites? Why is Juniper sick?”</p>
<p>He put his hoof over the shaking bundle of cloth and pony. “Juniper put up a shield much like yours, but all around her carpet. That helped a lot when we rose through the surface current. We hardly noticed any of the storm around us. But then we reached a region of windshear and crazy turbulences. The shield kept the wind out, but from the inside we could’ve easily fallen through it when one of those gusts whirled us around or pushed us down.</p>
<p>“She parried them with lightning speed, and always slanted the carpet just right to keep us and the instruments from falling off. It was pretty rad actually. We were still trying to rise higher where I suspected a slower, more uniform current because of the clouds there. The downdrafts kept pushing us back.</p>
<p>“It took us at least half an hour to get through that layer. We’ve all been sweating blood and water, literally. Well, except the blood part. Especially Juniper was worn out from the concentration. She pulled down the field and let the gentle breeze up there cool us. But it was really cold actually, and then she even doffed her jacket. My mom always warned me against that.</p>
<p>“Anyway, near Fillydelphia the aerial situation was more peaceful. Near the ground it was anything but. The parasprites weren’t eating the houses yet, but the farmers and the vendors in the market were all up in arms. Many unicorns also struggled to hold up fields to keep them out, but other ponies were defenseless unless a unicorn volunteered to help them. Some were trying to shoo them off. It was chaotic.</p>
<p>“Juniper helped with some of the fields, but she was already feeling weak. Meanwhile Pinkie Pie put together her one-pony band, and the rest was routine, or for Pinkie Pie anyway. She lured them into a forest to the south of the city, and then the ancient magic that keeps all the beasts in there must’ve taken hold of them.</p>
<p>“Juniper had already stayed behind during that tour. When we came back, she lay on the carpet half asleep. Somepony had brought her some tea and a blanket.</p>
<p>“When she noticed us, she forced herself erect. With her last ounce of strength brought us up into the air and up to the right speed to sail on the wind. Luckily Ponyville was almost straight west from there, just the direction of the surface current.</p>
<p>“Somehow that last ounce lasted almost a full hour until we reached Ponyville.”</p>
<p>They were almost upon Canterlot. Amber corrected their altitude slightly upward—the city was that high on the mountain—and braked softly. Juniper appeared to be asleep and her shivering had ceased. Amber wanted to keep it that way.</p>
<p>Much slower this time that she had guests on board, Amber steered her carpet through the hallways of the castle. The gate at the back of their hall was fully open now, and ponies were removing chairs and decoration. Only a few of the students were still around, and their parents were chatting with Cheerilee as Amber, Juniper, and September approached.</p>
<p>“You!” Two of the parents were charging toward her when Amber had hardly disembarked. Then they saw Juniper. “June! There you are!”</p>
<p>“Shh, she’s asleep,” Amber whispered to them.</p>
<p>“What have you done to her?” one of them whispered back.</p>
<p>Amber ignored the accusation. “We think she caught a cold. She has been shivering quite badly for a while. Probably fever too. We tried to bring her back here as quick as possible.”</p>
<p>Juniper’s parents took her upon themselves—she only groaned softly in response—and scurried off without another word.</p>
<p>That’s when her own parents noticed the commotion. Her brother must have been still in the hospital, she surmised. It only took a moment until she was hugged by two wings and legs.</p>
<p>When all welcoming hugs had been administered, Amber turned to Cheerilee. “How did the play go after I left?” She could have affected contrition but she found that she would make the same decision again, so such an affectation would have struck her as insincere.</p>
<p>“I’m the Element of Magic now. Quite a promotion for an earth pony.” Then Cheerilee looked more seriously. “You’re not the type to get stage fright. What happened?”</p>
<p>A few of the parents who were still around edged closer.</p>
<p>“I don’t know how secretive I should be about this. Can we talk in private, just you, September, and my parents?”</p>
<p>They found a little cranny in an adjacent hallway where they could sit down. Amber repeated much of what had happened over the last few days, and told Cheerilee about their campaigns in Ponyville and Fillydelphia.</p>
<p>“I understand why you decided that the play was less important. It had been very important to you personally, so it must’ve been a hard decision too. Don’t you think, though, that their struggle with the parasprites and the rebuilding of the town may have also held positive, formative experiences for some ponies, that some ponies may have learned important lessons from them?”</p>
<p>Amber had not considered that but neither did it seem relevant to her. “I can’t allow a definite catastrophe to be visited upon countless ponies just for some hypothetical and probably not even commensurate good it might bring for some of them.”</p>
<p>Cheerilee only nodded. “But I’m afraid you won’t be able to convince your classmates through rational argument. Many of them seemed rather hostile earlier when they talked about you. You better think of a good cover story.”</p>
<p>“In a way I deserve it,” Amber said.</p>
<p>“You could not have known that Little Cedar would be taken out of school.”</p>
<p>“Not that, but I picked her because she had no interest in performing. It was awfully selfish of me, of past me.” Now Amber’s contrition was sincere.</p>
<p>“I know. Such is the way of the limelight.” She sighed. “But you learned your lesson.” Cheerilee had already turned away when she hesitated and turned back to Amber. “Actually, you learned several lessons, didn’t you? Remember when I said that your cutie mark was like a first dan in your life? Maybe you took another step in that direction today. You weighed your priorities there on the stage and decided to defend our town against the parasprites even though no pony would know to thank you for it. You may even become an anathema to your classmates for almost ruining their play.”</p>
<p>Now it was Amber’s turn to nod in silence.</p>
<p>After a moment of silence, Cheerilee turned to September. “You haven’t said anything. Aren’t your parents waiting for you?”</p>
<p>“Naw, I don’t have a dad anymore, and my mom runs a huge soap business in Manehattan. She only comes to Ponyville on weekends. Some weekends. I’m staying here alone too. I like the solitude though. Lots of time to watch the clouds.” Amber couldn’t tell whether he sincerely liked it; he always had a smile on his face, but at some times it seems more rigid than at others.</p>
<p>“I can walk you home, to the hotel, if you like,” Cheerilee said.</p>
<p>The other parents had gone during their conclave. Outside the castle, the Roses, and Cheerilee and September also wished each other sweet dreams and went their separate ways.</p>
<p>Amber was tired too, but sleep was not on her agenda for this day. When she had had a few hours to spare waiting at Sugarcube Corner, she had made a mental list of catastrophes, calamities, and similar doozies that she remembered had befallen her family and friends. As soon as they reached the hotel, she got to work writing them down.</p>
<p>At one point there had been that series of accidents in Ponyville that this mysterious Mare Do Well alicorn prevented before she vanished again. She could still remember her feats, but there was nothing for Amber to do there.</p>
<p>There was also the time Twilight Sparkle’s dragon got big all of a sudden. She might read up on dragon physiology to see if there’s anything that could be done to prevent that, but the literature on dragons was limited to begin with.</p>
<p>The lightning bolt that struck Town Hall? The roof structure had been awfully ramshackle. It had been a stroke of luck that it collapsed when no pony was inside rather than during a session of the senate. She dismissed that one.</p>
<p>She remembered the time when Twilight Sparkle was visited by her future self and got everyone worried about some impending disaster when all she had tried to tell herself was not to worry. She could warn Twilight, or unwarn her, as it were. <em>Naw, better not mess with time.</em></p>
<p>For a few more hours, Amber went farther and farther into the future, her past, and jotted down pages of notes. Her recollection was full of holes. Especially dates were hard for her to recall, often even the relative order of different events. Then she remembered the time when Discord wreaked havoc on Equestria. Their house had floated around for a while, so she could not get out.</p>
<p>When she tried to think of any way in that she could prevent or influence this event, it occurred to her that Discord had not just toyed with Ponyville but with much of Equestria. All the catastrophes on her list had imprinted themselves so strongly on her brain because they concerned herself or friends and family, all of them ponies from her home town, ponies she knew personally. This was the first event that broke out of these limited confines.</p>
<p>Amber knew that she had read about countless accidents and a few outright catastrophes in the newspapers throughout the years, but her memory of them was even more hazy. There was not a single such accident that she could recall in sufficient detail to even begin to think about ways to avert it.</p>
<p>It was so fundamentally unjust. The majority of the ponies in her country would have to suffer all their hardships again because their misery had not left enough of an imprint on Amber’s memory, because Amber’s empathy had been too selective.</p>
<p>Amber could not imagine what it would be like to dedicate her coming years to saving her friends from a few falling flower pots while on some far away ends of Equestria dam failures swept away a whole towns. She felt like she was condemning all these other ponies—almost all ponies throughout all of Equestria—as second-class equines just because they had not been lucky enough to be born in her home town.</p>
<p>It would be more just of her to sit back and let all the catastrophes strike again, Amber thought, but it was a justice that served no pony. She cursed her past self for not memorizing all the papers like her brother did. If only she had had some sort of written record on herself when Fleur sent her back. <em>No, I’m not an “if only” pony. I’ll just have to remember harder!</em></p>
<p>Remembering harder, however, she soon got too drowsy to keep her eyes open. Her parents’ soft snoring mingled into intermittent flashes of dreams she still tried to shake off. Soon she felt the top of the something-dale desk under her head. Dream ponies whispered to her. She thought she should move over to her bed.</p>
<h2 id="the-heists"><a class="toclink" href="#the-heists">The Heists</a></h2>
<p><em>I do have a written record!</em> It was early dawn. Amber had only slept for a few hours but she was wide awake. She was proud to have come to this realization while she was still in Canterlot.</p>
<p>There was another thing she would have to do in their final week of vacation there. A calamity that she could prevent almost whenever she wanted. A very convenient calamity. But for that she first needed more information, more than she could remember.</p>
<p>Once she had a more thorough record of the future, she could intervene in catastrophes from the San Palomino Desert to the Neighagra Falls, and it would also help her plug holes of half a year or more that gaped in her scribbled pages from the previous night. A pony with her calling could not possibly be forced to waste half a year just waiting. She burned to get to the castle.</p>
<p>She left a note for her parents and brother not to worry while she engaged in trespassing and theft. She elided the last part.</p>
<p>She emptied her backpacks on the floor, put them on, and galloped off to Juniper’s hotel.</p>
<p>A minute later and halfway there she braked hard, hooves raking through the dust. She stood in front of the hypnotherapist’s practice that she would pass eight years from then on her way to the university. She stepped to the door and knocked.</p>
<p>A middle-aged pony with calm, confident eyes opened for her, and soon they sat on a cushy couch sipping vegan mango lassis. If lassis were a time measure, then Amber did not want to stay for more than one of them, so she got straight to the point.</p>
<p>“I need to remember a lot of things that happened over the past eight years and I need to recall them in detail. Can you hypnotize me and then make me wander back there and recount them on tape?”</p>
<p>The hypnotist shook her head. “You’d be much more likely to make up fake memories that would be indistinguishable from real memories to you. You don’t want that.”</p>
<p>Amber pondered that for a moment. Her glass was not even half empty, so she had time. And then she had a new idea.</p>
<p>“There’s something else you might be able to do for me.”</p>
<p>Half an hour later Amber reached Juniper’s hotel and learned that Juniper’s parents were still blaming her for their daughter’s cold. They told her to scram, and Amber complied rapidly when she saw a potted petunia being levitated out their window. It crashed to the ground inches behind her flying tail.</p>
<p>She only knew September’s hotel, not the room number, but the colt was already up and was dreamily glancing out of his window when Amber approached.</p>
<p>The performances were past and the weather must have been very ordinary as well, because he was excited to learn of Amber’s new plans and eager to escape the boredom of the empty hotel room.</p>
<p>Armed with empty backpacks, both ponies trotted to the castle.</p>
<p>The backdoor to the executive wing that Fleur had taken to get to her office had hardly changed over the eight years no matter the temporal direction. Amber recognized it and with it recognized a crucial flaw in her plan. The door was locked just as it would be in eight years, and of course she had no keys. She pressed the handle and rattled the door, all to no avail. How had she imagined they would be able to rob a governmental building with nothing but backpacks and good intentions?</p>
<p>She stared at the lock as if she could scare the driver pins into the hull. She cursed under her breath. “Darn, we’ll have to find another way in.”</p>
<p>Just as she turned to leave, September tried the door a final time. It opened.</p>
<p>“How’d you do that?” Amber asked.</p>
<p>“Dunno.”</p>
<p>Up the staircase and then left. Or was it straight and the next one left? All the hallways looked the same, gray carpets, white walls, and doors to the many offices every ten or twenty steps. Amber did not want to waste time vacillating. She took a left turn. After a few score steps she realized her mistake and they turned. So straight after all.</p>
<p>The new route looked more promising. A few ponies passed them by, but they seemed to be on important errants and hardly took notice of the two young ponies. Soon they came upon one of the hallways that was patrolled by a guard. Amber thought that maybe the guard would let them pass if only they exuded the right kind of routine confidence. She was not sure if September could pull off an act like that, and due to their age, they would probably seem out of place no matter what they did, so she decided to wait in a smaller, arching corridor until the guard was out of sight. <em>Didn’t Fleur have to unlock another door on our way here? It must’ve stood open or we would’ve noticed it.</em></p>
<p>The guard would not go out of sight. He walked a few hundred steps to the far side, then returned. When he threatened to pass the door to their corridor, they retreated back a few steps until they were hidden by the curve. His hoofsteps passed the door.</p>
<p>A few seconds later Amber thought she heard voices, many of them. The guard still had his back turned on the the mouth of their corridor, so they peeked out to see where they were coming from. A pony with a loud, clear voice was walking ahead of maybe a dozen others, explaining various aspects of the castle and of Canterlot. The visitors, tourists probably, were either listening or chatting among themselves.</p>
<p>When they passed the door, Amber felt a tug on her backpacks. “Come, this is our chance,” September whispered to her.</p>
<p>A minute later they still walked among the group, pretending to listen attentively. The guide even flashed them a smile when she saw them. This part of the castle must be so boring that she had to fill the time with general factoids about the architecture and later the intellectual heritage of the city. Another minute later, Amber indicated to September that it was time for them to excuse themselves again. They were almost there.</p>
<p>Amber definitely saw one of the previously locked doors standing open now. They were lucky indeed. The office itself was of course as unlocked as she had left it a few days earlier.</p>
<p>They swiftly scooped up all the detritus of partially singed newspapers and deposited it in their backpacks. Amber observed with dismay that they bulged conspicuously.</p>
<p>Now they only had to get out. It should be possible to avoid the patrol, Amber thought, since there must be many ways out. She led the way along the hallway. Looking through the windows in the doors she saw that only a hoofful of the offices seemed to be in use, just as she remembered it from eight years from then.</p>
<p>They passed a hallway to the right, and Amber recognized it as the one she had mistaken a few days earlier for the hallway through which they had just come. On that tour she had not met a soul in the whole wing until she came upon the tower where Praetoria found her. It had been earlier in the morning then, but maybe it was always as empty. She yearned for the safety of the public streets outside. If only she had been able to learn teleportation. <em>Buck “if only,”</em> she reminded herself.</p>
<p>In her reverie she walked out into an intersection of two hallways, and as if to punish her doubly for her carelessness, not one but two guards stood only a few doors away. For a moment she forgot to act suave, stopped dead in her tracks, then stumbled forward when September bump into her.</p>
<p>“Hmm. Are you lost, madam?” the guard still wearing her armor asked. “Sir?” she added when she saw September.</p>
<p>“Depends,” Amber replied. “Is this the correct direction to the archivist’s office?” She pointed at random, straight as it happened. <em>There got to be archivists here, right?</em></p>
<p>“Hmpf. What archivist?” the guard asked. Her intonation was so flat that Amber could not tell whether she wanted to know which of several archivists Amber meant or whether she did not know of any.</p>
<p>“The one of …” Amber searched her memory for anything she knew about the place that did not sound utterly inane. “… of the Hypogeum Absciticious.” <em>Thanks, bubs!</em></p>
<p>“Hmm. Oh, you mean the curator of the gallery?”</p>
<p>“Or <em>curator</em>, yeah.” She intoned it as if it were an inconsequential distinction.</p>
<p>“Hmm,” the guard started, and Amber thought how expressive that sound could be if only she put any sort of intonation into her voice. “Hmm. I think that’s this way, don’t you think?” she asked her colleague.</p>
<p>“Whew, Hypogeum Ab–, uh… Lower certainly. There’s a staircase this way.” He tried to be helpful. “But weren’t you part of that tourists group a few minutes ago?”</p>
<p><em>Oops! Why do all those guards have to look so darn similar in those helmets!</em> Amber started to sweat. She hated to lie, but this seemed like a situation that demanded it. She tried to think of a believable cover story.</p>
<p>September was quicker. “Nah, we were just interested in what the guide had to say. We’re on an errant for the c-archivist.”</p>
<p>That was a fib that Amber could build upon. “Yeah, she mentioned that the prevalent architectural style that was used for most of the governmental buildings in the city is a reproduction of the style of a much earlier period, long before the reign of the royal sisters. They discovered the ripped, ogival vaults and arches, the flying buttresses, and the ornate stained-glass windows during archeological explorations of their own, and much later decided to fashion their palace in the same spirit just as they reused many names from earlier periods as well. We found that fascinating and followed her along for a little while.”</p>
<p>After that, the guards were silent long enough for Amber and September to pass them. But they did not get far.</p>
<p>“But your packs were empty earlier. Can we have a look what you’re transporting there for the archivist?”</p>
<p><em>So he can’t recall a simple word like </em>abscititious<em> but he does remember the state of our backpacks. Darn again.</em> “Only some scraps of old, scorched newspapers, see.” Amber levitated one side open. She hoped they would not look any closer and discover the dates years in the future.</p>
<p>The guard on duty levitated one of the scraps to her eyes. “But this is years in the future. Where did you get this?”</p>
<p>“Uh, they’re just …” Amber looked at September for help, who looked back wide-eyed and helpless. “Run,” she whispered.</p>
<p>Amber grabbed the scrap from the guards magic before she could react and both of them raced along the winding corridor closely followed by both guards.</p>
<p>“Alert the guards at the entrance to the Hypogeum! I’ll follow them!” The armored guard shouted.</p>
<p><em>So let’s not run that way.</em> Amber took a sharp right turn at random, hoping it would not lead them into any sort of cul-de-sac. <em>Outward, upward, and find a carpet!</em></p>
<p>Amber had already noticed that doors favored them that day, but what happened then was hard to attribute to mere chance. The moment September’s tail passed one of the door that somehow always stood open for them, it slammed shut in front of the guard. Amber only heard the double <em>umpf</em>. Not a second later a light blue aura played around hinges and frame of another door to their left, then ripped it free and slammed it outward in a dust cloud of mortar. Through the hole, Amber recognized the tiles and ornaments of a tower. “This way!”</p>
<p>The door had come to a crashing halt at the balustrade of a balcony. Amber could not tell if it was the same one where had met Praetoria. Most of them probably looked the same.</p>
<p>“I’ve never tried this before,” Amber called to September over the noise of their hooves on the tiled floor, “but this door already proved it could fly!”</p>
<p>A second later she just threw a short glance behind her to ascertain that September was on board, then accelerated under several times their own weight outward and upward. They were pressed against the unyielding material of the door. Amber understood why Juniper preferred soft carpets. For a moment she thought she saw Celestia’s mane wave from a window in the tower, then the city vanished behind the curve of the mountain.</p>
<div> </div>
<p>She had recognized her magic as she knew she would. She was also the one who played Twilight in the theater play. Now she somehow thought it necessary to steal something from an empty office in the castle. <em>I should’ve asked her name.</em></p>
<p>The door opened. “Your Highness?”</p>
<p>“Yes?”</p>
<p>“It’s probably not important, but a filly and a colt just escaped us on a door, in a way. They seemed to be able to make it levitate underneath them. Anyway, it looks like they stole burned scrapes of newspapers that are dated in the future.”</p>
<p>“Yes, I saw them fly past. Interesting.” She could not suppress the flicker of a mischievous smile at her elisions. She had felt no ill intent from these ponies, but what <em>was</em> their intent? “Please find them. I would like to know what they’re up to.”</p>
<div> </div>
<p>It was like a sojourn in the land of Cockaigne for Amber. The scraps they had harvested contained reports on countless little and larger calamities, much more than she could possibly have remembered anymore. All she had to do was to make sense of the sometimes partial and partly burned articles and organize them in an efficient fashion.</p>
<p>Her brother helped her, even though he had been released from the hospital mere hours earlier, and September had chosen to stay at the Roses’ hotel for a while.</p>
<p>After they had rounded the mountain, they landed the door in the outskirts on the opposite side of Canterlot, then furtively made their way back to the hotel. There had been no pegasus guards following them, none of the guards had asked their names, and with their bulging backpacks they might just pass for really studious little ponies who were not at school for some reason. Still they decided to strip the backpacks as soon as possible and then lay low for a while.</p>
<p>Money was not an issue for September. He pondered talking a room in the hotel and paying somepony to move his belonging over, but Amber assured him that it would not be necessary.</p>
<p>They were making good progress too. Damask read all the articles and fragments of articles, marking each that seemed relevant however tangentially. September and Amber then categorized them, building little heaps of all articles that touched on the same event, which Amber sorted by how many ponies and other sentient beings would get hurt if she did not intervene. If none of the articles provided hard numbers, she would make her own estimates.</p>
<p>All the while, they also protocolled these data for each heap on a notecard and agreed on titles for each card. The nature of the incident, as well as the date, location, and their best estimate for the harm it caused were crucial. In some cases it so happened that the scraps were so large or an article so encompassing that it was relevant for more than one event. In those cases it was helpful to quickly assess whether it actually added new information about one of the events. Amber would compare it with the collected data on each of the relevant cards and assign it to the heap where she felt it provided the greatest new insights. That way, she had to add cross-reference on the cards in only a hoofful of cases.</p>
<p>Early in the process already, Amber noted with some surprise and embarrassment that during her own brainstorming she had overlooked one particularly devastating and lasting catastrophe that she should have been perfectly aware of. The sorts of incidents that she had scoured her memory for were sudden, like plagues, fires, and breaking dams. They were also all within Equestria. All the while she had failed to recall the thousands of griffons who had to sell their beautiful eyries to be able to afford some of the last overpriced food for their families. They still had five years to build reserves for the drought period—if Amber could warn them. Again she was glad she was still in Canterlot.</p>
<p>They worked until late into the night while Amber’s parents supplied them with food and drink.</p>
<p>“Well, Amber, it seems your dad has assumed control of the hotel kitchen. Don’t expect to see him again tonight,” her mother said and set down a tray of poêlée with fennel, Brokaw avocado pudding, toasted cashews, pea tendrils, and sweet pepper vinaigrette. Her dad took food seriously.</p>
<p>“He’s so cool. He’s been nerding out with the chef all day!” September said. He had not yet learned that it was inevitable whenever they visited a restaurant or checked into a hotel that offered quality cuisine.</p>
<p>It should turn out that Amber’s mother was right once again. Amber woke up first again and found her dad sleeping on the couch, his bed being occupied by September. She carefully draped her duvet over him since he only seemed to have found a thin quilt for himself. She did not have to wake anypony for what she had planned for this new day.</p>
<p>Thanks to several newspaper clippings, she finally had enough data to tackle the calamity that she had concluded was one of the most convenient for her to avert. It was still almost a year till it would torment—or annoy—her home town, but with a little luck she could forestall it right away.</p>
<p>She went through the newspaper clippings again to ascertain that the notecard was not missing any important bits. “I Thought It Was Just a Unicharm!” titled a national daily paper. It was a direct quote as the subtitle revealed: “An interview with the merchant who sold Ponyville for a satchel of gold.” The text did not contain any pertinent information, but on one of the many embedded photos, Damask had recognized an alley in Canterlot that they had passed on their way to the hotel.</p>
<p>One borough of the city lay in a wide vertical cleft that ran all the way from the top of the mountain to the plateau that formed the foundation of the city. As such, it only saw direct sunlight during a short period of the day—unless clouds occluded the sky. Since it also lay on the windward side of the mountain, moist air from the ocean was constantly pressed upward along its sheer face, soon condensated in the thinner, cooler air, and thus nurtured thick clouds, fogs, and almost daily rainfalls. Amber would wrap herself in her cape and fill all her bits into the pocket. She shuddered at the thought of trotting there alone.</p>
<p>Then she trotted there alone.</p>
<p>As she rounded the mountain, the early dawn was first blocked by the granite of the cliff wall, then shrouded in fog. A few of the stores were already open. The deeper Amber went into the borough, the more frequent became the stores and pubs that had not yet closed. She saw a few ponies reeling and staggering homeward from the bars, or so she guessed, but they were too far gone to reciprocate the seeing. She felt as if her dark, hooded cloak camouflaged her against the tenebrous backdrop. Then the rain started.</p>
<p>Not half an hour later she found the shop. It was dark but the door was unlocked. A bell chimed when she opened it, but no pony appeared to welcome or intercept her.</p>
<p>The rain had soaked through the seams of her cloak and more water had splattered up her legs. She shivered, yet she would have to leave the door open to have any light inside and to keep the bell from chiming again.</p>
<p>The interior was filled with carelessly piled up assortments of lamps, books, and useless séance paraphernalia. Amber tiphoofed the narrow path along the center of the room. <em>Meh.</em> <em>I’m never going to find it here.</em></p>
<p>The narrow path widened and transitioned into a wider area littered with smaller heaps much like an estuary. Amber reached the bay, a mostly uncluttered semicircle around a long counter. The items behind the counter looked more valuable and were arranged with greater care. High up, the metallic gleam of the Alicorn Amulet caught her eye.</p>
<p>The light came on. Before Amber could bolt for the door, a calm voice spoke from the shadows.</p>
<p>“May I help you, young traveller?” the voice asked the young traveler. A bespectacled pony stepped from the shadows. Amber could not place the accent. Was he from Mane? Or from Rode Island? She had never been there. Or was it a posh Canterlot affectation?</p>
<p>Her heart was still beating double time, but she forced her voice to equal calmness. “I see you have a unicharm in stock.”</p>
<p>“Oh, that is no unicharm, young lady. That is the legend’ry Alicorn Amulet. It is one of the most mysterious and powerful of all the known magical charms.”</p>
<p>“Pff! Pull the other one! I know a unicharm when I see one,” Amber said. As if to accentuate her words, a gust of wind slammed the door shut.</p>
<p>“I see you are versed in the dark arts far beyond your years. Still, few have the skill to forge a unicharm anymore, so each is worth its weight in gold.”</p>
<p>“What spell does it enhance?” Amber was bluffing. She barely knew what a unicharm was.</p>
<p>“That is for the owner to find out.”</p>
<p>“So you’re saying you lost the manual? There are millions of magical spells. No pony can try them all out to find the one spell it would boost. Without the manual the thing is scrap metal.”</p>
<p>“Oh no, the ruby alone is worth more than the rest of the rubble in here!”</p>
<p>Good, at least she had knocked him down to the raw material value. She had one hundred fifty-five bits on her. “20 bits.”</p>
<p>“Hah! One kilobit and not a bit less!”</p>
<p>“I don’t have a thousand bits. I have thirty bits on me. Take it or leave it.”</p>
<p>“Nine-fifty and I’ll gift-wrap it!”</p>
<p>“Forget it,” Amber said. “Next item on the list: books on amniomorphic spells. Do you have any?” That would buy her time until she knew what to do next.</p>
<p>“Sure, the whole shelf over there.” He pointed. “Some are almost undamaged. Thirty bits if you take all of them.”</p>
<p>Amber climbed over a moraine of rubble until she could read some of the titles, grabbed one at random, and levitated it to her. She leafed through it, pretended to read. The shop owner waited and watched. She put it back and selected a second one. He was still watching. She actually read a chapter of this one, the only one that had not molded away. Just when she was about to put it back, he seemed to get bored of staring at nothing. Amber saw from the corners of her eyes how he turned and vanished into another room. Careful not to trigger a loud rubble avalanche, she climbed down.</p>
<p>The Alicorn Amulet was inside a glass cloche. She hovered it toward her, opened it, and deposited the amulet in the inner pocket of her cloak. She hesitated, the cloche still hovering. Finally she put the one hundred bits from Fleur where the amulet had been and lowered the cloche.</p>
<p>Her next destination was the castle. She would prepare an anonymous letter for Celestia and Luna asking them to lock the relic into a safe forever. She would seal it in an envelope to be opened by the princesses alone and deposit it right at their tower so it would not be intercepted. Hopefully, no pony would recognize her. She snuck to the door.</p>
<p>“Young traveller!” The merchant stood behind her halfway between the door and the cloche. He made no attempt to intercept her, which scared Amber more than a running tackle would have. “You are not from around here, are you, silly filly.” It was not a question.</p>
<p>Amber did not ask what he was insinuating. She jumped up the steps to the door. At the same moment she passed through into the open, a shrill whistling set the house and the very earth vibrating. Amber’s legs gave way for a moment and she fell into the mud.</p>
<p>She jumped up and threw a last glance at the merchant. It was a simple whistle that had caused the sound; surely it was magically enhanced. He stowed it back into his work coat. Still he made no attempt to pursue her.</p>
<p>Amber ran.</p>
<h2 id="the-possession"><a class="toclink" href="#the-possession">The Possession</a></h2>
<p>The rain splashed in her face and the mud spluttered up from the soaked ground wherever she stepped.</p>
<p>Seconds later she noticed the effect of the whistle. Half the borough must have heard the signal, and all the shop owner, bartenders, and their bouncers stepped before their doors. It was too late for Amber to switch to her lost filly act. Everypony’s eyes were on her. They knew her for the whistlee.</p>
<p>They were in no hurry. They blocked the street ahead of her and closed in behind her. The closest street corners were cut off behind walls of a dozen ponies in both directions.</p>
<p>To her left she saw a narrow gap between two houses. It was just wide enough for a pony to pass through but more likely intended to keep fires from spreading.</p>
<p>She jumped, spun ninety degrees in the air, braked hard with her hindhooves, and dove into the cleft before the owner of the house could close in on her.</p>
<p>She had to jump over the trash ponies had thrown into the dead interstice. Every jump had her scraping along one of the walls if she did not keep precisely in the center. The raw bricks and mortar ripped her cloak and grated her flanks.</p>
<p>No pony had thought to conveniently discard a carpet in the gap. Amber cursed herself for not having prepared her escape while she was in the shop. <em>C’mon, think of something!</em></p>
<p>Then she reached the other side. A last jump and she was out in the open. In midair a muscular foreleg caught her around her neck, blocked her momentum like a tree trunk, and slammed her to the ground. A coal gray stallion of Big Mac’s proportion loomed over her. A dozen more joined him.</p>
<p>The crash to the ground had paralysed her and sent her vision spinning. The huge stallion was thorough though and also pinned her down with one hoof. She could hardly breathe.</p>
<p>“Caught it!” he shouted.</p>
<p>“Thank you, good sir,” came the voice of the merchant from far away.</p>
<p>Amber labored to turn her head in the mud. Vertigo and the rain falling in her eyes made it hard for her to see him.</p>
<p>“First I need to take something back that is mine,” the merchant added, much closer now, “then she’s yours.”</p>
<p>Amber was not eager to find out what that meant. She had never managed to teleport, and with the pain blurring her vision and derailing her thoughts she might not be able to sustain a simple levitation. If she attempted to use magic to fight back, half a dozen unicorns around her would block her attempts. She had only one option.</p>
<p>Would the amulet force her out of her body and take over the reins? Or would she be able to control it? Had Trixie just been easily corruptible? Would whatever happens happen instantaneously or gradually? Would she first have to learn to use the amulet before she could do anything with it? Amber decided it was time for certainty.</p>
<p>With one swift magical motion she levitated the amulet from her inner pocket, along under her cloak, and toward her neck. She almost missed for dizziness, but when it touched her neck it righted itself.</p>
<p>The healing spell worked instantaneously. Her pain was gone. Her vision was clear.</p>
<p>Still covered in mud, pinned under a mighty hoof, and half strangled in a ripped cloak, she enjoyed for a moment that these dozens of ponies thought themselves her captors when in reality they were all at her mercy now. She noticed without judgment that her eyes narrowed and that a mischievous smile reached her lips.</p>
<p>The merchant stood over her now. “Please remove her cloak,” he told the bulky stallion.</p>
<p>“Bored now,” Amber said instead. Slowly she let fire emanate from beneath her like a lake of burning lava that flowed outward frictionlessly. She observed with delight that she had tinted it in yellow and red hues, just like her coat.</p>
<p>The stallion jumped away and stomped his hooves to extinguish the flames around them. The merchant took a step back as well. The rain evaporated when it hit the ground, so Amber was soon shrouded in a thin cloud of vapor.</p>
<p>Applying magic recursively, she levitated herself, turned herself upright, and set her hooves down in the center of the flames. They lapped up to her knees but did not harm her. Her erstwhile pursuers retreated several more steps.</p>
<p>Amber observed that she did not have to learn to use the amulet. It just did exactly what she wanted it to, even before she knew that she wanted it.</p>
<p>From one second to the next, the sea of fire sent out filaments that split before and met behind all the ponies that ringed her. Then they widened and coalesced until each pony was standing on just a tiny islet in an expanding sea of flames. Amber had to giggle at the beautiful symmetry. First they had caught her; now she had caught them. <em>Tit for tat!</em></p>
<p>Four pegasi leaped into the air at once, but not two beats of their wings later, they were slammed back to the ground by Amber’s magic. She was amazed at her own swiftness. Her magic was precise yet fast as a reflex, and it took no conscious thought.</p>
<p>A few unicorns tried their tricks on her, but she never even learned what they had in mind. The moment the faintest aura shone around their horns, she had already stifled it in her magic grasp. Again she only truly noticed what she was doing when her magic already engulfed their horns. It was automatic, just like the times when she galloped down some steep stony slope and her hooves seemed to find the stable stones and plain patches before she even saw them.</p>
<p>Or did the amulet take hold of her in exchange for the illusion of control? She could pose the question but it was hard to think about it. Then she saw the flames again. Light yellow and dark red, almost unnaturally dark. Those had been her favorite colors for as long as she could remember. This had to have been her choice.</p>
<p>They had been captured for about as long as she had been captured under the iron hoof. Her business here was done.</p>
<div> </div>
<p>Teleportation was based on a pony’s ability to not only convey magical energies into the temporal world but also to convey elements of the temporal world, themselves, into the realm of pure magic. This eversion of the boundary between the worlds was what made teleportation fundamentally different from other sorts of magic and inaccessible to most unicorns.</p>
<p>It was also dangerous. If ponies did not protect themselves with sufficiently strong magical shields while immersed in the seas of energy, their hair would be singed and scorched upon ejection.</p>
<p>Amber paid no heed to the intricacies of the technique until after it was all over. With one poof of red, she stood in her hotel room. Then she gaped. Naturally, she had never experienced anything like it.</p>
<p>“Oh, there you are,” her mother greeted her. “When did you learn to teleport?”</p>
<p>“It comes with age,” she replied without hesitation.</p>
<p>“And what’s with the necklace?”</p>
<p>Amber materialized a scarf around it. “Just a trinket. Where is the rest of the family?”</p>
<p>“Your dad’s back in the hotel kitchen, and your brother is getting a blowtorch for him for a crème brûlée. He should be back any moment.”</p>
<p>Then it struck Amber. Her brother was the enemy. He, and only he, had the power to divest her of the amulet although he did not know it. He must never find her!</p>
<p>“Actually, I’m just on a quick sojourn here to pick up some of my cards.” Amber tried to quickly scan the notecards they had prepared for future events. “Or I’ll just take all of them.”</p>
<p>That moment the door opened. “I got the blowtorch, but do you know where I can get a gas tank?”</p>
<p>With one magical swipe, Amber snatched and bundled up all the cards even while she already jumped out of the window. Once in the air, she levitated herself, her aura glowing vivid red, and shot up far above the city almost the height of the mountain summit.</p>
<p>The nothingness beneath her hooves made her queasy, so she materialized a carpet for herself to sit down on. Then she went through the cards.</p>
<p><em>Eight million starving griffons. Eight million.</em> She could not grasp that huge number but the numeral with its googly eyes looked kind of cute for a moment. She turned the card ninety degrees to see it better. Then she got bored and tossed it out into the wind.</p>
<p><em>Bat pony filly discriminated against during school baseball match.</em> She wanted to kick some juror’s tail! But then a mental image of a bat pony stared her into her mind’s eye. Those narrow pupils! Just like Nightmare Moon’s! She tossed the card out in panic.</p>
<p><em>Dam failure.</em> That one was similar to the one Mare Do Well repaired, but she had the advantage of foreknowledge and could prevent it before any water could break free. This one, however, was far away from Ponyville. She did not even know where that town was. The name sounded completely unfamiliar. She tossed the card away.</p>
<p>Ten minutes later she had reduced the stack to a few cards. “Perfect!” she exclaimed.</p>
<p>The famous River Eddy’s little dog would go missing during the Equestria Games in Manehattan only a few days from then. She remembered the adorable Yorkshire terrier from billboards and she had also seen a photo in the newspaper the day before. She even knew his name. New Amber could not believe that Old Amber had completely dismissed this card. <em>Tsk-tsk, Old Amber.</em></p>
<p>Manehattan was a bit too far for teleportation but the flight was quick enough. A strong shield protected her from cold winds and low pressure as she overflew Foal Mountain. Hollow Shades and its northeastern peak lay to her left and she could already see the skyline of Manehattan dimly through the atmosphere. A few more minutes and the suburbs passed by underneath her. She could even see the great tanks of September’s mother’s company where the liquid soap was stored.</p>
<p>She would never be allowed to work there. Liquid soap production was a job almost completely reserved to earth ponies, bat ponies, and pegasi, because so close to such great amounts of liquid soap, even a small magical sparkling, be it only of a unicorn that sneezed a little clumsily, would be amplified by the liquid crystals and sent out like a beacon over half of Equestria. Dragons—as well as a certain Ponyville fashion designer—had a fine sense for the scent of gemstones in the magical field, and when they arrived to find only soap, they sometimes burned the factories to the ground in their wrath.</p>
<p>Once she arrived in the city center, she materialized a few bits and rented a hotel room. Her family would be worried and start a search for her. She felt guilty about that. But then she remembered the adorable critter and his owner all distraught. <em>Priorities, Amber!</em></p>
<div> </div>
<p>“Your Highness, the ponies you requested to see.”</p>
<p>“Thank you,” Celestia said, and when the guard was about to leave she added, “You may stay, Praetoria.” Wide eyes looked at her. <em>Hehe, it always catches them off guard, as it were, when I know their names.</em></p>
<p>She looked at her guests. A pegasus with scarlet mane; a thin, slightly pinched-looking earth pony; a unicorn colt with the blackest coat she had ever seen but white, almost luminescent mane and tail; and another earth pony with pink coat and brown mane, who, she could tell from their faces, must be September, the colt who stayed with the Rose family. He seemed scared.</p>
<p>“Not to worry,” Celestia said to the group. “I’ve asked you here because I need your help in finding your daughter and friend, Amber Rose.” She gestured for the ponies to sit. “A few hours ago, the Berry parents wanted to file a restraining order against your daughter.” She allowed some time for the parents to gasp. “They claimed that your daughter infected their daughter with a nasty flu. These are already unusual grounds for a restraining order, but moreover their daughter, Juniper Berry, was vehement that the illness she suffered was due to her own negligence. Further evidence in favor of her version is that Amber was—and continued to be—perfectly healthy when Juniper fell ill. Her parents’ version was only supported by their assertion that ‘Somepony had to be behind it.’</p>
<p>“Needless to say, the case was complicated enough that soon a small legion of my staff, including several guards, became involved as they deliberated the proper course of action. It was then, that some of them noticed that the family’s description of Amber Rose fit exactly that of a filly I’ve been observing on and off over the past few days.</p>
<p>“Only two days ago, for example, staff members alerted me to a strange sight, a pony who fit her description hovering over the city on a carpet, so high that only with a telescope I should see her clearly. Unfortunately she soon flew away.</p>
<p>“But I understand that that must’ve been around the time when you last saw her. You started the search a day later. According to the records, you saw her last when she leaped out of a second-floor window and then flew upward unaided. You also stated that she wore some sort of necklace. Do you remember what the necklace looked like?”</p>
<p>Three ponies looked at the son of the Rose family, Damask Rose. He shook his head. “She was wearing a scarf when I came in. When I saw her, she was already outside the window.”</p>
<p>“Did you see anything peculiar about the color of her magic? She must’ve had a magical aura around her when she levitated herself.”</p>
<p>“Allow me to go back for a minute,” he said. This confused the princess at first but then she saw that he drifted off into a state of reverie. First his unfocused eyes shone and a smile played around his lips. It was just of the sort of joyous expression she had often seen ponies wear who just obtained a working blowtorch at an economical price. A moment later she saw a colt’s recognition of his mother mirrored there, then an expression of confusion and concern. Damask returned. “Yes, her aura was vivid red, not auburn as it used to be.”</p>
<p>“Did any of you notice something strange about her eyes by any chance?”</p>
<p>It was the mother who answered. “Now that you mention it, yes. I thought it was just an optical illusion, but there was one moment when for a brief second her eyes seemed to shine reddish.”</p>
<p>“Then my suspicion was likely correct, namely that your daughter is under the influence of an old magical artifact that I believed destroyed, the Alicorn Amulet.” <em>Then the one I destroyed must’ve been just another unicharm. I never </em>can<em> tell those things apart.</em></p>
<p>She counted only two gasps. “You are not surprised?” she asked Damask and September.</p>
<p>They hesitated. Finally September replied. “Well, we had been reading about the Alicorn Amulet and it was clear that earlier or later she would try to find it to destroy it or give it to you for safekeeping. She did not tell us about her concrete plans.”</p>
<p>“Good. I will have more questions about that later, but for now I would like you to travel to Manehattan to continue your search there. Praetoria will accompany you.” The ponies looked uncomprehending. “When the filly I now know to have been your daughter first hovered over the city and then took off, her direction was that of Manehattan. In between the only sites of importance are a mountain range and the Hollow Shades, so I think it is likely that she was headed for the city.</p>
<p>“The Berry family refused their assistance in the matter. The parents did, to be precise. They seemed relieved about Amber Rose’s disappearance and explicitly specified that they would not allow her daughter to be in contact with any of the <em>Rose family</em>.” She looked significantly at September. The colt seemed just confused.</p>
<div> </div>
<p>The stadium was easy to find. From her hotel window, Amber could already see its lighting towers, and after a few minutes of walk, the oval of its membrane roof became visible over the roofs of the city.</p>
<p>A long double line of ponies wound around almost half the stadium when Amber arrived. She thought she should have come earlier, maybe even on the day before to sneak in and hide overnight, but she had not made any plans that she was aware of. She would have to improvise. It had been working out well for her over the past days.</p>
<p>Getting in line and waiting for half an hour or more only to be told that no tickets were left did not feel right at all. Amber just moseyed past all the sports enthusiasts hoping that some idea would occur to her. Many of them wore items of apparel that expressed their allegiance; others were more formally dressed but still wore a tie pin for the same purpose.</p>
<p>Soon she saw a black stretch carriage drawn by two steeds in black come to a hold on the stadium’s carriage parking lot. The two black-clad unicorns removed their harness and one of them opened a door. Amber hurried to get closer.</p>
<p>There she was! The famous River Eddy stepped out of the carriage, cradling her dog under one wing. She wore an extravagant dress that made her look like some sort of exotic plant.</p>
<p>Amber teleported on the opposite side of the carriage. That allowed her to get close to them quickly without having to fall into a hurried, undignified gallop. She also appreciated that she had materialized a gown for herself in imitation of a silken dress with red seams that she had taken a liking to when she inspected the ponies in the line.</p>
<p>She walked around the carriage. “Oh! Can it be? River Eddy! It’s such an honor to be able to welcome you to our stadium today!” <em>Our stadium?</em> Amber was not sure where she was going with this, but it had felt right. Yet she felt a slight flinch within.</p>
<p>The ponies in black spun around. River Eddy turned more slowly. “Yes, it’s quite an honor for me as well. I’ll be holding a speech later.” She inspected Amber. “Do you work at the stadium or did you mean <em>your city’s</em> stadium?”</p>
<p>“My dad’s the catering manager.” Amber again felt that flinch she could not comprehend.</p>
<p>“Oh splendid! Every stadium I’ve been to either had a great chef or a great medic!”</p>
<p>Amber wanted to correct her that he was the catering manager not the chef, but since it was a lie to begin with, she let it slide. Then she noticed the joke.</p>
<p>They reached an inconspicuous door that was promptly opened for them from within by a pony in livery. “Ms. Eddy. An honor. Welcome to the Manehattan Stadium. I will escort you to your suite.” He hesitated for a moment. “According to my list, you wanted to come with two bodyguards, but the filly is not listed.”</p>
<p>“We just met. She’s the daughter of the catering manager,” River Eddy replied.</p>
<p>“I didn’t even know she had a daughter! Great to get to know you then. What’s your name?” the liveried pony asked. At the same moment, Amber noticed how Ms. Eddy’s bodyguards looked at their client who winked to them in reply.</p>
<p>“It’s Amber.” She felt that River Eddy’s gesture had been significant but she did not understand why.</p>
<p>On their way to River Eddy’s suite, they passed a number of different luxurious rooms, all high up and all with floor-to-ceiling windows out onto the field. Some of these rooms could easily accommodate several score ponies. Others would be crowded with even a mere dozen ponies in them, but to make up for their smaller size, they were decorated even more lavishly and equipped with large, cushy couches.</p>
<p>It was one of those latter rooms that they eventually entered. The corner to the right of the door was reserved for the catering. A number of fixtures for preparing fresh food were installed there, but more trays on heating elements as well as yet unused heating elements indicated that more food was brought to the room from some central kitchen.</p>
<p>Thanks to her father’s influence, Amber recognized and could have named most of the dishes. For just a short moment, she considered reinforcing her cover story by presenting them to River Eddy, but a feeling swept the thought away. The feeling told her that it was unnecessary. She let herself fall into one of the couches.</p>
<p>The opening ceremonies had hardly started when one of the coordinators came up the aisle to escort Ms. Eddy down to a podium they had erected at one side of the stadium. She set her dog on her couch, told him to stay put, and went down to the podium together with her bodyguards and the coordinator.</p>
<p>Amber noticed the waves of laughter going through the audience, but she did not pay attention to the speech for long because after a few minutes the dog hopped from the couch and followed some whiff that Amber could not smell out of the door to the hallway. She followed.</p>
<p>The dog paid little attention to Amber, so she followed closely. A minute later, they were in a large room that was furnished like a restaurant. At that moment, however, hardly anypony was interested in eating. Most of the ponies were gathered at the fieldward window front. Amber saw that another speaker had taken Ms. Eddy’s place on the podium. He was talking turns reciting a prepared speech together with a filly, probably his daughter.</p>
<p>Amber was about to follow Ms. Eddy’s dog out through an inconspicuous door when she saw his owner fly down to the podium and land with a loud thud. Her bodyguards were hurrying to keep up.</p>
<p>“I’m going to let you finish,” she said to the speaker, “but this is an emergency.” She turned the microphone toward herself. “Citizens of Manehattan, valued guests! It is with a heavy heart that I have to inform you of the tragedy that has befallen us on this day of celebration.” She paused dramatically. “My dog has vanished. If you see the little Yorkshire terrier desperately searching the confines of this stadium trying to find his way back to me, please lend him your guidance and bring him back to my lounge.” She indicated her suite. “I promise a thousand bits for whoever finds him and brings him back to me. He listens to the name …,” but Amber could not hear the rest of her sentence. At the mention of the thousand bits finder’s reward, the stadium erupted into chaos as thousands of ponies started to search all niches and crannies for the pet, and pagasi circled outside and even within the hallways.</p>
<p>For a moment longer, Amber watched the other speaker hopelessly trying to gain the audience’s attention back, then she turned to follow the diminutive dog through the door it had managed to press open.</p>
<p>She entered a part of the building where a laminate floor replaced the soft carpet she had grown accustomed to, and where the walls were of béton brut and no longer hung with series of framed photos. She caught up with him just as he entered the kitchen.</p>
<p>The dog stopped close to the entrance, where at least a dozen trays of fried soy cutlets filled the air with spicy odors. They were too high up for him to reach them.</p>
<p>A waiter restocked a cart with a few of the trays and was already hurrying off again when she noticed Amber. “Dogs are not allowed in here.”</p>
<p>As if he had heard her, the terrier abandoned the trays on the table and followed the cart instead.</p>
<p>“Uhm, we’re about to leave, I guess,” Amber said, following the dog.</p>
<p>The waiter took a staff elevator one floor up. Locked in the small room, Amber had to tug the dog softly on the tail to keep it from pilfering the soy cutlets.</p>
<p>The waiter brought trays to a few of the luxury suites. Soon Amber recognized one of them. An exuberant pegasus flew toward her.</p>
<p>“Amber! You are an angel!” Ms. Eddy swept up her dog while landing in front of her. “Wherever did you find him?”</p>
<p>“Uhm, I kind of just, uh,” Amber explained.</p>
<p>“Come on down! We have to bring these glad tidings of joy to the world!”</p>
<p>River Eddy grabbed Amber in one foreleg, her dog in the other and flew all the way down to the podium again. The speaker with his daughter were still unable to continue their speech, but the stadium security had booted enough ponies from the premises to restore some first signs of order. Many of the rows were almost empty now.</p>
<p>“Ponies of Manehattan! Ponies of Equestria! Nonponies of the world! The state of emergency is hereby lifted: My dog has been restored to me!” Slowly, hundreds of disappointed ponies returned to their seats. “This filly, this fairy, this angel, has granted my dearest wish and has reunited me with him.” She held her dog up on one hoof and Amber on the other. Amber tried not to look as floppy up there as the dog did. “Amber! Please tell the world how you granted this dearest of wishes to me, my dear.” She held her toward the microphone.</p>
<p>“It’s just what I do,” Amber said. Her voice reverberated back to her from all sides. A stadium full of thousands of ponies had just heard her voice. It was intoxicating.</p>
<p>“Granting wishes is just what you do. A little fairy indeed!”</p>
<p>Amber yearned to say more to this stadium full of attentive ponies, but she could not think of anything. Then she noticed the other filly, the daughter of the other speaker, craning her neck toward the microphone.</p>
<p>“Amber?” she asked. “If you grant wishes, could you grant me one wish too?”</p>
<p>“Of course. Anything you’d like. What is it that you wish for?” Amber again enjoyed hearing her voice from all sides and seeing the thousands of spectators in rapt attention.</p>
<p>“I wish to see a real live dragon!”</p>
<p>“Of course. I will show you a real live dragon before these games draw to a close!” As much as she enjoyed herself, Amber was also shocked at her own promise. Where would she find a dragon? Those large, pleading eyes with three reflections—she could not possibly let that poor filly down.</p>
<p>“Thank you,” the filly said.</p>
<p>It was time to yield the stage to the legitimate orators. With a heavy heart, Amber followed Ms. Eddy up the stairs back to the suite. Climbing the stairs, she suddenly remembered the soap factory she had overflown. She knew how she would get that filly her dragon!</p>
<p>When she heeded her surroundings again, Ms. Eddy was climbing next to her. She folded one wing over Amber. “Who are you? I know you’re not the daughter of the catering manager.”</p>
<p><em>How did she know!</em> “I’m just another filly.”</p>
<p>“C’mon.” A glottal fry accentuated her disbelieve.</p>
<p>Amber felt uncomfortable in the situation. Spontaneously she teleported away.</p>
<p>She found herself on the roof of the stadium again. The membrane was so taut that it hardly gave way under her hooves and the view of the city was glorious. Her position was tactically advantageous as well, because from this higher vantage point, she could make out the soap factory.</p>
<p>Even with her amulet-enhanced magic, it took her some concentration to shape a tornado, but when it hit the ground, she was satisfied with the result. Two of the tanks bust into shreds of metal and the liquid crystal was sucked up into the clouds. A minute of maneuvering later, the mass of soapy water reached the patch of sky above the stadium, held in place by a gigantic aura of red.</p>
<p>A third speaker, maybe some Manehattan politician Amber did not know, occupied the podium when the sun was blotted out by the vivid red cloud. A few ponies in the audience noticed the filly at the edge of the roof, her horn shining and sparkling in the same distinctive tone of red. Some may have even noticed the amulet glowing in the same color beneath Amber’s silken dress. The politician interrupted her speech and stared skyward.</p>
<p>Amber was sweating. The exercise was demanding even with amulet, and she was glad that she could now just release her hold. The red aura vanished, and the cloud shone in thousands of iridescent colors in the sky.</p>
<p>She had not expected that it would take so long for the soapy water to reach the ground. The scintillating play of the sun on the crystal structures captivated the audience for half a minute longer, then the politician could gain their attention back. A few paragraphs later, the stadium, the field, a third of the audience, and Amber herself were bathed in soapy rain. Everything shone in brilliant colors. It was beautiful.</p>
<p>A few ponies realized what it was and ran toward the exits. Most were trying to contrive of actual or magical umbrellas to protect themselves from the sticky downpour. The politician ran for cover.</p>
<p>Soon Amber saw the first dragon on the southern horizon. She teleported to the podium. “Within a few minutes, a dragon will arrive at this stadium. Your …,” Amber looked around but could not find the filly who had wished for a dragon. “The filly’s wish has been granted.”</p>
<p>Amber noticed a commotion behind her and turned to see September come through a gate. Praetoria was talking to some of the security guards.</p>
<p>The dragon was faster than she had expected. While Amber’s back was turned on much of the stadium, he arrived and circled it twice. He noticed the ruse, and he did not enjoy being fooled.</p>
<p>His fiery breath set several sections of the roof membrane ablaze. The thin film of soapy water hardly slowed the fire down. “Bucking soap!” he roared.</p>
<p>Now thousands of ponies ran toward the exits and onto the field, countless pegasi took flight, often with loved ones in their grasp. A few dexterous unicorns merged their fields to shield the masses from the drops of molten roof membrane.</p>
<p>The little filly had wished to see a dragon. Amber had granted her wish. That was all that counted.</p>
<p>Another burst of fire melted some steel cables, and a lighting array came crashing down next to the stage just as September reached it. It missed him by inches.</p>
<p>“September! What are you doing here?” Even though she had her back turned to it, the microphone picked up her cries. Three more ponies emerged from the gate and ran toward her. “No! Damask! Don’t come closer!”</p>
<p>“Hei sisko?” he said tentatively.</p>
<p>Amber did not reflect on what she did next. She walked straight toward her brother and stopped inches from him looking him straight in the eyes, their noses almost touching.</p>
<p>With a quick motion of her right hoof she ripped the amulet off her neck. “There you go.”</p>
<p>He took it into his magic.</p>
<p>Not at once but within just a few seconds Amber’s world collapsed twice. First the posthypnotic suggestion collapsed. <em>No! My amulet!</em> Then the power of the amulet itself collapsed. <em>No! What have I done!</em></p>
<p>Amber looked into the stunned faces of her parents, the security ponies, and some of the audience who heard the exchange over the din of the conflagration. She wanted to apologize, she wanted to flee to some part of Equestria where no pony knew her name, she wanted to escape into the past again and make it all undone.</p>
<p>She decided that it was not the right time for any of these plans. She shot lines of fire across the floor of the podium while she turned toward her young earth pony friend.</p>
<p>“September, rally up some pegasi and coordinate them to create a huge deluge over the stadium. Try to find something acidic to mix into the water. We need to extinguish the fire, and dissolve and dilute the soap.” She jumped over one of her fiery lines. “I’ll distract the dragon.”</p>
<p>The lengths of carpet had burned through along Amber’s lines of fire and she ripped a segment from the floor, pulled it taut, jumped onto it, and shot out over the field.</p>
<p>“Nothing but bucking soap!” the dragon shouted in a voice almost too deep for ponies to perceive, except by pallesthesia. He kicked a lighting tower. The half-molten trusses bent and snapped. The tower hit a burning section of the stadium roof. It caved in on itself and scattered debris across the field.</p>
<p>Amber dove down to the field, dodged a falling floodlight, flew so low that the carpet scooped up some of the puddles of soapy water, then gained altitude again and concentrated hard to recreate the complete shields that the amulet had allowed her to hold during flight. It was not a question of raw magical power, she found, but of technique, and she remembered some of the technique that she had used so effortlessly when she was under the curse.</p>
<p>A second, smaller dragon stopped with two powerful beats of his wide wings and regarded the stadium from aloft. “What is your substance, whereof are you made, that millions of strange shadows on you tend?” he rumbled.</p>
<p>“Just bucking soap!” the first dragon replied.</p>
<p>“I know. It felt too big anyway.”</p>
<p>The first dragon tackled the stadium and deformed the elliptical structure into an egg. The remaining sections of roof membrane snapped. “I hate soap!”</p>
<p>“Dude, chill. It’s not like they spilled all that soap on purpose.”</p>
<p>“Yes they did! They just love messing with us! Don’t believe their lies!”</p>
<p>The shield appeared around Amber. Now it was simple to shape it to resemble the sharp edges of a cut gemstone and opaque it from the outside in ruby colors.</p>
<p>The second dragon pointed. “Look! Where did that come from? Ruby, ruby, ruby, ruby!”</p>
<p>“Another ruse, I’m telling ya! Rubies don’t fly.” He stared at the gem. “It’s mine!”</p>
<p>“Whoever catches it first!” the second dragon countered.</p>
<p>Amber saw the two dragons dive at her from the front, slightly to port. They had come from a southeastern direction, so that’s where she had to lure them back to.</p>
<p>Upward acceleration hurt the most. Amber aimed just slightly above the outer edge of the stadium wall, right through the structure that had held the stadium roof only minutes before. The dragons changed their course to intercept their gem.</p>
<p>Amber was pressed into the slightly slanted carpet with two times her weight. She could no longer stand, yet it seemed to her that the first dragon would intercept her anyway. She slightly corrected her course, then bit her teeth together and cranked the acceleration up by another notch.</p>
<p>“The gem’s escaping!”</p>
<p>“No!”</p>
<p>Her ribs compressed her lung painfully. The carpet, although magically tautened, bent under her multiplied weight. She shot through the bare roof structure.</p>
<p>One outstretched claw grasped for her. The longest talon scraped along the shield. Suddenly she was high above the city.</p>
<p>Amber continued accelerating but reduced the force enough so she could scramble to her hooves. The dragons lost time changing their course, but a few flaps of their muscular wings later they were in pursuit again.</p>
<p>Shortly later Amber dared to look back again. Manehattan lay far behind. Thick dark clouds were building up above the city. <em>Good job, September.</em></p>
<p>But the dragons were also farther away than she had expected. They were already flying as fast as they could. She had to make sure they did not give up on the chase.</p>
<p>She braked the carpet and turned to the bow again.</p>
<p>A third dragon was dead in front of her, on collision course, her mouth wide open. Amber shot upward at almost forty-five degrees, having her body squeezed against the carpet again. The remaining soapy water was pressed off to the sides.</p>
<p>“Just soap again in Manehattan,” one of the pursuers warned the newcomer.</p>
<p>“But this one looks yummy! Why does it fly?” she asked.</p>
<p>“It’s mine!” the largest dragon shouted at her.</p>
<p>The new dragon reacted immediately and climbed at an even steeper angle while she spiraled around her own axis.</p>
<p>Amber whizzed past above her, but as she passed, the dragon turned again and engulfed the flying ruby in her fiery breath.</p>
<p>The carpet and Amber’s dress steamed in the heat and caused her eyes to tear up. She blinked it away. “Shields holding,” she reassured herself.</p>
<p>Beneath her was only the wide ocean now. If the shield collapsed, the airstream would knock her off the carpet at once. To the right, almost behind her, she could see the skyline of Fillydelphia. Baltimare was already visible as well.</p>
<p>All three dragons were behind her now, the female one much closer than the males.</p>
<p>A second burst of fire hit her from behind.</p>
<p>The aft shields failed at once and the carpet, Amber’s tail, and her formally elegant gown were set ablaze. She ripped off the clothes and waved her tail to extinguish the fire, but the carpet was almost dry again and the burning material filled the remaining shields with noxious smoke that burned in her eyes and lung.</p>
<p>“It’s just another one of those ponies!” The first dragon joined into the fire-spitting.</p>
<p>Amber stood at the bow of the almost consumed carpet when she saw the third wave of fire roll toward her.</p>
<p>She felt her consciousness on the verge of slipping away as she gasped for air, still she managed to jerk the remaining carpet up behind her as a flimsy substitute for the shield. The airflow hit it much harder than it hit her, and she was catapulted out into the air.</p>
<p>The third wave of fire burned the carpet away within the second but it shielded the heat enough for Amber to get away with some more singed hair. Now she had fresh air as well, too much of it, in all directions.</p>
<p>The first dragon drew breath again, but the others stopped him.</p>
<p>“She’s not gonna survive this anyway,” Amber heard one of them say. Then she had fallen too far to hear them.</p>
<p>She tried frantically to levitate herself like she had with the amulet. The gasses still burned in her lungs, her eyes burned worse, her head hurt. She tried to find something out on the open sea that she could levitate up to her as a carpet substitute. She could see nothing and even if she did, it would have been out of her reach until it was much too late.</p>
<p>Still dazed by the gasses, she appreciated the foreboding ambiguity in the term <em>terminal velocity</em>. She shook her head. <em>No, don’t give up!</em></p>
<p>She tried to create a shield below herself to use the air resistance to slow down her fall. If it had any effect, she did not notice it. The waves, their surface surely hard as stone at her speed, became visible. Mere seconds away.</p>
<p>One slip in her concentration and the shield lost its precarious balance against the airflow and spun about setting Amber spinning as well. Up and down changed too quickly, she could not recreate it.</p>
<p>Sky flashed past, then the water, sky again, water. The surface raced toward her. Amber held her breath just on the off-chance that she might survive the plunge.</p>
<p>Something touched her softly from below and stopped her spinning. Then less softly. A second more and it was hard as a brick wall.</p>
<p>Her vision was reeling, she could not see where she was, but it was not water, she could breath. Then the force of the brick wall increased further and she could not breath anymore.</p>
<p>Then it was all over. The brick wall was soft again, and she hung limply from its edge. She puked into the water below. It was so close, she could almost touch the wave crests with her hooves.</p>
<p>A different hoof touched her shoulder.</p>
<p>“Amber, you look awful.” It was Juniper who greeted her so politely.</p>
<p>At once the two hugged.</p>
<p>“How do you do that, saving my life all the time?”</p>
<p>“September said they would go looking for you in Manehattan, he and your family that is. I was feeling better and wanted to come along, but my parents wouldn’t allow it.” Juniper made a spiraly gesture toward her head. “I snuck out on the hotel mattress as you can see. Good thing too, so I could catch you much more softly.”</p>
<p>“Why weren’t you with my family at the stadium then?” Amber asked.</p>
<p>“They had a few hours head start. When I got there, Damask told me that the three dots on the horizon were you with two dragons in pursuit, so I figured you might need help. When I saw you fall from the sky, I went so fast my shield started to glow. So cool!” She beamed at the scorched and dizzy Amber. “Oh, keep an eye out for your mom on the way back. She couldn’t keep up.”</p>
<div> </div>
<p>“Your Highness, you wanted to see me?”</p>
<p>“Amber Rose, yes.” Princess Celestia motioned for her guard to close the door through which she had escorted Amber. “I want to thank you for finding the amulet for me. In the wrong hooves, it could’ve caused great harm.”</p>
<p>The princess set down a plate of cloud cake and gulped down her last bite. She indicated a zafu next to the one she was sitting on. Amber sat down.</p>
<p>“It has caused great harm in my hooves,” Amber said. “Did anypony get hurt?” <em>Of course they did.</em> Her mother had treated Amber’s shallow burns on their train ride back to Canterlot because the hospitals in Manehattan were all over capacity.</p>
<p>“Many burns and bruises but nothing serious. The evacuation went very smoothly.” The princess looked out of the window at her beautiful sunset. “However, rebuilding the stadium will take a few years, and the Equestria Games will have to move to a different city this time if they can even still take place.” She turned back to Amber. “You better stay away from Manehattan for a few years.”</p>
<p>“Then I’m not going to prison?” Amber felt constricted as if she were pressed into a carpet during a strong acceleration. <em>What will be my punishment?</em></p>
<p>“Punishment has no place in restorative justice. The taking of a pony’s freedom is an extreme measure that is reserved as a very last resort when there is no hope that the criminal will amend their ways and they pose a threat to society. There have been very few cases throughout the centuries that required such extraordinary means.” She gulped, then her aspect lightened up. “And I don’t think you would do all that nonsense again.”</p>
<p>But Amber was only partially relieved. “It must seem as if the amulet had controlled me, but throughout I always felt that I was controlling it, not vice versa.” Now Amber gulped too and continued in a <em>sotto voce</em> she had rarely heard herself use. “Some things I did even now seem natural to me though of course I would never <em>actually</em> do them.”</p>
<p>To her surprise Celestia just nodded. “Yes. I think I’ve gained some insight into the way the amulet works. Central to its ability to channel such tremendous magical energies through common unicorns appears to be that it forces extreme spontaneity by suppressing some conscious functions—and thereby also the conscious parts of what in the past has been termed the <em>superego</em>, the ethical component of your personality.</p>
<p>“But no one component of your personality makes you. It takes all of them to form the Amber that stands in front of me, and that Amber I trust to make decisions that are loving <em>and</em> rational.</p>
<p>“Just as I trust that without the amulet, Trixie will stop short of enslaving a whole town, ignoring the fact that she won’t be able to anyway.”</p>
<p>“You know about that?” Amber was not actually surprised this time.</p>
<p>“Your brother briefed me on your plans, which is the main reason I wanted you to visit me here. We should put your endeavors on a little more, dare I say it, professional basis.</p>
<p>“I have established two teams, one to investigate the unusual mode of time travel that you have experienced and one to do what you did, to try to avert any catastrophes that can be reconstructed from the material that traveled with you. To redeem yourself, you will be required to advise both teams in any way you can.</p>
<p>“I also restored your graduation from school. If your advisory functions should bore you, you will be required to pick up your studies again, so you can continue to redeem yourself past the eight years in which your memories will be of value.”</p>
<p>Amber was an itsy bit overwhelmed. “Thank you. I will assist in any way I can. Has my brother also briefed you on the problems that will befall the Griffon Kingdom in about five years? You should be in a position to warn them.”</p>
<p>“Good to know the amulet has left no permanent damage,” Celestia said. “Yes, he has and I will, as soon as I have evaluated the evidence. Your brother is already with the second of the teams recreating your notecards from memory. You should join them. My guard will show you the way.”</p>
<p>“Yes, I’ll double-check them. Sometimes his memory plays tricks on him, though he doesn’t like to admit it.”</p>
<p>A few valedictions later, Amber was on the way to pick up where she had left off a few days earlier. Eight years seemed like a very long time to her, but what did Celestia expect her to do once she could no longer predict any disasters? Or would she have to watch the decades of her life dwindle away all in the knowledge that the only eight years of her life that had held any purpose were gone? Life like that seemed utterly pointless.</p>
<p>They passed the guarded door to Luna’s bedchamber just when the night shift took over. The guard who escorted Amber greeted all of them, then she guided Amber down a staircase and they were out of sight again.</p>
<p>No, there were a million ways for her to leave a positive impact in the world even without her foreknowledge. She could support campaigns for bat pony rights, organize marches against cutie mark readings, expose companies that cut corners in dam safety, and further the research into earth pony magic. Amber had her work cut out for her, finally.</p>Mapping Directions of Aggressions in The Bluest Eye2013-09-15T18:00:00+00:002013-09-15T18:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2013-09-15:/mapping-directions-of-aggressions-in-the-bluest-eye.html<p>The present paper analyzes directions and qualities of aggressions in Toni Morrison’s <em>The Bluest Eye</em>. To this end it establishes purview and methodology before presenting the visualized graph that results from an annotated adjacency list of the aggressions.</p><style>
body {counter-reset: section}
.entry-content h2 {counter-reset: sub-section}
.entry-content h3 {counter-reset: composite}
.entry-content h4 {counter-reset: detail}
.entry-content h2:before {
counter-increment: section;
content: counter(section) " ";
}
.entry-content h3:before {
counter-increment: sub-section;
content: counter(section) "." counter(sub-section) " ";
}
.entry-content h4:before {
counter-increment: composite;
content: counter(section) "." counter(sub-section) "." counter(composite) " ";
}
.entry-content h5:before {
counter-increment: detail;
content: counter(section) "." counter(sub-section) "." counter(composite) "." counter(detail) " ";
}
</style>
<div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#definitions-of-aggression">Definitions of Aggression</a></li>
<li><a href="#classifications-of-aggressions">Classifications of Aggressions</a><ul>
<li><a href="#buss">Buss</a></li>
<li><a href="#ramirez">Ramirez</a></li>
<li><a href="#sue-et-al">Sue et al.</a></li>
</ul>
</li>
<li><a href="#types-of-textual-representations-of-aggressions">Types of Textual Representations of Aggressions</a></li>
<li><a href="#observations">Observations</a><ul>
<li><a href="#applicability">Applicability</a></li>
<li><a href="#composite-aggressions">Composite Aggressions</a></li>
</ul>
</li>
<li><a href="#the-graph">The Graph</a></li>
<li><a href="#appendix-table-of-aggressions">Appendix: Table of Aggressions</a></li>
<li><a href="#endnotes">Endnotes</a></li>
<li><a href="#works-cited">Works Cited</a></li>
</ul>
</div>
<h2 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h2>
<p>A number of structural patterns in nature and direction of the
interactions that Toni Morrison portrays in <em>The Bluest Eye</em> stood out
to me upon my first reading of the novel. Many of these interactions
related to aggressive behaviors in more or less apparent ways. These
individual aggressions, however, may have just attracted my attention
due to a possible predisposition on my part for the corresponding
patterns, in which case confirmation bias would have lead me to
involuntarily ignore other aggressive behaviors contradicting them.<sup id="fnref:1"><a class="footnote-ref" href="#fn:1">1</a></sup></p>
<p>In particular it seemed that it was often an older generation that acted
aggressively toward a younger generation along the cascade of the at
least three generations that the book traces upstream, as it were, into
the past. What rendered this hypothesis consciously plausible to me was
the relative impressionability and vulnerability of children, and their
parents’ position of authority (<a href="#vissing1991verbal">Vissing et al.</a>; <a href="#huntsman2008parents">Huntsman</a>).</p>
<p>To minimize the influence of personal biases and thus solidify the
empirical basis of further inferences, this term paper attempts to
record aggressions in <em>The Bluest Eye</em> as comprehensively as possible
and visualize them algorithmically to make any such flows of aggressions
apparent, if they exist.</p>
<p>One hurdle in this respect is that aggressions can vary widely in their
intensity, from China’s teasing of Marie
(<a href="#morrison1999bluest">Morrison 39, 41</a>), an adult, over the
microaggression by Claudia’s parents on their daughters when “Frieda and
[Claudia] were not introduced to [Henry] merely pointed out”
(<a href="#morrison1999bluest">10</a>), over Charles Breedlove’s traumatic
experience when white hunters forced Darlene and him to have sex in
front of them (<a href="#morrison1999bluest">116</a>), to the two times that
Charles Breedlove rapes his daughter Pecola
(<a href="#morrison1999bluest">128, 158</a>). Unfortunately, there are only
few pairs of cases that can be ordered with respect to their intensity
so that enough observers would agree with the order to call it
objective. What is helpful in this case is that there are several
established systems of classification of aggressions (see section 3)
that allow us to assign categories to the aggressions, categories that
are more readily distinguishable.</p>
<p>Another hurdle is that the representation of aggressions in the literary
text is sometimes not straightforward. Action or dialogue may convey
aggressions very directly, but it is also possible that an aggression,
while very explicit for the victim, is only implied for the reader. One
example is the scene at the “big white house with the wheelbarrow full
of flowers” where Paulina Breedlove works as domestic aid. At the end of
the scene, Pecola (together with Claudia and Frieda) is leaving while
the family’s child asks who they were. Mrs. Breedlove evades the
question, effectively denying her daughter in front of her
(<a href="#morrison1999bluest">85</a>).</p>
<p>Interior emotion or monologue may also be straightforward in its
descriptions of aggressions, but in other cases, the victims themselves
may not be aware of the aggressive nature of an interaction they
experienced.<sup id="fnref:2"><a class="footnote-ref" href="#fn:2">2</a></sup> Pecola’s encounter with Mr. Yacobowski, the owner of a
candy shop, during which Pecola is the point-of-view character, is told
in a quick alternation between action, dialogue, interior emotion, and a
little interior monologue. She is unable to speak in his presence and
sentences such as “His nails graze her damp palm” reveal her fear, or
more precisely shame, as it is made explicit in the following sentence.
Then she does feel a sense of vague anger, but only momentarily before
it is replaced again by shame. Underneath lies Mr. Yacobowski’s
aggression in his disrespectful and contemptuous behavior toward her.
This is amplified by her identification with dandelions. She initially
liked them, but after the experience acquiesces to popular opinion and
decides that they—and by implication she herself—are weeds and ugly,
whereby she blames herself for Mr. Yacobowski’s show of contempt<sup id="fnref:3"><a class="footnote-ref" href="#fn:3">3</a></sup>
(<a href="#morrison1999bluest">47–50</a>).</p>
<p>Similarly, in the wide field of description and narrative summary, some
aggressions are not themselves described or summarized but strongly
implied as premise for the described. On page five already, Claudia
describes first the travail she and her family put themselves through to
collect “the tiny pieces of coal lying about” near the railroad tracks
that lead to a steel mill. In the following paragraph, she describes
their home as “cold.” Only a few pages later, she explains the concept
of “outdoors” in more detail and mentions that “to be put outdoors by a
landlord was one thing—unfortunate, but an aspect of life over which you
had no control, since you could not control your income,” which
highlights the discrimination of people of color on the job market.
Taken together, these scraps of summary and description portray clearly
the intersectional, multifaceted forms of aggression that Claudia’s
family and community, and by extension people of color throughout the
<span class="caps">US</span>, were (or are) exposed to, the aggression of racial oppression from
whites and capitalist oppression from the owning class.</p>
<p>In the following, section 2 will first establish a reasonably
precise definition of aggression as the term is used in this paper.
Then, section 3 will introduce the three classification systems that
will serve as guide for a qualitative analysis of the aggressions in
<em>The Bluest Eye</em>. The distinctions in the textual representation of
aggressions that I will observe will be explained in section 4.
On this basis, section 5 will turn to observations on the process
of applying these typologies to the text. The final section, section 6,
will present the result, the graph, which will also answer my opening hypothesis.</p>
<p>The full annotated classification table that forms the basis for the
graph is attached as appendix to the paper.</p>
<h2 id="definitions-of-aggression"><a class="toclink" href="#definitions-of-aggression">Definitions of Aggression</a></h2>
<p>Polysemy, “the greedy habit some words have of taking more than one
meaning for themselves,” (<a href="#McKean2007">McKean</a>) is a core feature of
natural languages. The term <em>aggression</em> is certainly not an epitome of
polysemy, like <em>set</em> or <em>run</em>, but in different contexts the intended
meaning will also be subtly different. Hence, what this section seeks
to establish is not an encompassing definition of <em>aggression</em> but one
that will help describe the purview of this survey.</p>
<p>Requirements for such a definition are that it be specific enough to
allow the reader to decide in most of the cases found in <em>The Bluest
Eye</em> whether they constitute an aggression according to the definition;
that it be inclusive enough to capture most of the aggressions that have
significant effects on the characters; and that it be unencumbered by
the difficulties associated with passing moral judgment on aggressors,
since such an endeavor would go far beyond the scope of this paper.</p>
<p>Dictionary definitions such as (1) “An unprovoked attack; the first
attack in a dispute or conflict; an assault, an inroad,” (2) “The
practice of attacking another or others; the making of an attack or
assault,” (3) “Feeling or energy displayed in asserting oneself, or in
showing drive or initiative; aggressiveness, assertiveness,
forcefulness. (Usu. as a positive quality.),” or (4) “Behaviour intended
to injure another person or animal; an instance of this”
(<a href="#OEDaggression">Oxford English Dictionary</a>) are unfortunately alternately limited to
aggressions between countries, too broad to be useful in this context,
or entirely reliant on synonyms.</p>
<p>Within the academic literature, one popular definition appears to be the
following: “Aggression is any form of behavior directed toward the goal
of harming or injuring another living being who is motivated to avoid
such treatment.” (<a href="#baron2004human">Baron and Richardson 7</a>)</p>
<p>This definition makes five important choices: (1) that aggression is a
behavior rather than an emotion, attitude, or motive<sup id="fnref:4"><a class="footnote-ref" href="#fn:4">4</a></sup>; (2) that the
aggressor intends the aggression; (3) that it aims to harm or injure the
victim; (4) that the victim needs to be a “living being”<sup id="fnref:5"><a class="footnote-ref" href="#fn:5">5</a></sup>; and (5)
that the victim is motivated to avoid the aggression.</p>
<p>These observations are valuable, but one of them critically limits the
scope of the term. Although Baron and Richardson observe that the
notion of intend is a problematic one to include since it is not
externally observable, they argue that to exclude it would make it
“necessary to describe the actions of surgeons, dentists, and even
parents when disciplining their children as aggressive.”
(<a href="#baron2004human">9</a>) As such, however, the definition would
exclude a range of microaggressions—specifically many microinsults and
microinvalidations, a distinction that will be explained in the
corresponding subsection—from the definition, as they are often inflicted
negligently rather than intentionally. These microaggressions have
critical effects on many characters in the novel as well as in reality
(<a href="#Sue2007">Sue et al. 278</a>).</p>
<p>To resolve this conflict, it should be observed that, firstly, in the
cases of surgeons and dentists, the fifth criterion does not apply;
evidently, their patients are more strongly motivated to undergo
treatment than to avoid it.<sup id="fnref:6"><a class="footnote-ref" href="#fn:6">6</a></sup> Optionally, changing the definition to
say “predominantly motivated” would make this clearer, but introduce
another potentially vague term. Secondly, categorizing the action of
parents “disciplining their children,” whether verbally or physically,
as aggressive does not seem to conflict with the notion of aggression
when removed from its morally judgmental overtones, as I intend to do
here. A verbal chastisement may often be legitimate despite its
aggressive nature.<sup id="fnref:7"><a class="footnote-ref" href="#fn:7">7</a></sup></p>
<p>Importantly, they also point out that omission of any mention of intent
would remove from the scope of the definition any (intuitively)
aggressive actions that were foiled. An imperfect example of this might
be Claudia’s blow against Maureen, which missed. She instead,
accidentally, hit Pecola (<a href="#morrison1999bluest">Morrison 56</a>). None of this
would intuitively strip the aggressiveness from the behavior. This
argument, however, should not be read as an argument against the
exclusion of intent from the definition in any form but merely against
the exclusion of intent as sufficient condition for an aggression.</p>
<p>Consequently, I will describe as <em>aggression</em> in this text any form of
behavior negligently or intentionally harming or injuring another living
being who is motivated to avoid such treatment, or the attempt thereof.</p>
<p>It should also be noted that the definition does not make mention of the
precise nature of the perpetrator, so that it encompasses aggressions
that are inflicted by groups rather than individuals that could clearly
be pinpointed. Aggressions that will later be described as systemic or
societal count into this category.</p>
<p>What the definition excludes, however, are aggressions toward or by
inanimate objects. When Claudia dismantles her dolls, the aggression is
not one against the doll but one that bifurcates into, on the one hand,
a symbolic or at that point hypothetical rebellion against the
societally dictated ideal of beauty that the doll represents and, on the
other hand, a more concrete aggression against her parents, who may be
irked by her lacking appreciation of their gift.<sup id="fnref:8"><a class="footnote-ref" href="#fn:8">8</a></sup> Conversely, the
perceived aggression of the doll against Claudia (“It was a most
uncomfortable, patently aggressive sleeping companion”
(<a href="#morrison1999bluest">Morrison 14</a>)) does not count by itself constitute
an aggression of this sort by is rather another indication of the
systemic oppression that she perceives more subconsciously at her young
age and that causes her heightened state of irritability and vexation
with the doll.</p>
<h2 id="classifications-of-aggressions"><a class="toclink" href="#classifications-of-aggressions">Classifications of Aggressions</a></h2>
<p>Aggressions can have a physical and an emotional impact on the victim,
though it is hard to imagine a physical aggression that does not also
have an emotional dimension. When an aggression is effective purely on
an emotional level, it can often be described as microaggression
(<a href="#pierce1977experiment">Pierce et al. 65</a>), some of which are “ambiguous and
nebulous” as Sue et al. (<a href="#Sue2007">272</a>) described them, and thus
“difficult to identify and acknowledge.”</p>
<p>Categorizations of aggressions that I have encountered in the literature
have often been dichotomous, and Ramirez also observed
that they can usually be divided into systems “focused to distinguish
the form or mode of aggression [and] others interested in its function
or goal,” whereby his own study is one of the latter kind
(<a href="#ramirez2009">86</a>). The description of the “function or goal” is
one that scrutinizes the aggressor, while those focused on “form or
mode” (the next subsection will introduce
Buss’s system) concentrate on the aggression
itself. What is missing is a typology of its impact on the victim. At
least in the realm of microaggressions, Sue et al.’s system
fills this void.</p>
<h3 id="buss"><a class="toclink" href="#buss">Buss</a></h3>
<p>One taxonomy that is interested in the “form or mode” of the aggression
and that is also referenced by Baron and Richardson in their book
was proposed by Buss.
Griffin, O’Leary-Kelly, and Pritchard (<a href="#griffin2004dark">65</a>)
call it “the most widely recognized”
typology of aggressions, and Google Scholar lists more than 1,400
articles that cite Buss’s book. The typology
distinguishes three pairs of mutually exclusive modes of aggression, and
thus captures eight types in total.</p>
<div> </div>
<p><strong>Physical vs. verbal</strong>:
“Physical aggression … involves physical action on part of the actor,
whereas verbal aggression inflicts harm through words.”
(<a href="#griffin2004dark">Griffin, O’Leary-Kelly, and Pritchard 65</a>)</p>
<div> </div>
<p><strong>Active vs. passive</strong>:
“Active aggression requires the actor to do something to harm the
target, whereas passive aggression involves withholding something that
the target needs or values.” (<a href="#griffin2004dark">65</a>)</p>
<div> </div>
<p><strong>Direct vs. indirect</strong>:
Direct aggression has the actor harming the target directly, whereas
indirect aggression is proximate.<sup id="fnref:9"><a class="footnote-ref" href="#fn:9">9</a></sup> (<a href="#baron2004human">Baron and Richardson 10</a>)</p>
<h3 id="ramirez"><a class="toclink" href="#ramirez">Ramirez</a></h3>
<p>Following from his observation about the different foci of taxonomies,
Ramirez was able to integrate the systems he surveyed
and strip redundancies until his new system again distinguished three
pairs of alternative motivations or causes for aggressions. The terms
are fairly intuitive: hostile vs. instrumental, Impulsive vs.
premeditate,<sup id="fnref:10"><a class="footnote-ref" href="#fn:10">10</a></sup> and proactive vs. reactive.</p>
<div> </div>
<p><strong>Hostile vs. instrumental</strong>:
“Hostile aggression is an angry, unplanned act intended to harm another
person,” whereas “instrumental aggression is conceived as a premeditated
technique for obtaining a variety of objectives, such as some reward,
profit, or advantage for the aggressor.”<sup id="fnref:11"><a class="footnote-ref" href="#fn:11">11</a></sup></p>
<div> </div>
<p><strong>Impulsive vs. premeditate</strong>:
“Impulsive aggression tends to pursue immediate gratification, without
thinking or concern about consequences, delaying long-run costs,”
whereas “premeditate aggression … is planned, purposeful, intentional
and goal-directed.”</p>
<div> </div>
<p><strong>Proactive vs. reactive</strong>:
“Proactive aggression enacts aggression as an effective means for
obtaining external rewards and social goals, such as possession of
objects (i.e., instrumental) or dominating people (i.e., person-oriented
or bullying),” whereas “reactive aggression is a hostile reaction or
response to any perceived harm, threat or provocation.”
(<a href="#ramirez2009">Ramirez 87–93</a>)</p>
<div> </div>
<p>Unfortunately—and one can already recognize this in the definitions—the
pairs are not entirely independent. Ramirez observes
that the combination of instrumental, premeditate, and proactive and the
combination of hostile, impulsive, and reactive are disproportionately
common, based centrally on Polman et al.’s meta-analysis
of empirical studies and corroborated by three more studies cited in the
paper (<a href="#ramirez2009">93–94</a>).</p>
<h3 id="sue-et-al"><a class="toclink" href="#sue-et-al">Sue et al.</a></h3>
<p>A more specific category is that of microaggressions.
Sue et al., who developed a trichotomous taxonomy of
microaggressions, describe racial microaggressions as “brief and
commonplace daily verbal, behavioral, or environmental indignities,
whether intentional or unintentional, that communicate hostile,
derogatory, or negative racial slights and insults toward people of
color.” This concept can easily be extended to encompass any nonviolent
slights that oppressed groups are exposed to. They distinguish three
subgroups of microaggressions.</p>
<div> </div>
<p><strong>Microassault</strong>:
“Explicit racial derogations characterized primarily by a violent verbal
or nonverbal attack meant to hurt the intended victim through
name-calling, avoidant behavior or purposeful discriminatory actions.”</p>
<div> </div>
<p><strong>Microinsult</strong>:
“Behavioral/verbal remarks or comments that convey rudeness,
insensitivity and demean a person’s racial heritage or identity.”</p>
<div> </div>
<p><strong>Microinvalidation</strong>:
“Verbal comments or behaviors that exclude, negate, or nullify the
psychological thoughts, feelings, or experiential reality of a person of
color.” (<a href="#Sue2007">Sue2007 278</a>)</p>
<div> </div>
<p>These taxonomies by Buss, Ramirez, and Sue et al. are complementary,
with the first one focusing on “form or mode” (<a href="#ramirez2009">Ramirez 86</a>)
of the aggression, the second on their “function or goal”
(<a href="#ramirez2009">86</a>) for the aggressor, and and the last on their
impact on the victim. Apart from these differences, the first system
appears to be the most broadly applicable one. I will explain some of
the limitations of Ramirez’s taxonomy in
section 5, as they are not immediately obvious. What is
explicitly obvious, however, is that Sue et al.’s system only
describes microaggressions, and many aggressions in <em>The Bluest Eye</em> are
anything but “micro-.” While they may have ancillary aspects that fit
one of the definitions, applying the category to the whole of such an
aggression would be tantamount to a microinvalidation against the reader.</p>
<p>To reconcile these different systems, the applicable categories, and
only these, will be listed for each aggression.</p>
<h2 id="types-of-textual-representations-of-aggressions"><a class="toclink" href="#types-of-textual-representations-of-aggressions">Types of Textual Representations of Aggressions</a></h2>
<p>Just as aggressions can be taxonomically qualified, so can their
representations in a literary text. The section on motivation already
touched on some of the different contexts of aggressions in <em>The Bluest
Eye</em>in action and dialogue, in inner emotion or inner monologue, and in
description or summary—but what they all have in common is that they
either denote actual aggressions (within the partially fictional realm)
or hypothetical aggressions that take place on an embedded diagetic tier
as characters, including the narrator, merely imagine them.</p>
<p>What they also have in common is that they are either exacted at or in
the presence of their victim, or expressed in absentia, the latter mode
making their effect and their actual victims more difficult to determine.</p>
<p>The first chapter, for example, sees a number of unnamed characters
engaging in gossip and animadverting upon a woman named Peggy, who they
describe as “trifling” and a “heifer” (<a href="#morrison1999bluest">Morrison 8</a>).
One of them avers that Henry left another woman, Della, because of a
perfume she wore. Factuality of that assertion aside, they conclude that
Henry was a “nasty” “old dog” not for leaving someone over such a
trifling reason but for preferring someone’s natural over artificial
odors. This, of course, creates an adverse social climate for Peggy or
reinforces the existing one.<sup id="fnref:12"><a class="footnote-ref" href="#fn:12">12</a></sup></p>
<p>The effects are even more insidious, however. At the conclusion of the
dialogue sequence, Claudia reflects on the way she and Frieda perceived
the conversation as children: “We do not, cannot, know the meanings of
all their words, for we are nine and ten years old”
(<a href="#morrison1999bluest">10</a>). The prosodic qualities of the
dialogue were primary to them, so she described it as a “gently wicked
dance.” Thus Morrison shows how impressionable they are, bare of many of
the conscious safeguards adults develop, when they are subjected to the
normative influences that inform them that their body odors are noxious
and need to be drowned in “violet water”
(<a href="#morrison1999bluest">8</a>), a readily homemade perfume, lest they
earn the denigration of their community. This tendency of a societal
ideal to imbue the community with self-loathing, here olfactorically, is
reminiscent of the inimical ideal of visual beauty, whose deconstruction
is very central to the novel. Many of the aggressions throughout the
book are its product, just as the title alludes to it.</p>
<p>Another distinction that I would have found interesting but ultimately
dismissed is that between the concrete and the abstract description of
an aggression. Concrete would have been the aggressions portrayed in
action and dialogue with fleshed-out individuals involved, while
abstract would have been description that considers forms of aggression,
and other narrative modes where forms of aggression are described in
terms of groups or archetypes of people.</p>
<p>This distinction was of particular interest to me as it is a strong
determinant in how people react emotionally to information. Many studies
have shown that “people care more about identifiable than statistical
victims” (<a href="#kogut2005identified">Kogut and Ritov</a>;
<a href="#small2003helping">Small and Loewenstein</a>), which is a
central problem in the organization of philanthropy to combat
wide-spread hunger and desolation abroad. Potential donors concentrate
their generosity on helping out a single, well-publicized individuals
when the same donations could have saved scores or hundreds of lives of
“statistical victims.” Moreover, not even further education about these
victims’ situation helps; it merely reduces people’s readiness to donate
to the individual cases (<a href="#small2007sympathy">Small and Loewenstein</a>).</p>
<p>Toni Morrison surely found herself faced with the same problem as these
altruistic organizations, on the one hand wanting to emotionally
captivate and move the audience, and on the other show that the
conflicts she portrays were and are not individual cases but examples of
all-pervasive scourges of society. Claudia even reflects on their
comparative ease handing their plights so long as they remained abstract
(<a href="#morrison1999bluest">Morrison 11</a>).</p>
<p>Examples of concrete aggressions would have been Claudia’s parents
neglecting to introduce their children to Henry; the white hunters
forcing Darlene and Charles at gunpoint to have sex in front of them;
and Claudia imagining herself and Frieda beat up Rosemary. Examples of
abstract aggressions would have been the mention of “mothers [who] put
their sons outdoors” (<a href="#morrison1999bluest">11</a>); the description
of Charles’s “dangerously free” state that enabled him to commit various
kinds of abuse on people under his authority, whether hypothetical or
not; and the statement that “the white-collar occupations available to
black people” were few (<a href="#morrison1999bluest">135</a>).</p>
<p>While these aggressions seemed to fit rather neatly into the category,
many more were greatly more ambivalent and elusive when viewed through
its lens. One one level, by merit of its being a work of fiction, all
aggressions in the novel could be said to be abstract, made more
concrete in countless different ways only in the minds of readers. On
another level, the narration often intentionally equivocates here. When
Claudia introduces the reader to her intersectional positionality,
namely, to the ways their parents treated Frieda and her, to how the
community treated them and their parents, and how society treated their
community, all very succinctly on the first three pages before Mr.
Washington arrives, she uses the present tense to aggregate flash backs
that alternate between possible concrete examples, sometimes even direct
speech, (e.g., “Frieda and I lag behind, staring at the patch of color
surrounded by black” (<a href="#morrison1999bluest">5</a>)) and clearly
abstract descriptions (e.g., “When, on a day after a trip to collect
coal, I cough once …” (<a href="#morrison1999bluest">6</a>)), intermingled to
an extend that the distinction becomes blurred. This mode of narration
reminds the readers not to dismiss what they read as isolated incidents,
but to extrapolate from them to the anguish and later trauma of
Claudia’s whole ethnic group, class, and generation.</p>
<h2 id="observations"><a class="toclink" href="#observations">Observations</a></h2>
<h3 id="applicability"><a class="toclink" href="#applicability">Applicability</a></h3>
<p>There are four reasons that I have encountered that precluded the
applicability of a category to an instance of aggression in the novel:
(1) the aggression was outside its purview, (2) the (single) aggression
included several aspects contradictory in category, (3) insufficient
information was given to make the decision, and (4) the aggression kept
in such abstract terms that it could describe multiple concrete
aggressions of contradictory types. While the latter two are results of
the diegetic interpretation of the events, the first two are (at least
partially intentional) limitations of the taxonomies.</p>
<p>Types of microaggressions are typically inapplicable to physical
aggressions, which are not “micro-.” This is the most straightforward
case of inapplicability.<sup id="fnref:13"><a class="footnote-ref" href="#fn:13">13</a></sup></p>
<p>There were also cases where the categories of microinsult and
microinvalidation seemed to overlap. The abovementioned microaggression
of Claudia’s parents against her when they neglect to introduce their
children to Henry Washington could be seen simply as microinsult for its
“rudeness” and “insensitivity” toward their “identity” (quoting the
definition). This very identity, however, is that of individuals whose
“experiential reality” is that of conscious and “feeling” beings rather
than fixtures, an experience that their parents treatment “negates.”
Since the concept of the microinsult appears to be the more inclusive
one, I would tend to categorize these aggressions as microinvalidations.
Apart from this issue, the different types of microaggressions,
according to Sue et al.’s system, were for the most part
clearly discernible.<sup id="fnref:14"><a class="footnote-ref" href="#fn:14">14</a></sup></p>
<p>Buss’s typology performed similarly well.
There was only one problem that I encountered occasionally. The system
only distinguishes physical and verbal aggressions, so behavioral
aggressions would have to be counted as physical even though, to me,
they appear more similar to verbal aggressions or possibly constituting
a category of their own. I decided not to categorize them as either.</p>
<p>Finally, the classification according to Ramirez’s
system was the most challenging and its results should be viewed with
the greatest caution. One reason for this was surely that the assessment
of the motives of aggressors—central on for this system—is often left to
the reader. Although the narrator makes many of these motives explicit,
she introduces the story with the words “But since <em>why</em> is difficult to
handle, one must take refuge in <em>how</em>.” (Emphasis in original.) This
refuge in <em>how</em> is certainly frequent as well, so that the <em>why</em> is
sometimes open for debate.</p>
<p>Furthermore, the categories are, as mentioned above, interrelated, so
that some categories already imply or preclude categories other than
their counterpart. In particular the following three combinations are
impossible according to the definition, although the intuitive meaning
of the terms would in some cases leave room for some such
aggressions.<sup id="fnref:15"><a class="footnote-ref" href="#fn:15">15</a></sup></p>
<ol>
<li>An aggression cannot be hostile and premeditate, because hostile
aggression has to be “unplanned.” By implication, premeditate aggression
would have to be instrumental, but Claudia’s various actual or
hypothetical, planned attacks on Maureen Peal
(<a href="#morrison1999bluest">Morrison 48–50</a>) did not serve the goal of
obtaining any “reward, profit, or advantage.”</li>
<li>An aggression cannot be instrumental and impulsive, because instrumental
aggression has to be a “premeditated technique.” This seems intuitive,
but only because it implies that any amount of premeditation, however
short, satisfies the criterion. When Rosemary Villanucci spots Claudia,
Frieda, and Pecola in the bushes around their house and immediately
calls for Mrs. MacTeer (<a href="#morrison1999bluest">21</a>), she probably
has no clear conscious idea of what she is doing and why,<sup id="fnref:16"><a class="footnote-ref" href="#fn:16">16</a></sup> but the
definition would force us to assume premeditation.</li>
<li>An aggression cannot be instrumental and reactive, because reactive
aggression has to be “hostile.” However, anyone can readily imagine
aggressions that are reactive in that they are a “response to any
perceived harm, threat or provocation” and are yet not “unplanned.” When
the woman of the “family of slender means” terminated Mrs. Breedlove
employment and denied her the last outstanding salary as a form of
inverse incentive to leave Mr. Breedlove, she surely planned this
decision, yet it was clearly her reaction to Mr. Breedlove’s threatening
impression on her.</li>
</ol>
<p>Finally, there were also many aggressions that each have distinct
manifestations for different victims. One example of this are the
aggressions <em>in absentia</em> that were discussed in sec:representation, as
they can easily be microaggressive to mere bystanders why creating a
hostile environment for the intended victim. Another example are the
fights between Mrs. and Mr. Breedlove
(<a href="#morrison1999bluest">29, 96, …</a>), which not only constitute
violent aggression between the two of them but also left their children,
who observed the fights, deeply traumatized. If the different aspects of
an aggression appeared similarly important or likely, I often decided to
split them up and classify them separately.</p>
<h3 id="composite-aggressions"><a class="toclink" href="#composite-aggressions">Composite Aggressions</a></h3>
<p>“Composite aggressions” is my term for aggressions that can be broken up
into multiple or countless separate aggressions, limitations on the
available data notwithstanding.</p>
<p>The typical case are societal or systemic aggressions. These types of
aggressions are more or often less consciously perceived by the victims;
pinpointing any original perpetrator, however, is greatly more
difficult: “Couple the vulnerability of youth with indifferent parents,
dismissive adults, and a world, which, in its language, laws, and
images, re-enforces despair, and the journey to destruction is sealed”
(<a href="#morrison2007bluest">Morrison x</a>).</p>
<p>This world that “in its language, laws, and images, re-enforces
despair,” as Toni Morrison put it in the foreword of a 2007 print of her
book, is introduced right from the start when the MacTeers have to
collect “tiny pieces of coal” that have fallen off the coal
trains—tedious work in the cold of winter—only to turn around and see the
steel mill where the trains transport their never ending abundances of coal.</p>
<p>Other sections specifically indicate racial discrimination on the job
market, in education, and centrally in the accepted ideal of physical
beauty: “To be put outdoors by a landlord was one thing—unfortunate, but
an aspect of life over which you had no control, since you could not
control your income” (<a href="#morrison1999bluest">Morrison 11</a>) or “They go to
land-grant colleges, normal schools, and learn how to do the white man’s
work with refinement: home economics to prepare his food; teacher
education to instruct black children in obedience; music to soothe the
weary master and entertain his blunted soul”
(<a href="#morrison1999bluest">64</a>) or “Along with the idea of romantic
love, she was introduced to another—physical beauty. Probably the most
destructive ideas in the history of human thought. Both originated in
envy, thrived in insecurity, and ended in disillusion”
(<a href="#morrison1999bluest">95</a>) or “The men studied medicine, law,
theology, and emerged repeatedly in the powerless government offices
available to the native population” (<a href="#morrison1999bluest">133</a>)
or “… among all the waste and beauty of the world—which is what she
herself was. All of our waste which we dumped on her and which she
absorbed” (<a href="#morrison1999bluest">163</a>).</p>
<p>Especially the imposed ideal of beauty, for which you first had to be
white and blue eyed, takes center stage and is viewed there from at
least two perspectives, that of the Breedloves who accept it and are
devastated by it, and Claudia, who from her early childhood onward wages
war against it and tries to dismantle it—even literally at first and out
of a motivation to find and understand it.<sup id="fnref:17"><a class="footnote-ref" href="#fn:17">17</a></sup></p>
<h2 id="the-graph"><a class="toclink" href="#the-graph">The Graph</a></h2>
<p>In the graph below, individuals are indicated by oval vertices, groups
of people or abstract entities by boxes. Every solid directed edge
represents an aggression, while dashed edges represent hypothetical
aggressions, which, of course, do not count as actual aggressions.</p>
<p>The graph is arranged using the <em>dot</em> algorithm, which has the property
that it arranges the vertices in distinct ranks and minimizes edge
lengths, which implicitly leads to vertices with more incoming edges
tending toward the right and those with fewer tending toward the left:
“An optimal rank assignment assigns integer ranks to nodes such that the
sum of the ‘costs’ of edges is minimized. The ‘cost’ of an edge is the
product of its weight and its length, where the length is the rank of
its head minus the rank of its tail” (<a href="#gansner1990technique">Gansner, North, and Vo 4</a>).</p>
<p>This property is useful for assessing the overall “flow” of aggressions.
In view of my hypothesis that aggressions are often handed down from the
older generation to the younger, for example, the graph structure makes
clear that I could not find a single aggression of a member of a younger
generation on a member of an older one. Neither could I find any of a
person of color on an individual white person. Furthermore, we can see
that Pecola Breedlove has only incoming edges.</p>
<p>These observations are in line with the strategy of the author as
explained in her 2007 foreword: “The project, then, for this, my first
book, was to enter the life of the one least likely to withstand such
damaging forces because of youth, gender, and race” and “The main
character could not stand alone since her passivity made her a narrative
void” (<a href="#morrison2007bluest">Morrison x</a>).</p>
<p>However, to establish that there is at least in many cases a
cause-effect relationship between the aggressions of the oldest on the
younger and the aggressions of the younger on the youngest, the novel
proper needs to be scrutinized again.</p>
<p>Firstly, there are two sentences that describe the process very
succinctly: “When white men beat their men, they cleaned up the blood
and went home to receive abuse from the victim. They beat their children
with one hand and stole for them with the other”
(<a href="#morrison1999bluest">Morrison 108</a>).</p>
<p>Secondly, Toni Morrison also explains in her foreword that she “mounted
a series of rejections, some routine, some exceptional, some monstrous,
all the while trying hard to avoid complicity in the demonization
process Pecola was subjected to. That is, [she] did not want to
dehumanize the characters who trashed Pecola and contributed to her
collapse.” To this end she did not just create Pecola’s parents as mere
plot devices—a victim of delusions of a religious calling and an alcohol
addict—but gave them extensive backstory that showed much of the <em>how</em>,
and thereby some of the <em>why</em>, behind their callous and atrocious
aggressions, especially against their children.</p>
<p>We are lead through both their childhoods and the early years of their
relationship. We experience, proximately, the traumatic turning points
of their lives. We even hear Paulina Breedlove’s voice and Charles
Breedlove’s thoughts.<sup id="fnref:18"><a class="footnote-ref" href="#fn:18">18</a></sup> This intimacy builds rapport and lets us
empathize with them, so while we will still condemn their deeds, we can
no longer “demonize” them, as Toni Morrison put it, and hence can no
longer indulge in the complacent belief of having isolated the root
cause of Pecola’s descent into madness.</p>
<p>What happened, then, is that we got a glimpse of the <em>why</em> behind their
aggressive behaviors, a <em>why</em> that itself is born of aggressions and
their resulting traumata. Hence, it becomes clear that the destructive
force that caused Pecola to “literally fall apart”
(<a href="#morrison2007bluest">Morrison xii</a>), exacerbated by many of the same
aggressions that already beleaguered her parents, was yet the aggregate
of generations of oppression.</p>
<p><img alt="Graph of aggressions in The Bluest Eye" title="Graph of aggressions in The Bluest Eye" src="/images/mapping-directions-of-aggressions-in-the-bluest-eye/aggressions-in-the-bluest-eye.svg" width="100%" /></p>
<h2 id="appendix-table-of-aggressions"><a class="toclink" href="#appendix-table-of-aggressions">Appendix: Table of Aggressions</a></h2>
<p>Below the table of the aggressions in <em>The Bluest Eye</em>. For typographic
reasons, I had to resort of abbreviating the types of aggressions and
their forms of representation in the text. The abbreviations are unique
per column.</p>
<div> </div>
<p><strong>Representation</strong></p>
<ul>
<li><strong>hyp.</strong>: Hypothetical</li>
<li><strong>i.a.</strong>: In absentia</li>
</ul>
<p><strong>Buss</strong></p>
<ul>
<li><strong>phy., ver.</strong>: Physical, verbal</li>
<li><strong>act., pas.</strong>: Active, passive</li>
<li><strong>dir., ind.</strong>: Direct, indirect</li>
</ul>
<p><strong>Ramirez</strong></p>
<ul>
<li><strong>hos., ins.</strong>: Hostile, instrumental</li>
<li><strong>imp., pre.</strong>: Impulsive, premeditate</li>
<li><strong>pro., rea.</strong>: Proactive, reactive</li>
</ul>
<p><strong>Sue</strong></p>
<ul>
<li><strong>ass.</strong>: Microassault</li>
<li><strong>ins.</strong>: Microinsult</li>
<li><strong>inv.</strong>: Microinvalidation</li>
</ul>
<p>Furthermore I chose to abbreviate “composite” as “com.” and “see above”
as “v.s.” (<em>vide supra</em>).</p>
<table>
<thead>
<tr>
<th>Page</th>
<th>Perpetrator</th>
<th>Victim</th>
<th>Repr.</th>
<th>Buss</th>
<th>Ramirez</th>
<th>Sue</th>
<th>Comment</th>
</tr>
</thead>
<tbody>
<tr>
<td>5</td>
<td>Rosemary Villanucci</td>
<td>Claudia and Frieda MacTeer</td>
<td></td>
<td>ver., act., dir.</td>
<td>hos., pro.</td>
<td>ass.</td>
<td>Rosemary Villanucci initiates a conversation with Claudia and Frieda MacTeer to tell them that they cannot enter her family’s car.</td>
</tr>
<tr>
<td>5</td>
<td>Claudia and Frieda MacTeer</td>
<td>Rosemary Villanucci</td>
<td>hyp.</td>
<td>phy., act., dir.</td>
<td>hos., pre., rea.</td>
<td></td>
<td>Claudia and Frieda MacTeer fantasize about physical retribution on Rosemary Villanucci.</td>
</tr>
<tr>
<td>5</td>
<td>Society</td>
<td>MacTeer family</td>
<td></td>
<td>com.</td>
<td>com.</td>
<td>com.</td>
<td>The MacTeer family has to collect coal scraps in view of abundance.</td>
</tr>
<tr>
<td>5</td>
<td>Adults</td>
<td>Claudia and Frieda MacTeer</td>
<td></td>
<td>ver., act., dir.</td>
<td>imp., rea.</td>
<td>ins.</td>
<td>Parents blame children for accidents and illnesses.</td>
</tr>
<tr>
<td>6</td>
<td>Mrs. and Mr. MacTeer</td>
<td>Claudia MacTeer</td>
<td></td>
<td>v.s.</td>
<td>v.s.</td>
<td>v.s.</td>
<td>Examples of the above.</td>
</tr>
<tr>
<td>7</td>
<td>Mrs. MacTeer</td>
<td>Claudia MacTeer</td>
<td></td>
<td>v.s.</td>
<td>v.s.</td>
<td>v.s.</td>
<td>The above explained in terms of synechdochical blame.</td>
</tr>
<tr>
<td>8</td>
<td>Unnamed person</td>
<td>Peggy</td>
<td>i.a.</td>
<td>ver., act.</td>
<td>imp., pro.</td>
<td>ass.</td>
<td>The conversation reveals a deep-seated disdain of someone called Peggy.</td>
</tr>
<tr>
<td>8</td>
<td>Society</td>
<td>Women</td>
<td></td>
<td>com.</td>
<td>com.</td>
<td>com.</td>
<td>The conversation reveals a societal pressure to wear perfume, a possible indication of self-loathing.</td>
</tr>
<tr>
<td>8</td>
<td>Unnamed person</td>
<td>Aunt Julia</td>
<td>i.a.</td>
<td>ver., act.</td>
<td>hos., imp., rea.</td>
<td>ass.</td>
<td>One of the interlocutors pretends that someone “floating by” harmed them.</td>
</tr>
<tr>
<td>9</td>
<td>Unnamed person</td>
<td>Bella’s sister</td>
<td>i.a.</td>
<td>ver., act.</td>
<td>rea.</td>
<td>ass.</td>
<td>An interlocutor raises a suspicion of self-serving motives merely due to someone’s prolonged absence from family members.</td>
</tr>
<tr>
<td>10</td>
<td>Mrs. and Mr. MacTeer</td>
<td>Claudia and Frieda MacTeer</td>
<td></td>
<td>ver., act., dir.</td>
<td>imp., pro.</td>
<td>ins.</td>
<td>Claudia MacTeer’s parents treat their children like inventory.</td>
</tr>
<tr>
<td>11</td>
<td>Charles Breedlove</td>
<td>Wife and children</td>
<td></td>
<td>phy., act., dir.</td>
<td>hos., imp.</td>
<td></td>
<td>Description of Charles Breedlove’s violence toward his family.</td>
</tr>
<tr>
<td>11</td>
<td>Abstract mother</td>
<td>Abstract son</td>
<td></td>
<td>com.</td>
<td>com.</td>
<td>com.</td>
<td>Aggression by mother on son used as abstract example.</td>
</tr>
<tr>
<td>11</td>
<td>Society</td>
<td>Claudia MacTeer’s community</td>
<td></td>
<td>com.</td>
<td>com.</td>
<td>com.</td>
<td>A comment that indicates discrimination on the job market.</td>
</tr>
<tr>
<td>12</td>
<td>Society</td>
<td>Claudia MacTeer’s community</td>
<td></td>
<td>com.</td>
<td>com.</td>
<td>com.</td>
<td>Threat of homelessness.</td>
</tr>
<tr>
<td>14</td>
<td>Society</td>
<td>Claudia MacTeer</td>
<td></td>
<td>com.</td>
<td>com.</td>
<td>com.</td>
<td>The doll, symbolic of the oppressive ideals of beauty, is described as “patently aggressive sleeping companion..”</td>
</tr>
<tr>
<td>14</td>
<td>Adults</td>
<td>Claudia MacTeer</td>
<td></td>
<td>pas., dir.</td>
<td></td>
<td>ins.</td>
<td>No adult ever asked her what she wanted for Christmas but were content to guess for her.</td>
</tr>
<tr>
<td>15</td>
<td>Claudia MacTeer</td>
<td>White girls</td>
<td>hyp.</td>
<td>phy., act., dir.</td>
<td>pre., rea.</td>
<td></td>
<td>Claudia MacTeer imagines cruelty toward “little white girls..”</td>
</tr>
<tr>
<td>15</td>
<td>People</td>
<td>Claudia MacTeer</td>
<td></td>
<td>ver., pas., dir.</td>
<td></td>
<td>ins.</td>
<td>Claudia MacTeer resents never being regarded as, maybe, adorable.</td>
</tr>
<tr>
<td>16</td>
<td>Mrs. MacTeer</td>
<td>Pecola Breedlove et al.</td>
<td></td>
<td>ver., act., dir.</td>
<td>imp., rea.</td>
<td>ass.</td>
<td>Mrs. MacTeer engages in an extensive tirade against Pecola over her consumption of great amounts of milk.</td>
</tr>
<tr>
<td>18</td>
<td>“My man”</td>
<td>Narrator</td>
<td></td>
<td>phy., pas., dir.</td>
<td></td>
<td></td>
<td>A blues song provides a lyrical description of the abandonment of a woman by her boyfriend or husband.</td>
</tr>
<tr>
<td>21</td>
<td>Frieda MacTeer</td>
<td>Claudia MacTeer</td>
<td></td>
<td>ver., act., dir.</td>
<td>hos., imp.</td>
<td>ass.</td>
<td>Frieda MacTeer insults her sister.</td>
</tr>
<tr>
<td>21</td>
<td>Rosemary Villanucci</td>
<td>Claudia and Frieda MacTeer, Pecola Breedlove</td>
<td></td>
<td>ver., act., ind.</td>
<td>ins.</td>
<td></td>
<td>Rosemary Villanucci exploits a situation possibly to gain recognition from the adults, to harm Claudia and Frieda MacTeer, or to feel confirmed in her possibly unfavorable opinion about blacks generally.</td>
</tr>
<tr>
<td>22</td>
<td>Mrs. MacTeer</td>
<td>Frieda MacTeer</td>
<td></td>
<td>phy., act., dir.</td>
<td>imp.</td>
<td></td>
<td>Mrs. MacTeer beats Frieda.</td>
</tr>
<tr>
<td>23</td>
<td>Frieda MacTeer</td>
<td>Claudia MacTeer</td>
<td></td>
<td>ver., act., dir.</td>
<td>hos., imp.</td>
<td>ass.</td>
<td>Another slight slight by Frieda against her sister.</td>
</tr>
<tr>
<td>28</td>
<td>Society</td>
<td>Breedloves</td>
<td></td>
<td>com.</td>
<td>com.</td>
<td>com.</td>
<td>The white-dominated culture confirms the Breedloves in their belief in their own ugliness.</td>
</tr>
<tr>
<td>29</td>
<td>Paulina and Charles Breedlove</td>
<td>Paulina and Charles Breedlove</td>
<td></td>
<td>phy., act., dir.</td>
<td>rea.</td>
<td></td>
<td>A fight between the Breedlove parents.</td>
</tr>
<tr>
<td>29</td>
<td>Paulina and Charles Breedlove</td>
<td>Pecola and Sammy Breedlove</td>
<td></td>
<td>act., dir.</td>
<td></td>
<td></td>
<td>The fight is also aggressive towards their children.</td>
</tr>
<tr>
<td>31</td>
<td>White hunters</td>
<td>Charles Breedlove, Darlene</td>
<td></td>
<td>phy., act., dir.</td>
<td>hos., imp., pro.</td>
<td></td>
<td>A description of Charles Breedlove’s trauma.</td>
</tr>
<tr>
<td>34</td>
<td>Any girl at school</td>
<td>Bobby, Pecola Breedlove</td>
<td></td>
<td>ver., act., dir.</td>
<td></td>
<td></td>
<td>Teasing of “Bobby” (or anyone) as an example of mean-spirited behavior between children other than Pecola that is yet abusive toward her.</td>
</tr>
<tr>
<td>36</td>
<td>Mr. Yacobowski</td>
<td>Pecola Breedlove</td>
<td></td>
<td>dir.</td>
<td>hos., imp.</td>
<td>ins.</td>
<td>A complex experience reveals a very hostile kind of disrespect that Pecola is exposed to.</td>
</tr>
<tr>
<td>39</td>
<td>China</td>
<td>Marie</td>
<td></td>
<td>ver., act., dir.</td>
<td>imp., pro.</td>
<td>ass.</td>
<td>China fat-shaming Marie without hurtful intent.</td>
</tr>
<tr>
<td>40</td>
<td>John Dillinger</td>
<td>Random people</td>
<td>hyp.</td>
<td>phy., act., dir.</td>
<td>ins., pre.</td>
<td></td>
<td>A description of a violent bank robbery.</td>
</tr>
<tr>
<td>41</td>
<td>China</td>
<td>Marie</td>
<td></td>
<td>ver., act., dir.</td>
<td>imp., pro.</td>
<td>ass.</td>
<td>Another sizeist comment from China in conjunction with a probably jokular racist aspect.</td>
</tr>
<tr>
<td>43</td>
<td>China, Marie, and Poland</td>
<td>Men</td>
<td></td>
<td>com.</td>
<td>hos.</td>
<td></td>
<td>Several descriptions of aggressions against men, at one point referred to as “vengence..”</td>
</tr>
<tr>
<td>48</td>
<td>Claudia and Frieda MacTeer</td>
<td>Maureen Peal</td>
<td></td>
<td>ver., act., dir.</td>
<td>hos., rea.</td>
<td>ass.</td>
<td>The sisters tease Maureen Peal.</td>
</tr>
<tr>
<td>49</td>
<td>Claudia MacTeer</td>
<td>Maureen Peal</td>
<td></td>
<td>ver., dir.</td>
<td>hos., imp., rea.</td>
<td>ins.</td>
<td>Claudia MacTeer counters an invitation from Maureen Peal discourteously while not turning it down.</td>
</tr>
<tr>
<td>49</td>
<td>Claudia MacTeer</td>
<td>Maureen Peal</td>
<td>hyp.</td>
<td>phy., act., dir.</td>
<td>hos., pre., rea.</td>
<td></td>
<td>Claudia MacTeer fantasizes about damaging Maureen’s fur muff.</td>
</tr>
<tr>
<td>50</td>
<td>Bunch of boys</td>
<td>Pecola Breedlove</td>
<td></td>
<td>ver., act., dir.</td>
<td>pro.</td>
<td>ass.</td>
<td>Several boys mob and insult Pecola Breedlove.</td>
</tr>
<tr>
<td>50</td>
<td>Frieda MacTeer</td>
<td>Woodrow</td>
<td></td>
<td>phy., act., dir.</td>
<td>pre., rea.</td>
<td></td>
<td>Frieda MacTeer hits Woodrow with a book.</td>
</tr>
<tr>
<td>51</td>
<td>Claudia MacTeer, Bay Boy</td>
<td>Bay Boy, Claudia MacTeer</td>
<td></td>
<td>ver., act., dir.</td>
<td>imp., rea.</td>
<td></td>
<td>Claudia and Bay Boy insult each other.</td>
</tr>
<tr>
<td>53</td>
<td>Maureen Peal</td>
<td>Claudia and Frieda MacTeer</td>
<td></td>
<td>ver., pas., dir.</td>
<td></td>
<td>ins.</td>
<td>Claudia and Frieda MacTeer are disappointed over not getting to eat ice cream.</td>
</tr>
<tr>
<td>53</td>
<td>Society</td>
<td>Claudia and Frieda MacTeer</td>
<td></td>
<td>com.</td>
<td>com.</td>
<td>com.</td>
<td>At a movie theater they see a picture of a white actress smiling down at them.</td>
</tr>
<tr>
<td>55</td>
<td>Claudia MacTeer</td>
<td>Maureen Peal</td>
<td></td>
<td>ver., act., dir.</td>
<td>imp.</td>
<td>ass.</td>
<td>Claudia MacTeer employs one of her stock insults against Maureen Peal.</td>
</tr>
<tr>
<td>56</td>
<td>Claudia and Frieda MacTeer, Maureen</td>
<td>Claudia and Frieda MacTeer, Maureen, Pecola Breedlove</td>
<td></td>
<td>ver. and phy., act., dir.</td>
<td>hos., imp., rea.</td>
<td></td>
<td>A wild, nonsensical quarrel ensues, and they again use “black” as insult.</td>
</tr>
<tr>
<td>57</td>
<td>Claudia and Frieda MacTeer</td>
<td>Pecola Breedlove</td>
<td></td>
<td>ver., pas., dir.</td>
<td></td>
<td></td>
<td>Yet the sisters are irritated by Pecola’s self-loathing and fail to comfort her.</td>
</tr>
<tr>
<td>59</td>
<td>Claudia and Frieda MacTeer</td>
<td>Rosemary Villanucci</td>
<td>hyp.</td>
<td>pas., dir.</td>
<td>hos., pre., rea.</td>
<td>ins.</td>
<td>Claudia and Frieda MacTeer plan on making Rosemary Villanucci jealous.</td>
</tr>
<tr>
<td>60</td>
<td>People</td>
<td>Marie</td>
<td></td>
<td>ver., act.</td>
<td></td>
<td>ass.</td>
<td>A description of the slander the three prostitutes are exposed to.</td>
</tr>
<tr>
<td>61</td>
<td>Henry Washington</td>
<td>Claudia and Frieda MacTeer</td>
<td></td>
<td>ver., act., dir.</td>
<td>ins.</td>
<td></td>
<td>Henry Washington lies to the sisters.</td>
</tr>
<tr>
<td>64</td>
<td>Society</td>
<td>Certain type of women</td>
<td></td>
<td>com.</td>
<td>com.</td>
<td>com.</td>
<td>A description of the established educational obedience “brainwashing” Geraldine and others went through.</td>
</tr>
<tr>
<td>65</td>
<td>Certain type of women</td>
<td>Husbands</td>
<td></td>
<td>pas., dir.</td>
<td>pro.</td>
<td></td>
<td>Denial of love for their husbands.</td>
</tr>
<tr>
<td>67</td>
<td>Geraldine</td>
<td>Blacks, Junior</td>
<td></td>
<td>ver., act., dir./ind.</td>
<td>pro., pre.</td>
<td>ass.</td>
<td>Geraldine distinguishes between colored people and “niggers..”</td>
</tr>
<tr>
<td>70</td>
<td>Junior</td>
<td>Pecola Breedlove, cat</td>
<td></td>
<td>phy./ver., act., dir./ind.</td>
<td>hos., pre.</td>
<td></td>
<td>Junior devises a complex stratagem for hurting Pecola and his mother’s cat at the same time, possibly killing the latter.</td>
</tr>
<tr>
<td>72</td>
<td>Geraldine</td>
<td>Pecola Breedlove</td>
<td></td>
<td>ver., act., dir.</td>
<td>hos., rea.</td>
<td>ass.</td>
<td>Geraldine insults Pecola and leaves her thinking she believed</td>
</tr>
<tr>
<td>Junior.</td>
<td></td>
<td></td>
<td></td>
<td></td>
<td></td>
<td></td>
<td></td>
</tr>
<tr>
<td>75</td>
<td>Parents</td>
<td>Claudia and Frieda MacTeer</td>
<td></td>
<td>phy., act., dir.</td>
<td>rea.</td>
<td></td>
<td>An abstract description of the beating of children.</td>
</tr>
<tr>
<td>76</td>
<td>Elihue Micah Whitcomb</td>
<td>Young girls</td>
<td></td>
<td>phy., act., dir.</td>
<td>pre., pro.</td>
<td></td>
<td>The narrator apparently already knew of Elihue Micah Whitcomb’s crimes</td>
</tr>
<tr>
<td>76</td>
<td>Henry Washington</td>
<td>Frieda MacTeer</td>
<td></td>
<td>phy., act., dir.</td>
<td>pro.</td>
<td></td>
<td>Henry Washington’s transgression on Frieda</td>
</tr>
<tr>
<td>77</td>
<td>Parents</td>
<td>Henry Washington</td>
<td></td>
<td>phy., act., dir.</td>
<td>rea.</td>
<td></td>
<td>The MacTeer parents’ retaliation against Henry Washington</td>
</tr>
<tr>
<td>77</td>
<td>Frieda MacTeer</td>
<td>Rosemary Villanucci</td>
<td></td>
<td>phy., act., dir.</td>
<td>rea.</td>
<td></td>
<td>A description of Frieda’s actions against Rosemary Villanucci.</td>
</tr>
<tr>
<td>80</td>
<td>Marie</td>
<td>Claudia and Frieda MacTeer</td>
<td></td>
<td>phy., act., dir.</td>
<td>hos., imp., rea.</td>
<td></td>
<td>Marie throws a bottle more or less at the sisters.</td>
</tr>
<tr>
<td>82</td>
<td>Society</td>
<td>Blacks</td>
<td></td>
<td>com.</td>
<td>com.</td>
<td>com.</td>
<td>An explicit mention of segregation.</td>
</tr>
<tr>
<td>83</td>
<td>Frieda MacTeer</td>
<td>Pecola Breedlove</td>
<td></td>
<td>ver., act., dir.</td>
<td>imp., rea.</td>
<td>inv.</td>
<td>Frieda expresses lacking trust in Pecola.</td>
</tr>
<tr>
<td>84</td>
<td>Paulina Breedlove</td>
<td>Pecola Breedlove (et al.)</td>
<td></td>
<td>ver., pas., ind.</td>
<td></td>
<td>ins.</td>
<td>A girl calls Mrs. Breedlove “Polly” in front of Pecola.</td>
</tr>
<tr>
<td>84</td>
<td>Paulina Breedlove</td>
<td>Pecola Breedlove (et al.)</td>
<td></td>
<td>ver., act., dir.</td>
<td>hos., imp., rea.</td>
<td>ass.</td>
<td>Mrs. Breedlove insults Pecola.</td>
</tr>
<tr>
<td>85</td>
<td>Paulina Breedlove</td>
<td>Pecola Breedlove (et al.)</td>
<td></td>
<td>ver., pas., dir.</td>
<td>ins.</td>
<td>ins.</td>
<td>Mrs. Breedlove fails to claim motherhood of Pecola.</td>
</tr>
<tr>
<td>86</td>
<td>Her family</td>
<td>Paulina Breedlove</td>
<td></td>
<td>ver., pas., dir.</td>
<td></td>
<td>ins.</td>
<td>Mrs. Breedlove’s foot injury is met with indifference.</td>
</tr>
<tr>
<td>86</td>
<td>Her community</td>
<td>Paulina Breedlove</td>
<td></td>
<td>pas., com.</td>
<td></td>
<td>ins.</td>
<td>Mrs. Breedlove was left out of social activities.</td>
</tr>
<tr>
<td>87</td>
<td>Her community</td>
<td>Paulina Breedlove</td>
<td></td>
<td>pas., com.</td>
<td></td>
<td>ins.</td>
<td>Creative outlets were not offered to her as a child.</td>
</tr>
<tr>
<td>92</td>
<td>Her new community</td>
<td>Paulina Breedlove</td>
<td></td>
<td>ver., act., dir.</td>
<td>imp., rea.</td>
<td>ins.</td>
<td>Heckling between black women.</td>
</tr>
<tr>
<td>93</td>
<td>“Family of slender means”</td>
<td>Paulina Breedlove</td>
<td></td>
<td>pas., dir.</td>
<td>pro.</td>
<td>ins.</td>
<td>Mrs. Breedlove has to work for a family that takes their own problems too seriously.</td>
</tr>
<tr>
<td>93</td>
<td>“Family of slender means”</td>
<td>Paulina and Charles Breedlove</td>
<td></td>
<td>ver., act., dir.</td>
<td>ins., pre., rea.</td>
<td></td>
<td>A conflict ensues over Charles Breedlove turning up at their house.</td>
</tr>
<tr>
<td>95</td>
<td>Society</td>
<td>Paulina Breedlove</td>
<td></td>
<td>com.</td>
<td>com.</td>
<td>com.</td>
<td>Mrs. Breedlove is confronted with what is regarded as physical beauty.</td>
</tr>
<tr>
<td>96</td>
<td>Paulina and Charles Breedlove</td>
<td>Paulina and Charles Breedlove</td>
<td></td>
<td>phy., act., dir.</td>
<td></td>
<td></td>
<td>Fights between Paulina and Charles Breedlove</td>
</tr>
<tr>
<td>97</td>
<td>Doctor</td>
<td>Paulina Breedlove</td>
<td></td>
<td>ver., act., ind.</td>
<td></td>
<td>ins.</td>
<td>Insulting comments from a racist doctor.</td>
</tr>
<tr>
<td>98</td>
<td>Paulina Breedlove</td>
<td>Pecola Breedlove</td>
<td></td>
<td>act., dir.</td>
<td></td>
<td>com.</td>
<td>Paulina Breedlove has perceived Pecola as ugly from birth, which Pecola surely picked up on.</td>
</tr>
<tr>
<td>98</td>
<td>Paulina Breedlove</td>
<td>Charles Breedlove</td>
<td></td>
<td>act., ind.</td>
<td></td>
<td></td>
<td>Mrs. Breedlove engages in vengence through subtle force on Charles Breedlove.</td>
</tr>
<tr>
<td>99</td>
<td>Paulina Breedlove</td>
<td>Rest of the family</td>
<td></td>
<td>pas., dir.</td>
<td></td>
<td></td>
<td>Mrs. Breedlove neglects her family.</td>
</tr>
<tr>
<td>100</td>
<td>Paulina Breedlove</td>
<td>Children</td>
<td></td>
<td>phy., act., dir.</td>
<td></td>
<td></td>
<td>Mrs. Breedlove resorts to beating and disproportionate punishment.</td>
</tr>
<tr>
<td>100</td>
<td>Paulina Breedlove</td>
<td>Charles Breedlove</td>
<td></td>
<td>ver., act., ind.</td>
<td></td>
<td></td>
<td>Mrs. Breedlove points out her husband’s faults, but not to help him but to use him as negative example for her children.</td>
</tr>
<tr>
<td>102</td>
<td>Charles Breedlove</td>
<td>Paulina Breedlove</td>
<td></td>
<td>phy., act., dir.</td>
<td>imp., pro.</td>
<td></td>
<td>Practice by Charles Breedlove that may constitute rape depending on Paulina’s perception of it.</td>
</tr>
<tr>
<td>103</td>
<td>Mother</td>
<td>Charles Breedlove</td>
<td></td>
<td>phy., act., dir.</td>
<td>ins., pre.</td>
<td></td>
<td>Charles Breedlove’s mother places him on junk heap as baby.</td>
</tr>
<tr>
<td>103</td>
<td>Aunt Jimmy</td>
<td>Mother</td>
<td></td>
<td>phy., act., dir.</td>
<td>rea.</td>
<td></td>
<td>Mr. Breedlove’s aunt beats his mother.</td>
</tr>
<tr>
<td>103</td>
<td>Aunt Jimmy</td>
<td>Charles Breedlove</td>
<td></td>
<td>phy., pas., dir.</td>
<td></td>
<td></td>
<td>Mr. Breedlove’s aunt exposes him to noxious odors.</td>
</tr>
<tr>
<td>103</td>
<td>Aunt Jimmy</td>
<td>Charles Breedlove</td>
<td></td>
<td>ver., act., dir.</td>
<td>imp., rea.</td>
<td>ass.</td>
<td>Mr. Breedlove’s aunt calls him “foolish..”</td>
</tr>
<tr>
<td>104</td>
<td>Aunt Jimmy</td>
<td>Charles Breedlove (mother and father i.a.)</td>
<td></td>
<td>ver., act., ind.</td>
<td>rea.</td>
<td>ass., ins.</td>
<td>Mr. Breedlove’s aunt shames his heritage by constantly insulting his mother and father.</td>
</tr>
<tr>
<td>108</td>
<td>Everyone except Black women and children</td>
<td>Black women</td>
<td></td>
<td>com.</td>
<td>com.</td>
<td>com.</td>
<td>An abstract description of women being commanded around.</td>
</tr>
<tr>
<td>108</td>
<td>Black men</td>
<td>Black women</td>
<td></td>
<td>act., dir.</td>
<td></td>
<td></td>
<td>An explicit description of aggressions being handed on.</td>
</tr>
<tr>
<td>108</td>
<td>Men, probably</td>
<td>Black women</td>
<td></td>
<td>act., dir.</td>
<td></td>
<td></td>
<td>Implication of molestation.</td>
</tr>
<tr>
<td>113</td>
<td>Girls</td>
<td>Girls</td>
<td></td>
<td>ver., act., dir.</td>
<td></td>
<td>ass.</td>
<td>Girls engage in “a serious kind of making fun..”</td>
</tr>
<tr>
<td>114</td>
<td>Charles Breedlove, Jake</td>
<td>Darlene, Suky</td>
<td></td>
<td>act., dir.</td>
<td>imp., pro.</td>
<td>ass.</td>
<td>They two boys throw grapes at the girls.</td>
</tr>
<tr>
<td>116</td>
<td>White hunters</td>
<td>Charles Breedlove, Darlene</td>
<td></td>
<td>phy., act., dir.</td>
<td>imp., pro.</td>
<td></td>
<td>Charles Breedlove and Darlene are forced at gunpoint to have sex in front of white hunters.</td>
</tr>
<tr>
<td>120</td>
<td>Conductor</td>
<td>Charles Breedlove</td>
<td></td>
<td>ver., act., dir.</td>
<td>hos., imp., pro.</td>
<td></td>
<td>Insults from a conductor are used as exmaple for a descrition of them similar to forces of nature.</td>
</tr>
<tr>
<td>122</td>
<td>Aunt Jimmy</td>
<td>Charles Breedlove (mother i.a.)</td>
<td></td>
<td>ver., pas., dir.</td>
<td>ins., pre., rea.</td>
<td></td>
<td>Mr. Breedlove’s aunt further shames his mother and heritage by never telling him her name.</td>
</tr>
<tr>
<td>123</td>
<td>Samson Fuller</td>
<td>Charles Breedlove</td>
<td></td>
<td>ver., act., dir.</td>
<td>hos., imp., rea.</td>
<td>ass.</td>
<td>Mr. Breedlove’s father behaves rude and insultingly toward him.</td>
</tr>
<tr>
<td>125</td>
<td>Charles Breedlove</td>
<td>Anyone under his authority</td>
<td>hyp.</td>
<td>com.</td>
<td>com.</td>
<td>com.</td>
<td>Abstract description making mention of various aggressions.</td>
</tr>
<tr>
<td>128</td>
<td>Charles Breedlove</td>
<td>Pecola Breedlove</td>
<td></td>
<td>phy., act., dir.</td>
<td>imp., pro.</td>
<td></td>
<td>Mr. Breedlove rapes his daughter.</td>
</tr>
<tr>
<td>132</td>
<td>Whitcombs</td>
<td>Themselves, each other</td>
<td></td>
<td>act., dir.</td>
<td></td>
<td>com.</td>
<td>The Whitcombs are afflicted with some sort of partial self-loathing of their native roots.</td>
</tr>
<tr>
<td>133</td>
<td>Society</td>
<td>Natives</td>
<td></td>
<td>com.</td>
<td>com.</td>
<td>com.</td>
<td>It is stated that only powerless government offices are available to the native population.</td>
</tr>
<tr>
<td>133</td>
<td>Whitcombs</td>
<td>Anyone</td>
<td></td>
<td>com.</td>
<td>com.</td>
<td>com.</td>
<td>Those of the family in government offices are described as corrupt and lecherous.</td>
</tr>
<tr>
<td>134</td>
<td>Elihue Micah Whitcomb’s father</td>
<td>Elihue Micah Whitcomb</td>
<td></td>
<td>phy., act., dir.</td>
<td>ins.</td>
<td></td>
<td>Elihue Micah Whitcomb’s father uses violence for education and discipline.</td>
</tr>
<tr>
<td>135</td>
<td>Society</td>
<td>Blacks</td>
<td></td>
<td>com.</td>
<td>com.</td>
<td>com.</td>
<td>Few white-collar occupations are available to blacks.</td>
</tr>
<tr>
<td>136</td>
<td>Elihue Micah Whitcomb</td>
<td>Young girls</td>
<td></td>
<td>phy., act., dir.</td>
<td>pre., pro.</td>
<td></td>
<td>A description of molestation, possibly rape, of young girls.</td>
</tr>
<tr>
<td>138</td>
<td>Elihue Micah Whitcomb</td>
<td>Bob</td>
<td></td>
<td>phy., act., ind.</td>
<td>ins., pre., rea.</td>
<td></td>
<td>Elihue Micah Whitcomb has Bob, the dog, poisoned.</td>
</tr>
<tr>
<td>138</td>
<td>Elihue Micah Whitcomb</td>
<td>Pecola Breedlove</td>
<td></td>
<td>act., dir.</td>
<td>ins., pre., rea.</td>
<td></td>
<td>Elihue Micah Whitcomb tricks Pecola into performing the poisoning.</td>
</tr>
<tr>
<td>149</td>
<td>Random people</td>
<td>Pecola Breedlove</td>
<td>i.a.</td>
<td>ver., act.</td>
<td>pro.</td>
<td>ass.</td>
<td>An example of victim blaming.</td>
</tr>
<tr>
<td>149</td>
<td>Random people</td>
<td>Pecola Breedlove</td>
<td>i.a.</td>
<td>ver., act.</td>
<td>hos., pro.</td>
<td>ass.</td>
<td>People insult the Breedloves and the unborn child as ugly.</td>
</tr>
<tr>
<td>149</td>
<td>Random people</td>
<td>Pecola Breedlove</td>
<td></td>
<td>pas., dir.</td>
<td></td>
<td></td>
<td>The community shows a complete lack of empathy.</td>
</tr>
<tr>
<td>154</td>
<td>Paulina Breedlove</td>
<td>Pecola Breedlove</td>
<td></td>
<td>pas., dir.</td>
<td></td>
<td>ins.</td>
<td>Mrs. Breedlove is said to avoid looking at Pecola.</td>
</tr>
<tr>
<td>158</td>
<td>Charles Breedlove</td>
<td>Pecola Breedlove</td>
<td></td>
<td>phy., act., dir.</td>
<td>pro.</td>
<td></td>
<td>Mr. Breedlove’s second rape of his daughter is revealed.</td>
</tr>
<tr>
<td>162</td>
<td>Random people</td>
<td>Pecola Breedlove</td>
<td></td>
<td>com.</td>
<td>com.</td>
<td>com.</td>
<td>People avoid Pecola or laugh at her.</td>
</tr>
<tr>
<td>163</td>
<td>Society</td>
<td>Pecola Breedlove</td>
<td></td>
<td>com.</td>
<td>com.</td>
<td>com.</td>
<td>A description of figurative waste having been dumped on Pecola Breedlove.</td>
</tr>
<tr>
<td>164</td>
<td>Society</td>
<td>Pecola Breedlove</td>
<td>i.a.</td>
<td>ver., act.</td>
<td></td>
<td>ass.</td>
<td>An abstract form or victim blaming is exposed.</td>
</tr>
</tbody>
</table>
<h2 id="endnotes"><a class="toclink" href="#endnotes">Endnotes</a></h2>
<div class="footnote">
<hr>
<ol>
<li id="fn:1">
<p>A process the novel comments on
(<a href="#morrison1999bluest">Morrison 134</a>). <a class="footnote-backref" href="#fnref:1" title="Jump back to footnote 1 in the text">↩</a></p>
</li>
<li id="fn:2">
<p>Another factor to consider is Claudia MacTeer’s possible
unreliability as narrator. However, the implications, if any, are
not clear to me. The chapters that are narrated from a first person
perspective, moreover, are set in ragged right in the two editions I
have at my disposal, possibly to indicate the inchoate nature of a
child’s worldview. In these chapters, Claudia assumes an internal
focalization on her younger self, while she, or the narrator, seems
greatly more perspicacious and self-aware in the chapters set in
regular flush left and right. <a class="footnote-backref" href="#fnref:2" title="Jump back to footnote 2 in the text">↩</a></p>
</li>
<li id="fn:3">
<p>Just possibly, the act of consuming the Mary Janesa brand of
sweets with the “smiling white face” of a girl named Mary Jane on
them, with “blue eyes looking at [Pecola] out of a world of clean
comfort” ay be akin to a vorarephilic sexual experience for Pecola,
which would turn the act of buying them into an intensely intimate
endeavor to begin with. (Her milk binge from the Shirley Temple cup
might be interpreted in a similar vein.) This interpretation is
reinforced by the description of Pecola’s eating them as “nine
lovely orgasms with Mary Jane” (<a href="#morrison1999bluest">Morrison 50</a>). <a class="footnote-backref" href="#fnref:3" title="Jump back to footnote 3 in the text">↩</a></p>
</li>
<li id="fn:4">
<p>On the diagetic level, I will later introduce the <em>hypothetical</em>
aggression, which is merely imagined or discussed, not actualized. <a class="footnote-backref" href="#fnref:4" title="Jump back to footnote 4 in the text">↩</a></p>
</li>
<li id="fn:5">
<p>Luckily, the edge cases of who can be considered “living” will be
of limited importance for this paper. <a class="footnote-backref" href="#fnref:5" title="Jump back to footnote 5 in the text">↩</a></p>
</li>
<li id="fn:6">
<p>Cases where a guardian decides in favor of a surgical procedure
against the will of a child or a senile elderly person may
constitute an edge case but may also be covered by the vague
reference to “motivated to avoid.” <a class="footnote-backref" href="#fnref:6" title="Jump back to footnote 6 in the text">↩</a></p>
</li>
<li id="fn:7">
<p>To avoid the connotational baggage in cases like this, one might
opt for alternatives like “assertive” or “forthright.” <a class="footnote-backref" href="#fnref:7" title="Jump back to footnote 7 in the text">↩</a></p>
</li>
<li id="fn:8">
<p>This lack of appreciation, of course, was owed to their own role
in perpetuating this societal ideal that ran counter to Claudia’s
and, moreover, through the microaggression of never asking her for
her own wishes. <a class="footnote-backref" href="#fnref:8" title="Jump back to footnote 8 in the text">↩</a></p>
</li>
<li id="fn:9">
<p>Examples—borrowed from Baron and Richardson—for the indirect
case are “hiring an assassin to kill an enemy” (physical, active,
and indirect) or “failing to speak up in another person’s defense
when he or she is unfairly criticized” (verbal, passive, and
indirect). The other modes are fairly self-explanatory. <a class="footnote-backref" href="#fnref:9" title="Jump back to footnote 9 in the text">↩</a></p>
</li>
<li id="fn:10">
<p><em>Premeditate</em> as adjective seems to be a neologism along the
lines of such words as <em>aggregate</em> or <em>separate</em>, possibly coined to
avoid having to break the parallelism with “premeditated” as the
only participial adjective. “Premeditative” seems to apply more to
the agent than to the act. <a class="footnote-backref" href="#fnref:10" title="Jump back to footnote 10 in the text">↩</a></p>
</li>
<li id="fn:11">
<p>The distinction into hostile and instrumental aggression is a
controversial one, albeit widely adopted.
Bushman and Anderson (<a href="#bushman2001time">275</a>) make a strong case for its
limited applicability, and I will refrain from applying it in cases
that appear ambiguous. <a class="footnote-backref" href="#fnref:11" title="Jump back to footnote 11 in the text">↩</a></p>
</li>
<li id="fn:12">
<p>Henry, seemingly, has remained unscathed by this, possibly by
dint of his gender, so that Claudia’s mother was still “all ease and
satisfaction in discussing his coming.”
(<a href="#morrison1999bluest">Morrison 7</a>) <a class="footnote-backref" href="#fnref:12" title="Jump back to footnote 12 in the text">↩</a></p>
</li>
<li id="fn:13">
<p>One aggression where I decided to make an exception to this rule
was the scene where young Charles Breedlove and Jake throw grapes at
Suky and Darlene (<a href="#morrison1999bluest">Morrison 114</a>) because the
physical aspect of these small, light objects pelting their clothes
seemed negligible compared to the humiliation of being treated in
such a way, especially in view of the one-sided nature of the
playful fight. <a class="footnote-backref" href="#fnref:13" title="Jump back to footnote 13 in the text">↩</a></p>
</li>
<li id="fn:14">
<p>In addition, it was a very interesting experience to apply the
three categories, which have previously been used to describe racial
(<a href="#Sue2007">Sue et al.</a>; <a href="#Solorzano2000">Solorzano, Ceja, and Yosso</a>;
<a href="#pierce1977experiment">Pierce et al.</a>) and LGB microaggressions
(<a href="#Burn2005">Burn, Kadlec, and Rexer</a>; <a href="#Silverschanz2008">Silverschanz et al.</a>;
<a href="#Thurlow2001">Thurlow</a>; <a href="#nadal2011">Nadal et al.</a>), to
microaggressions by parents who use their authority irresponsibly at
times, a phenomenon that Toni Morrison may well have used as
extensively as she did in order to build an empathetic bridge for
readers who themselves suffer few oppressions and are not conscious
of the oppressions their class or ethnic group exacts. Most of them
will remember situations from their childhood where they have been
wronged and felt powerless against the aggressor. <a class="footnote-backref" href="#fnref:14" title="Jump back to footnote 14 in the text">↩</a></p>
</li>
<li id="fn:15">
<p>Another problem was that hostile aggression has to be “angry”
according to the definition, while in the book aggressions that
would intuitively fit that classification are often motivated by
hate or disgust rather than anger. Since this may well be just due
to imprecise wording in the definition, I chose to ignore this
distinction. <a class="footnote-backref" href="#fnref:15" title="Jump back to footnote 15 in the text">↩</a></p>
</li>
<li id="fn:16">
<p>The reason may be that she derives personal satisfaction from
seeing the two of them beaten by their mother or that ostensible
confirmation of their “dirty” (<a href="#morrison1999bluest">Morrison 21</a>)
nature accords with the “values” her parents instill in her. <a class="footnote-backref" href="#fnref:16" title="Jump back to footnote 16 in the text">↩</a></p>
</li>
<li id="fn:17">
<p>There is no beauty to this physical beauty for her, she merely
went “from pristine sadism to fabricated hatred, to fraudulent
love,” which were “adjustment[s] without improvement”
(<a href="#morrison1999bluest">Morrison 16</a>). Ergo her assessmentdisillusioned
itselfthat all romantic love and physical beauty “originated in
envy, thrived in insecurity, and ended in disillusion,” because such
was the modus operandi of all she ever got presented as “beauty” by
society. On the final pages, however, she acknowledges a new kind of
beauty, reminiscing about who Pecola was: “And all of our beauty,
which was hers first and which she gave to us”
(<a href="#morrison1999bluest">163</a>). <a class="footnote-backref" href="#fnref:17" title="Jump back to footnote 17 in the text">↩</a></p>
</li>
<li id="fn:18">
<p>The section from page 86 onward contains several blocks set in
italics and narrated by Paulina Breedlove in the first person. The
section from page 103 frequently reveals thoughts and feelings of
Charles Breedlove, a clear internal focalization. <a class="footnote-backref" href="#fnref:18" title="Jump back to footnote 18 in the text">↩</a></p>
</li>
</ol>
</div>
<h2 id="works-cited"><a class="toclink" href="#works-cited">Works Cited</a></h2>
<p><a name="baron2004human"></a>
Baron, Robert A. and Deborah R. Richardson. <em>Human Aggression</em>. Springer, 2004. Print. Perspectives in Social Psychology.</p>
<p><a name="Burn2005"></a>
Burn, Shawn M., Kelly Kadlec, and Ryan Rexer. “Effects of Subtle Heterosexism on Gays, Lesbians, Bisexuals.” <em>Journal of Homosexuality</em> 49.2 (2005): 23–38. Print.</p>
<p><a name="bushman2001time"></a>
Bushman, Brad J. and Craig A. Anderson. “Is it Time to Pull the Plug on the Hostile Versus Instrumental Aggression Dichotomy?.” <em>Psychological Review</em> 108.1 (2001): 273. Print.</p>
<p><a name="buss1961psychology"></a>
Buss, Arnold H. <em>The Psychology of Aggression</em>. Wiley, 1961. Print.</p>
<p><a name="gansner1990technique"></a>
Gansner, Emden R., Stephen C. North, and Kiem-Phong Vo. “Technique for Drawing Directed Graphs.” (Aug. 1990). <span class="caps">US</span> Patent 4,953,106. Print.</p>
<p><a name="GSbuss"></a>
Google Scholar. “Buss: The Psychology of Aggression. – Google Scholar.” <a href="http://scholar.google.com/scholar?as_sdt=1,5&hl=en&as_vis=1&cites=10933256569654438107">Web</a>. 30 Aug. 2013.</p>
<p><a name="griffin2004dark"></a>
Griffin, Ricky W., Anne O’Leary-Kelly, and Robert D. Pritchard. <em>The Dark Side of Organizational Behavior</em>. Wiley, 2004. Print. J-B <span class="caps">SIOP</span> Frontiers Series.</p>
<p><a name="huntsman2008parents"></a>
Huntsman, Leone. “Parents with Mental Health Issues: Consequences for Children and Effectiveness of Interventions Designed to Assist Children and their Families.” Pandora Electronic Collection (2008). Print. Pandora Electronic Collection.</p>
<p><a name="kogut2005identified"></a>
Kogut, Tehila and Ilana Ritov. “The ‘Identified Victim’ Effect: An Identified Group, or Just a Single Individual?.” <em>Journal of Behavioral Decision Making</em> 18.3 (2005): 157–167. Print.</p>
<p><a name="McKean2007"></a>
McKean, Erin. “Erin McKean: The Joy of Lexicography.” 2007. <a href="http://www.ted.com/talks/erin_mckean_redefines_the_dictionary.html">Web</a>. 30 Aug. 2013.</p>
<p><a name="morrison1999bluest"></a>
Morrison, Toni. <em>The Bluest Eye</em>. Random House Group Limited, 1999. Print.</p>
<p><a name="morrison2007bluest"></a>
―. <em>The Bluest Eye</em>. Knopf Doubleday Publishing Group, 2007. Print. Vintage International.</p>
<p><a name="nadal2011"></a>
Nadal, Kevin L., et al. “Sexual Orientation Microaggressions: ‘Death by a Thousand Cuts’ for Lesbian, Gay, and Bisexual Youth.” <em>Journal of <span class="caps">LGBT</span> Youth</em> 8.3 (2011): 234–259. Print.</p>
<p><a name="OEDaggression"></a>
Oxford English Dictionary. “aggression, n.” <a href="http://www.ted.com/talks/erin_mckean_redefines_the_dictionary.html">Web</a>. 30 Aug. 2013.</p>
<p><a name="pierce1977experiment"></a>
Pierce, Chester M., et al. “An Experiment in Racism <span class="caps">TV</span> Commercials.” <em>Education and Urban Society</em> 10.1 (1977): 61–87. Print.</p>
<p><a name="polman2007meta"></a>
Polman, Hanneke, et al. “A Meta-Analysis of the Distinction Between Reactive and Proactive Aggression in Children and Adolescents.” <em>Journal of Abnormal Child Psychology</em> 35.4 (2007): 522–535. Print.</p>
<p><a name="ramirez2009"></a>
Ramirez, J. Martin. “Some Dychotomous Classifications of Aggression According to its Function.” <em>Journal of Organisational Transformation and Social Change</em> 6.2 (2009): 85–101. Print.</p>
<p><a name="Silverschanz2008"></a>
Silverschanz, Perry, et al. “Slurs, Snubs, and Queer Jokes: Incidence and Impact of Heterosexist Harassment in Academia.” <em>Sex Roles</em> 58.3-4 (2008): 179–191. Print.</p>
<p><a name="small2003helping"></a>
Small, Deborah A. and George Loewenstein. “Helping a Victim or Helping the Victim: Altruism and Identifiability.” <em>Journal of Risk and Uncertainty</em> 26.1 (2003): 5–16. Print.</p>
<p><a name="small2007sympathy"></a>
Small, Deborah A., George Loewenstein, and Paul Slovic. “Sympathy and Callousness: The Impact of Deliberative Thought on Donations to Identifiable and Statistical Victims.” <em>Organizational Behavior and Human Decision Processes</em> 102.2 (2007): 143–153. Print.</p>
<p><a name="Solorzano2000"></a>
Solorzano, Daniel, Miguel Ceja, and Tara Yosso. “Critical Race Theory, Racial Microaggressions, and Campus Racial Climate: The Experiences of African American College Students.” <em>The Journal of Negro Education</em> 69.1/2 (2000): 60–73. Print.</p>
<p><a name="Sue2007"></a>
Sue, Derald Wing, et al. “Racial microaggressions in everyday life: implications for clinical practice.” <em>The American Psychologist</em> 62.4 (2007): 271–286. Print.</p>
<p><a name="Thurlow2001"></a>
Thurlow, Crispin. “Naming the ‘Outsider Within’: Homophobic Pejoratives and the Verbal Abuse of Lesbian, Gay and Bisexual High-School Pupils.” <em>Journal of Adolescence</em> 24.1 (2001): 25–38. Print.</p>
<p><a name="vissing1991verbal"></a>
Vissing, Yvonne M., et al. “Verbal Aggression by Parents and Psychosocial Problems of Children.” <em>Child Abuse <span class="amp">&</span> Neglect</em> 15.3 (1991): 223–238. Print.</p>Train2012-09-29T12:00:00+00:002012-09-29T12:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2012-09-29:/train.html<p>Between the skyscrapers and factories of a large city, a foal is waiting for a train. The train may well be the opposite of Godot.</p><p>I let my eyes trail along the tracks. They seemed to converge in the distance, but much quicker they vanished into the gray of the fog and the fumes. That’s what I call horizon—the point where the air becomes opaque.</p>
<p>These rails snake their way all through the city and beyond, my mom says, wedged between the looming towers whose summits are always shrouded.</p>
<p>I waited.</p>
<p>“New-fangled”—isn’t it ironic that this word would sound so quaint? After the last train had gone, I noticed that somepony left the barrier ajar at the end of the station, and I dared to take a few steps down the stair that lead onto the railway. But only a few steps. There I waited, eyes fixed on the horizon, while the station behind me was crowded with ponies, some waiting to be reunited with their loves ones, some waiting for the moment to say goodbye.</p>
<p>Was there a darker shade emerging or was it my imagination? No, it was the engine! The black of the engine stood out clearly now; I could almost hear it over the noises of the city.</p>
<p>The smoke from the engine was pressing down on the wagons behind it only to coalesce into the air moments later. It had almost reached the station.</p>
<p>I was surprised by my relative calm. I took the few steps back up the stair, never taking my eyes off the wagons. There! The first showcase! A spot of brilliant colors in the side of one wagon, unmissable, unmistakable.</p>
<p>My hooves hit the station the very moment the train reached it, but my eyes were peeled on the wagon with the showcase. Finally! I could make out the first item they had on display: “giant.” The decorators had made it the biggest one, fittingly, yet it seemed tiny compared to its meaning. On one of the lower shelves I could make out “set” in all its iridescence. Beautiful!</p>
<p>The wagon rushed past, but I, trying to take in all its thousands of words could catch none.</p>
<p>Another showcase neared. With the train slowed and my strategy revised, I could discern “sculpture” and “accomplish.” My mane stood on end with excitement!</p>
<p>The third showcase slid closer, and only a few seconds later it came to a stop just a few steps in front of me as I knew it would. Now I could read most of them. I tried to take in as many as I could, which was only a small fraction. Constantly ponies entering and leaving the train—heedless of the abounding wonders!—obstructed my view, so that I tried to edge a little closer to the glass. With my nose pressed against it, I noticed the tiny “minute” the decorators had cunningly used as apostrophe for a much larger “4 o’clock.” I giggled to myself.</p>
<p>No! I mustn’t lose time! I need to read as many as I can! The doors were already closing and I had read only a few score of thousands. What wastefulness!</p>
<p>The train accelerated and was almost lost in the gray of the fog. As its black silhouette faded, so faded the excitement I had felt for each word when I first recognized it.</p>
<p>I would wait for the next train. It shouldn’t be more than an hour.</p>
<p>Some day, maybe, I’ll work as a decorator myself.</p>Microaggressions2012-07-05T18:00:00+00:002012-07-05T18:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2012-07-05:/microaggressions.html<p>Microaggressions are “brief and commonplace daily verbal, behavioral, or environmental indignities” (Sue et al., 2007) that are particularly pernicious as attackers are often unaware of the offense and thus unwilling to acknowledge it and apologize. This neglect can communicate further invalidation to the victim.</p><div class="admonition type">
<p class="admonition-title">Dated Content</p>
<p>I tend to update articles only when I remember their content and realize that I want to change something about it. But I rarely remember it well enough once about two years have passed. Such articles are therefore likely to contain some statements that I no longer espouse or would today frame differently.</p>
</div>
<div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#definition">Definition</a></li>
<li><a href="#homonymy-and-polysemy">Homonymy and Polysemy</a></li>
<li><a href="#effects-of-microaggressions">Effects of Microaggressions</a></li>
<li><a href="#causes-of-microaggressions">Causes of Microaggressions</a></li>
</ul>
</div>
<h3 id="introduction"><a class="toclink" href="#introduction">Introduction</a></h3>
<p>In this article, I want to give a little background on the subject of
microaggressions and provide references to more formal studies, for the
topic has recently been the center of several heated debates on
DeviantArt and image boards.</p>
<p>One caveat: I will not euphemize any pertaining terms. For further
information on the use-mention distinction, see
<a href="http://en.wikipedia.org/wiki/Use%E2%80%93mention_distinction">Wikipedia</a>
and <a href="http://languagelog.ldc.upenn.edu/nll/?p=3431">Language Log</a>. A
trigger warning may still apply.</p>
<h3 id="definition"><a class="toclink" href="#definition">Definition</a></h3>
<p><a href="http://eus.sagepub.com/content/10/1/61">Pierce et al. (1977)</a> observed
that “the chief vehicle for proracist behaviors are microaggressions.
These are subtle, stunning, often automatic, and nonverbal exchanges
which are ‘put-downs’ of blacks by offenders,” which is probably the
first use of the term. Since, the definition, specifically of “racial
microaggressions,” has been widened to “brief and commonplace daily
verbal, behavioral, or environmental indignities, whether intentional or
unintentional, that communicate hostile, derogatory, or negative racial
slights and insults toward people of color.” (<a href="http://www.olc.edu/%7Ejolson/socialwork/OnlineLibrary/microaggression%20article.pdf">Sue et al.,
2007</a>)
They observe furthermore that “perpetrators of microaggressions are
often unaware that they engage in such communications when they interact
with racial/ethnic minorities.”</p>
<p>This definition can be further expanded to include a wider range of
minorities. <a href="http://www.cesa1.k12.wi.us/cms_files/resources/microagressions_article.pdf">Kevin L.
Nadal</a>,
author and associate professor of psychology at John Jay College of
Criminal Justice, adapted the definition to “Microaggressions are brief
and commonplace daily verbal, behavioral, or environmental indignities,
whether intentional or unintentional, that communicate hostile,
derogatory, or negative slights and insults toward members of oppressed
groups.” (Emphasis added)</p>
<p>I doubt that mules are oppressed in Equestria, but apart from that,
there are two instances in which two of the mane six inadvertently
insulted a mule (in “Applebuck Season” and “Hurricane Fluttershy”). Of
course they promptly apologized (“No offense.”) when they noticed their
faux pas. What this example, by analogy, also shows is that, as <a href="http://digitalcommons.calpoly.edu/cgi/viewcontent.cgi?article=1033&context=psycd_fac">Burn et
al.,
(2005)</a>
put it, “heterosexuals, deprived of seeing whom their comments
ultimately harm, are not inclined to carefully monitor their colloquial
speech (<a href="http://faculty.washington.edu/thurlow/papers/Thurlow%282001%29-JofA.pdf">Thurlow,
2001</a>).”
(The mule was <a href="http://www.ponychan.net/chan/arch/res/6286.html#6442">voiced by James
Wootton</a>, by the way.)</p>
<p><a href="http://www.tandfonline.com/doi/abs/10.1080/19361653.2011.584204">Nadel et al.
(2011)</a>
conducted a qualitative study to test and refine a originally
seven-point, now eight-point taxonomy of microaggressions. They term
these categories “themes,” as they are largely perpendicular to a
previously established taxonomy of three “types,” microassault,
microinsult, and microinvalidation (see <a href="http://www.olc.edu/%7Ejolson/socialwork/OnlineLibrary/microaggression%20article.pdf">Sue et al.,
2007</a>).
It should be noted, however, that “this study focused solely on sexual
orientation microaggressions or microaggressions experienced in everyday
lives by <span class="caps">LGB</span> people. Although there may be similarities between sexual
orientation and transgender microaggressions (i.e., microaggressions
experienced by transgender or gender-nonconforming people), there are
complex differences between these two groups, warranting separate
studies. In addition, because it is important not to conflate sexual
orientation and gender identity, the researchers believed it was
important to examine the experiences of these two groups separately.”</p>
<p>The following are the eight classifications taken directly from the
study, which illustrate and exemplify very well what forms
microaggressions can take.</p>
<div> </div>
<p><strong>“Use of heterosexist terminology.”</strong> This very central theme refers to
both intentional and unintentional denigration of <span class="caps">LGB</span> individuals, e.g.,
through heterosexist jokes or comments. In particular, “participants
described certain words such as ‘faggot’ and ‘dyke’ as denigrating to
them, sending the message that it is inferior or undesirable to be <span class="caps">LGB</span>.”
A quote from a participant:</p>
<blockquote>
<p>I recently opened up to my friend about [being gay] and he’s a guy …
and just the other day I was at his house and were talking about other
people and he would describe them as like, “faggot,” and it would get
to me.</p>
</blockquote>
<p>The researchers observed that in some environments it still seemed
socially acceptable to use such language, be it directly (“When you’re
angry at someone, you can call them a ‘faggot’ and that’s still okay.”)
or what <a href="http://www.lsa.umich.edu/psych/lilia-cortina-lab/silverschanz%20et%20al.%202008%20.pdf">Silverschanz et al.
(2008)</a>
call ambient heterosexual harassment: “Participants also described that
peers used the word “gay” in negative contexts (i.e., as synonymous to
‘bad’ or ‘weird’). According to the participants, hearing such remarks
was hurtful, distressful, and made them feel uncomfortable or unsafe.”</p>
<div> </div>
<p><strong>“Endorsement of heteronormative culture/behaviors.”</strong> Another form of
microaggression occurs when <span class="caps">LGB</span> persons have to hide or disguise their
sexual orientation, be it within the social context of their families,
at the workplace, or publicly. One participant described her experiences
with the first:</p>
<blockquote>
<p>[My mother] knows that I’m a lesbian, but she is in denial. She
doesn’t want to see it, so I have to act a certain way. You know, act
heterosexual, not mention anything about me having a girlfriend or
anything like that to make her feel uncomfortable or make her say
anything offensive toward me. So, I have to act completely different
at home.</p>
</blockquote>
<div> </div>
<p><strong>“Assumption of universal <span class="caps">LGBT</span> experience.”</strong> This theme describes the
degradation experienced by <span class="caps">LGB</span> persons when others expect them to comply
with societal stereotypes they don’t identify with. Two participants reported:</p>
<blockquote>
<p>They’re making it seem like everybody is sexual and smutty-buddy and,
you know, every gay man is an interior decorator or sells hair for a
living, you know. Even at my job … if you work in stocks, then it’s a
straight guy, but if you’re on the beauty floor and/or you’re a
cashier, you’re gay.</p>
<p>Since I dress a little bit more feminine than most other lesbians they
might take [my identity as a lesbian] as a joke or make offensive statements.</p>
</blockquote>
<div> </div>
<p><strong>“Exoticization.”</strong> This theme is closely related to the previous one,
but describes the aspect of objectification and dehumanization <span class="caps">LGB</span>
individuals are exposed to, for example, when they are treated like a
stereotype. Again the researchers point out that “although the intention
of the perpetrator is to be complimentary, the victim experiences a microinsult.”</p>
<blockquote>
<p>This woman came up to me one night and she said … I think I made some
joke or something and she said, ‘Do you know who you remind me of?’
and I knew what was coming, I just knew what was coming. She’s like,
“You’re just like that Jack on Will and Grace. You’re so funny.” And I
looked at her, and I said, “Ma’am, no offense, but that’s actually not
a compliment.” And she was like, “What do you mean? What do you mean?
No, no, I was saying you’re funny, and you’re cute, and you dress nice.”</p>
<p>A lot of guys would think, you know, because I’m into both guys and
girls that I’ll be like down with the threesome kinda thing, and it’s
like ugh, get over yourself.</p>
</blockquote>
<div> </div>
<p><strong>“Discomfort/disapproval of <span class="caps">LGBT</span> experience.”</strong> This name is fairly
self-explanatory. One of the participants in the study gives an example:</p>
<blockquote>
<p>I was in college, and I came out to a friend who was very conservative
Christian, and she didn’t say “I’m going to stop being your friend,”
but she did say she was sorry to hear that because “I believe you are condemned.”</p>
</blockquote>
<p>Others gave examples of disapproving glances, overt ridicule, and also
legislation such as the overturned <em>Don’t Ask Don’t Tell</em> policy.</p>
<div> </div>
<p><strong>“Denial of the reality of heterosexism.”</strong> Sometimes people fail to
realize the extent to which individuals with minority sexual
orientations are subjected to heterosexism. Through overt disregard for
their situation they can feel invalidated, or children and adolescents
may even be plunged into self-doubt about the validity of their
perceptions. This theme also encompasses situations in which people who
have perpetrated microaggressions fail to admit their error.</p>
<div> </div>
<p><strong>“Assumption of sexual pathology/abnormality.”</strong> This theme occurs when
societal stereotypes lead uninformed individuals to assume that <span class="caps">LGB</span>
people were suffering from psychological disorders related to their
sexual behavior (being “oversexualized, sexual deviants, or both”) or
were suffering from <span class="caps">HIV</span>/<span class="caps">AIDS</span>.</p>
<blockquote>
<p>Well, in my case, I’ve actually had some friends stop being my
friends, because they were like, “Oh, since you’re bisexual and you
might try come on to me” so they stopped being my friend.</p>
</blockquote>
<div> </div>
<p><strong>“Threatening behaviors.”</strong> The participants reported another kind of
experiences that did not neatly fit into any of the categories
established by the previous themes. They often found themselves in
threatened with physical assaults. Such assaults, of course, could not
be classified as microaggressions, but the hostile environment created
by the threat can. These situations can be temporary as in the example
below, but may also be permanent and recurring, for example, in the case
of children and adolescents constantly threatened by the presence of
bullies at their school or in their class.</p>
<blockquote>
<p>I was walking with three friends of mine—all male, all gay. … It was
pretty late, probably about two in the morning, and two guys were
standing near a park, and we just passed by them, and one of them
said, “What did you say?” We didn’t respond … and then they continued
and said, “Did you just call me gay? Did you just call me a faggot?”
They followed us for about two blocks and tried to start a fight.</p>
</blockquote>
<p>What is also interesting to observe is how many of these themes of
heterosexist microaggressions can easily be extended to accommodate
forms of animadversion upon bronihood (or pony fan identity). The
analogue to “endorsement of heteronormative culture/behaviors” could be
the propensity of certain
<a href="http://www.breitbart.com/Big-Hollywood/2012/04/25/my-little-pony">authors</a>
or <a href="http://www.youtube.com/watch?v=RwcJorJp_Aw">pundits</a> to endorse the
idea that adults mustn’t like a show intended for children (even though
Lauren Faust of course created it <a href="http://furrymedia.livejournal.com/321145.html">with the parents in
mind</a>) or that men
mustn’t like a show intended for a female audience. The counterpart to
the “assumption of universal <span class="caps">LGBT</span> experience” could include perpetuating
the cliché of the immature, disheveled brony, the homosexual brony, and
the exclusively male brony. “Discomfort/disapproval of <span class="caps">LGBT</span> experience”
could find its counterpart in disapproval of pony T-shirts, cosplay, or
the public unpacking of merchandise. “Denial of the reality of
heterosexism,” of course, could translate into <a href="http://nymag.com/daily/intel/2011/11/4chan-based_bronies_continue_t.html">some
authors’</a>
idea that we are all just facetious about our love for the show.</p>
<h3 id="homonymy-and-polysemy"><a class="toclink" href="#homonymy-and-polysemy">Homonymy and Polysemy</a></h3>
<p>When different words (or rather lexemes) share the same spelling and
pronunciation, it’s called homonymy; when the same word (or rather
lexeme) has different meanings, we talk about polysemy, “the greedy
habit some words have of taking more than one meaning for themselves,”
<a href="http://www.ted.com/talks/erin_mckean_redefines_the_dictionary.html">as Erin McKean put
it</a>.
(The notion of identity between lexemes is beyond the purview of this article.)</p>
<p>Most common words have several meanings, and some (like <a href="http://www.nytimes.com/2011/05/29/opinion/29winchester.html">“set” and
“run”</a>)
even several hundred. This is relevant to the topic of microaggressions
as especially the term “gay” has acquired very different uses in recent
decades. Researchers of the University of Canberra and the Australian
National University have studied the usage of the term among different
age groups and found that, in addition to the meaning of “gay” as
synonymous with “homosexual,” older people still recognized the sense of
“ ‘happy,’ ‘carefree’ and ‘frivolous,’ ” while younger Australians
increasingly understood it to mean “ ‘stupid’, ‘lame’ or ‘boring.’ ”</p>
<p>Usually, of course, the intended denotation is clear not only from the
pragmatic context of the conversation, but also from its immediate
grammatical context, as “young people (18–30 year olds) understand the
meaning of ‘gay’ differently depending upon whether the subject is
animate (e.g. ‘he’, ‘she’) or inanimate (e.g. ‘that film’); whether it
is used with the [linking] verb ‘to look’ or the copula ‘to be’; and
whether the word ‘gay’ is used in conjunction with the intensifier ‘so’
(e.g. ‘They’re gay’ compared to ‘They’re so gay’).” (<a href="http://www.tandfonline.com/doi/abs/10.1080/07268600701522764">Lalor et al.,
2007</a>) The
microaggression, instead, lies in the implication that the homosexual
person lives amidst a society that would allow a term for their sexual
orientation to acquire a pejorative meaning.</p>
<p>The term “fag” is a different case, as it has different meanings that
are <a href="http://www.thefreedictionary.com/fag">typically regarded as separate
lexemes</a> as they don’t share the
same etymological roots, the one meaning “drudgery,” the one in
“fag-end,” which is also used to denote a cigarette, and the disparaging
and offensive term for a homosexual, derived from “faggot.”
Distinguishing “meanings” (of homonyms) and “senses” (of polysemes),
<a href="http://www.sciencedirect.com/science/article/pii/S0749596X01928104">Rodd et al.
(2002)</a>
observe in their experiments that “while multiple word senses do produce
faster responses, ambiguity between multiple meanings delays
recognition. These results suggest that, while competition between the
multiple meanings of ambiguous words delays their recognition, the rich
semantic representations associated with words with many senses
facilitate their recognition.”</p>
<p>Interestingly, the third sense of “fag” is not only widely used as
insult, but has also acquired neutral and positive meanings in some
limited sociolects, especially within compounds. All these senses and
meanings, however, are usually well-distinguished by the context of
their usage, and thus exist in parallel with minimal influence on one another.</p>
<p>Even though different senses of words can exist in parallel without one
dominating or supplanting the other, and the intended meaning is usually
clear from the context, sufficiently advanced speakers of a language are
easily aware of the various shades of any given term. Any pun can attest
to that: “Atheists don’t solve exponential equations because they don’t
believe in higher powers.”</p>
<h3 id="effects-of-microaggressions"><a class="toclink" href="#effects-of-microaggressions">Effects of Microaggressions</a></h3>
<blockquote>
<p>Compared to the overt hate violence that gays may experience, the use
of derogatory terms for gays by heterosexuals to refer to each other
may seem innocuous and minor. However, this behavior perpetuates
anti-gay prejudice and violence by suggesting that it is socially
acceptable to exhibit bias against gays. In other words, it
contributes to heterosexism, which <a href="http://jiv.sagepub.com/content/5/3/316.abstract">Herek
(1990)</a> defines as
the denigration and stigmatization of any nonheterosexual form of
behavior, identity, relationship, or community. (<a href="http://www.tandfonline.com/doi/abs/10.1300/J082v40n02_01">Burn,
2000</a>)</p>
</blockquote>
<p>Numerous studies linked to minority stress theory have also confirmed
that “socially marginalized groups, including sexual minorities, can
experience mental and physical health problems resulting from negative
social environments created by stigma, prejudice, and discrimination
(e.g., <a href="http://psycnet.apa.org/journals/cou/46/3/395/">Fischer and Shaw
1999</a>; <a href="http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1447127/">Gee
2002</a>; <a href="http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1447724/">Harrell et
al. 2003</a>;
<a href="http://midus.wisc.edu/findings/pdfs/52.pdf">Kessler et al. 1999</a>;
<a href="http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2072932/">Meyer 2003</a>;
<a href="http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1447717/">Williams et al.
2003</a>),”
(<a href="http://www.lsa.umich.edu/psych/lilia-cortina-lab/silverschanz%20et%20al.%202008%20.pdf">Silverschanz et al.,
2008</a>)
and a study of the behavior of 14- to 15-year-old high school students
confirmed the “increasingly well-documented daily assault on the
psychological health of young homosexual people.” (<a href="http://faculty.washington.edu/thurlow/papers/Thurlow%282001%29-JofA.pdf">Thurlow,
2001</a>)
In their qualitative study cited above, <a href="http://www.tandfonline.com/doi/abs/10.1080/19361653.2011.584204">Nadal et al.
(2011)</a>
list a few of these mental and physical effects.</p>
<ul>
<li>Among the emotions that participants reported to have experienced
due to microaggressions were “anger, frustration, … sadness, …
belittlement and hopelessness.”</li>
<li>In several cases, these experiences also led to “detrimental
relationships with their family members, friends, coworkers, and others.”</li>
<li>Participants also discussed the detrimental role of microaggressions
in “their ability to be comfortable with their <span class="caps">LGB</span> identities.”
Especially adolescents’ self-worth can suffer, as Nadal et al. point
out, affecting their personal and professional prospects: “For
example, many studies discussed how <span class="caps">LGB</span> youth who experience school
violence, heterosexist threats, or damage to their property may
avoid going to school altogether (<a href="http://www.pediatricsdigest.mobi/content/101/5/895.full">Garofalo, Wolf, Kessel, Palfrey,
<span class="amp">&</span> DuRant,
1998</a>;
<a href="http://www.tandfonline.com/doi/abs/10.1080/19361650802379748">Grossman et al.,
2009</a>).”</li>
<li>Also discussed were “chronic mental health effects,” including
“post-traumatic stress disorder (<span class="caps">PTSD</span>).” They also cite reference
literature that documents that “<span class="caps">LGB</span> youth also experience higher
rates of emotional distress, higher amounts of suicide attempts,
risky sexual behavior, and substance abuse (<a href="http://www.pediatricsdigest.mobi/content/101/5/895.full">Garofalo et al.,
1998</a>;
<a href="http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1508407/pdf/amjph00013-0059.pdf">Remafedi, Frendh, Story, Resnick, <span class="amp">&</span> Blum,
1998</a>;
<a href="http://jama.ama-assn.org/content/278/10/823.full.pdf">Resnick et al.,
1997</a>).”</li>
</ul>
<p>In much the same vein, <a href="http://digitalcommons.calpoly.edu/cgi/viewcontent.cgi?article=1033&context=psycd_fac">Burn et al.
(2005)</a> write:</p>
<blockquote>
<p>Even if heterosexist language is not used to intentionally harm <span class="caps">LGB</span>
persons, it may be experienced as anti gay harassment and contribute
to psychosocial stress. Experiences of negative treatment in society
and resultant lack of self-acceptance culminate to produce abnormally
high chronic stress for <span class="caps">LGB</span> persons (<a href="http://cpmcnet.columbia.edu/dept/healthandsociety/events/ms/year4/pdf/sh_Meyer%20IH.pdf">Meyer,
1995</a>).
Research finds a number of negative effects from the stress related to
stigmatization based on sexual orientation (<a href="http://www.apa.org/pi/lgbt/resources/guidelines.aspx"><span class="caps">APA</span> Division 44,
2000</a>). This
stress is linked to depression, higher suicide rates among <span class="caps">LGB</span> persons
during young adulthood (<a href="http://homes.chass.utoronto.ca/%7Etkennedy/Courses/38H3/Daugelli.pdf">D’Augelli,
1992</a>;
<a href="http://books.google.de/books/about/I_thought_people_like_that_killed_themse.html?id=Hf8DAQAAIAAJ">Rofes,
1983</a>;
<a href="http://jar.sagepub.com/content/9/4/498.short">Rotheram-Borus, Hunter, <span class="amp">&</span> Rosario,
1994</a>; <a href="http://psycnet.apa.org/journals/ccp/62/2/261/">Savin-Williams,
1994</a>), high-risk
sexual behaviors (<a href="http://psycnet.apa.org/psycinfo/1993-45985-001">Folkman, Chesney, Pollack, <span class="amp">&</span> Phillips,
1992</a>;
<a href="http://www.sciencedirect.com/science/article/pii/S0140197185710020">Rotherman-Borus, Reid, Rosario, <span class="amp">&</span> Kasen,
1995</a>),
eating disorders (<a href="http://www.tandfonline.com/doi/abs/10.1300/J082v12n03_09">Brown,
1986</a>),
school problems (Bendet, 1986), substance abuse, running away, and
prostitution (<a href="http://www.apa.org/pi/lgbt/resources/guidelines.aspx"><span class="caps">APA</span> Division 44,
2000</a>). The
experience of anti-gay harassment has been found to be more common
among gay and bisexualmale adolescents who had attempted suicide than
among those who had not (<a href="http://jar.sagepub.com/content/9/4/498.short">Rotheram-Borus et al.,
1994</a>).</p>
</blockquote>
<p>Another interesting dimension is added by <a href="http://www.lsa.umich.edu/psych/lilia-cortina-lab/silverschanz%20et%20al.%202008%20.pdf">Silverschanz et al.
(2008)</a>,
who, in contrast to the other studies examined in this article, extended
the purview of their study to include heterosexuals and the ways in
which they are influenced by personal and ambient heterosexist
harassment. Their analysis proved that “heterosexual identification does
not shield individuals from heterosexist harassment and its association
with negative outcomes.” They further hypothesize:</p>
<blockquote>
<p>Because heterosexual students experiencing <span class="caps">HH</span> overwhelmingly reported
ambient rather than personal <span class="caps">HH</span> (84% and 16%, respectively), it is
possible that our results can be explained by bystander stress. In
other words, heterosexual students tended to suffer from overhearing
others make negative remarks about sexual-minority people.</p>
</blockquote>
<p>“Sticks and stones may be more likely to break their bones,” <a href="http://faculty.washington.edu/thurlow/papers/Thurlow%282001%29-JofA.pdf">Thurlow
(2001)</a>
concludes, ”but the relentless, careless use of homophobic pejoratives
will most certainly continue to compromise the psychological health of
young homosexual and bisexual people by insidiously constructing their
sexuality as something wrong, dangerous or shameworthy.”</p>
<h3 id="causes-of-microaggressions"><a class="toclink" href="#causes-of-microaggressions">Causes of Microaggressions</a></h3>
<p>Perhaps counter-intuitively, only some of the perpetrators of
microaggressions are actually ideologically heterosexist. <a href="http://www.tandfonline.com/doi/abs/10.1300/J082v40n02_01">Burn
(2000)</a> observes:</p>
<blockquote>
<p>The use of terms such as “fag” or “queer” in heterosexual friendship
groups is in many cases normative. That is, it is part of the group’s
culture. If the individual wishes to be identified as an ingroup
member, s/he must participate in the group’s culture. The terms, which
may be quite creative (e.g., “butt-pirate” and “fudgepacker”), are
reinforced through laughter and frequency of use. In this way, this
expression of anti-gay prejudice serves what <a href="http://jiv.sagepub.com/content/5/3/316.short">Herek
(1990)</a> would call a
social-expressive function by helping individuals win approval from
important others and affirm their status as “insiders.” Similarly,
<a href="http://www.tandfonline.com/doi/abs/10.1080/00224545.1991.9713823">Sigelman et al.
(1991)</a>
suggest that anti-gay behaviors may arise as individuals try to
distance themselves from stigmatized persons out of a concern that
they will be stigmatized by association (what Goffman called “courtesy
Stigma”). By using anti-gay language, individuals distance themselves
from this stigmatized social group. <a href="http://www.stthomas.edu/diversity/files/Masculinity_as_Homop.pdf">Kimmel
(1994)</a>
suggests that the fear men have of being perceived as homosexual
propels them to enact all kinds of masculine behaviors and attitudes,
such as homophobic remarks, to make sure that no one gets the “wrong
idea” about their manliness.</p>
</blockquote>
<p>Ergo, what these groups lack is a Fancy Pants, an insider of a social
group who is confident in their standing inside the group and courageous
enough to question and to stand up to its memetic aberrations.</p>
<p>Another remedy against stereotypes is personal information. <a href="http://digitalcommons.calpoly.edu/cgi/viewcontent.cgi?article=1033&context=psycd_fac">Burn et
al.,
(2005)</a>,
cite research indicating that “people will set aside their stereotypes
and judge people on an individual basis when personal information is
available to them (<a href="http://psycnet.apa.org/journals/psp/57/2/201/">Hilton <span class="amp">&</span> Fein,
1989</a>; <a href="http://psycnet.apa.org/journals/psp/66/4/658/">Lord et al.,
1994</a>) and when they are
highly motivated to form an accurate impression of someone (<a href="http://books.google.com/books?hl=en&lr=&id=sNadYmHFsUAC&oi=fnd&pg=PA235&dq=Hilton,+J.+L.,+%26+Darley,+J.+M.+%281991%29.+The+effects+of+interaction+goals+on+person+percep%C2%AD+tion.++&ots=qEtsBhob5L&sig=BtUFq3NhchT91chXn4avh7VA1UY">Hilton <span class="amp">&</span>
Darley,
1991</a>;
<a href="http://books.google.com/books?hl=en&lr=&id=z9vHEy0osBAC&oi=fnd&pg=PA67&dq=Snyder,+M.+%281992%29.+Motivational+foundations+of+behavioral+confirmation+&ots=Pfrrm0iWez&sig=bj9UfYWpct87zBH4rXKgy2UH9VE">Snyder,
1992</a>).”
Hence they conclude that “coming out is, perhaps, the most potent method
of reducing anti gay sentiments (<a href="http://books.google.de/books?id=BoBjkgAACAAJ">Davis,
1992</a>; <a href="http://books.google.de/books/about/Psychological_Perspectives_on_Lesbian_an.html?id=XgL89Ve48t4C">Garnets <span class="amp">&</span> Kimmel,
1993</a>;
<a href="http://books.google.de/books/about/Homosexuality.html?id=NCQbAAAAYAAJ">Gonsiorek <span class="amp">&</span> Weinrich,
1991</a>;
<a href="http://books.google.de/books?id=GailsZNFdS8C">Klein, 1993</a>).”</p>
<p>In other instances where people “with otherwise accepting attitudes
toward homosexuality call their heterosexual friends derogatory names,”
it is possible that they “merely have not thought of its contribution to
antihomosexual bias,” so that “simpler awareness efforts should succeed
in changing behavior.” For instance, “many people professed that they
had simply never thought of this type of behavior as gaybashing and were
quite ashamed of themselves upon reflection.” (<a href="http://www.tandfonline.com/doi/abs/10.1300/J082v40n02_01">Burn,
2000</a>)</p>
<p>Within a fandom made up of people who are open-minded enough to profess
their love for well-animated equines, such awareness efforts should fall
on fertile ground. Hence, the simplest and most effective remedy and
prevention measure for microaggressions may well be to politely point
them out when one notices them in the discourse of others and to train
greater consciousness of the “social-expressive functions” of one’s own
language. If you are a member of a tight-knit online or real-life group
of friends, then you carry responsibility for the culture you
cultivate—responsibility that stems from the influence you wield.</p>The Intransitive to Warn2011-03-27T12:00:00+00:002011-03-27T12:00:00+00:00Dawn Dreschertag:impartial-priorities.org,2011-03-27:/intransitive-warn.html<p>The intransitive use of <em>to warn</em> – when no recipient of the warning is specified – apparently sounds odd to some native speakers in some contexts (e.g., <a href="http://languagelog.ldc.upenn.edu/nll/?p=2623">Pullum</a> and <a href="https://languagelog.ldc.upenn.edu/nll/?p=2625">Zimmer</a>). This document is a compilation of various scraps of data on the issue – focusing on current usage, recent history, and potential differences between British and American English – for anyone who is interested in this sort of thing.</p><style>
.half-width img,
.half-width p {
display: block;
max-width: 50%;
}
</style>
<div class="toc"><span class="toctitle">Contents</span><ul>
<li><a href="#corpora">Corpora</a><ul>
<li><a href="#corpus-of-contemporary-american-english">Corpus of Contemporary American English</a></li>
<li><a href="#british-national-corpus">British National Corpus</a></li>
<li><a href="#corpus-of-historical-american-english">Corpus of Historical American English</a></li>
<li><a href="#google-books-ngram-viewer">Google Books Ngram Viewer</a></li>
<li><a href="#american-english-corpus">American English Corpus</a></li>
<li><a href="#british-english-corpus">British English Corpus</a></li>
</ul>
</li>
<li><a href="#great-authors">Great Authors</a></li>
<li><a href="#etymology">Etymology</a></li>
<li><a href="#modern-english-dictionaries">Modern English Dictionaries</a></li>
</ul>
</div>
<h2 id="corpora"><a class="toclink" href="#corpora">Corpora</a></h2>
<p>The ultimate usage authorities for a certain dialect are its speakers, and corpora are the searchable aggregates of these people’s voices. Unlike dictionaries, however, corpora leave most of the interpreting to you. In that I will follow their example.</p>
<p>The query I used, <code>-[be] [warn].[v*] ,|.|of|against|that</code>, is optimized to exclude as many transitive constructions as possible, not to capture all intransitive ones. <code>-[be]</code> excludes passive constructions, <code>[warn].[v*]</code> matches all verb forms of <em>to warn</em> (<code>.[v*]</code> specifies the tags), and matches any of those five tokens.</p>
<p><code>[warn].[v*]</code> also matches the potentially -ambiguous <em>warning</em>, relying completely on the corpus’s tagging to make the distinction. In order to confirm the plausibility of the results, I compared them to the results of <code>-[be] warned|warns|warn ,|.|of|against|that</code>.</p>
<p>I will not distinguish between intransitive and absolute (also called <em>elliptical</em>) usages of <em>to warn</em>, as I see in this case no reliable way to make the distinction (save, perhaps, asking the speaker) and certainly none I could translate into a corpus query. Furthermore, absolute usages are extremely rare in any case. Absolute verbs are verbs whose objects are implied but not stated. Discussing <em>to remind</em>, Merriam-Webster’s Dictionary of English Usage(1995, p. 812) gives an example from the The Times Literary Supplement of February 11, 1983: “Your correspondence … reminds of a question.”</p>
<h3 id="corpus-of-contemporary-american-english"><a class="toclink" href="#corpus-of-contemporary-american-english">Corpus of Contemporary American English</a></h3>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/corpora/coca_chart.png"
alt="COCA genre chart"
/>
<p>COCA genre chart.</p>
</div>
<p>The <a href="http://corpus.byu.edu/coca/help/intro_e.asp">Corpus of Contemporary American English</a>, <span class="caps">COCA</span>, “is the largest freely-available corpus of English, and the only large and balanced corpus of American English. It was created by Mark Davies of Brigham Young University in 2008.” It “contains more than 410 million words of text and is equally divided among spoken, fiction, popular magazines, newspapers, and academic texts. It includes 20 million words each year from 1990–2010 and the corpus is also updated once or twice a year ….”</p>
<p>On January 23, 2011, my query turned up 7,369 tokens, about 35% of the 20,918 tokens matching <code>[warn].[v*]</code>. I manually checked 100 tokens out of a cross section of all years and genres, none of which was transitive. There are, however, infrequent constructions such as “When field reviews began in the 1990s, states were repeatedly warned that they were failing to comply with the law warnings that continue to this day” (<a href="http://corpus.byu.edu/coca/x4.asp?t=4024646&ID=468017976">source</a>) and “Lincoln warned that able organizer but dilatory strategist, General George B. McClellan, ‘Your enemies will probably use time as advantageously as you can’” (<a href="http://corpus.byu.edu/coca/x4.asp?t=2019848&ID=140203212">source</a>) that do slip through.</p>
<p>On the same day, my comparison query yielded 7,066 tokens, about 37% of the 18,917 tokens for . The difference of two percentage points (about 6%) is barely statistically significant, and probably negligible for the purposes of this investigation. Again, my hand-checked sample of 100 tokens included no false-positives.</p>
<h3 id="british-national-corpus"><a class="toclink" href="#british-national-corpus">British National Corpus</a></h3>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/corpora/bnc_chart.png"
alt="BNC genre chart"
/>
<p>BNC genre chart.</p>
</div>
<p>The <a href="corpus.byu.edu/bnc/">British National Corpus</a>, <span class="caps">BNC</span>, contains about 100 million words of texts written between the 1970s and 1993.</p>
<p>In the <span class="caps">BNC</span>, the query yields 2,067 tokens, about 34% of the 6,134 tokens for <code>[warn].[v*]</code>. (Excluding <em>warning</em>, as above: <span class="math">\(\frac{1,988}{5,587} \approx 36\%\)</span>.) This time, however, the query tuned up <a href="http://corpus.byu.edu/bnc/x4.asp?t=3686&ID=136584753">one false-positive</a>, due to a mistagged <em>warning</em> in the 100 entries that I checked manually.</p>
<p>Note that <a href="http://corpus.byu.edu/bnc/help/compare_bnc.asp">the genre balance of the is not as even as that of the <span class="caps">COCA</span></a>, hence the <em>frequencies per million words</em> rather than the raw frequencies should be used for comparisons of genres.</p>
<h3 id="corpus-of-historical-american-english"><a class="toclink" href="#corpus-of-historical-american-english">Corpus of Historical American English</a></h3>
<div class="lightbox-block align-center">
<img
src="/images/intransitive-warn/corpora/coha_chart.png"
alt="COHA, per year"
width="80%"
/>
<p>COHA, per year.</p>
</div>
<p>The <a href="http://corpus.byu.edu/coha/help/intro_e.asp">Corpus of Historical American English</a>, <span class="caps">COHA</span>, “is the largest structured corpus of historical English (or any language, for that matter).” It contains “more than 400 million words of text of American English from 1810 to 2009.”</p>
<p>Here, the query yields 4,046 tokens, 22% of the 18,747 tokens matching <code>[warn].[v*]</code>, which is interpreted as <code>warn|warns|warned</code> in this corpus, rendering the comparison query redundant.</p>
<h3 id="google-books-ngram-viewer"><a class="toclink" href="#google-books-ngram-viewer">Google Books Ngram Viewer</a></h3>
<p>When the <a href="http://ngrams.googlelabs.com/">Google Books Ngram Viewer</a> launched in December 2010, it immediately joined many a linguist’s league of favorite toys. As the name suggests, it is not a feature-rich corpus, but an interface to a database of n-grams, series of <span class="math">\(n\)</span> (currently one to five) words, gleaned from some of the 15 million books scanned by Google. <a href="http://googlesystem.blogspot.com/2010/12/google-books-ngram-viewer.html">And it is gargantuan</a>: “The datasets we’re making available today to further humanities research are based on a subset of that corpus, weighing in at 500 billion words from 5.2 million books in Chinese, English, French, German, Russian, and Spanish.”</p>
<p>Since the past tense forms of <em>to warn</em> are significantly more frequent than the present tense ones, I opted to use separate graphs with differing scales. Note also that there is no way to exclude passive constructions in this search.</p>
<h3 id="american-english-corpus"><a class="toclink" href="#american-english-corpus">American English Corpus</a></h3>
<div class="lightbox-block align-center">
<img
src="/images/intransitive-warn/corpora/ngram_past_american.png"
alt="Past tense of to warn in American Google Books n-gram database"
/>
<p>Past tense of *to warn* in American Google Books n-gram database.</p>
</div>
<div class="lightbox-block align-center">
<img
src="/images/intransitive-warn/corpora/ngram_present_american.png"
alt="Present tense of to warn in American Google Books n-gram database"
/>
<p>Present tense of *to warn* in American Google Books n-gram database.</p>
</div>
<h3 id="british-english-corpus"><a class="toclink" href="#british-english-corpus">British English Corpus</a></h3>
<div class="lightbox-block align-center">
<img
src="/images/intransitive-warn/corpora/ngram_past_british.png"
alt="Past tense of to warn in British Google Books n-gram database"
/>
<p>Past tense of *to warn* in British Google Books n-gram database.</p>
</div>
<div class="lightbox-block align-center">
<img
src="/images/intransitive-warn/corpora/ngram_present_british.png"
alt="Present tense of to warn in British Google Books n-gram database"
/>
<p>Present tense of *to warn* in British Google Books n-gram database.</p>
</div>
<h2 id="great-authors"><a class="toclink" href="#great-authors">Great Authors</a></h2>
<p>These are excerpts from books by people I hold in high esteem. (I have only read some of the books, though.)</p>
<blockquote>
<p>Jefferson specifically warned against “banking institutions and monied incorporations” (what we would now call “corporations”) and said that if they grow, the aristocrats will have won and the American Revolution will have been lost. —Noam Chomsky, <em>Secrets, Lies, and Democracy</em> (Odonian Press, 1994)</p>
<p>She warned that launching an attack on Iraq would undermine counterterrorism efforts. —Amy Goodman and David Goodman, <em>Static: Government Liars, Media Cheerleaders, and the People Who Fight Back</em> (Hyperion Books, 2007)</p>
<p>She says that a possible outcome of the Act is “the criminalization of legitimate political dissent” and warns that it “grants the executive branch unprecedented, and largely unchecked, surveillance powers, including the enhanced ability to track email and Internet usage, conduct sneak-and-peak searches, obtain sensitive personal records, [and] monitor financial transactions.” (addition in original) —Howard Zinn, <em>Terrorism and War</em> (Seven Stories Press, 2002)</p>
<p>Worse, a 1999 report had already warned that aj Qaeda was looking into using planes as missiles with the intention of crashing them into government buildings. —Michael Moore, <em>Dude, Where’s My Country?</em> (Warner Books, 2003)</p>
<p>While the army corps of engineers and others warned of the urgent impending danger to New Orleans and put forward proposals for funding to reinforce and protect the city, the Bush administration, in every year since 2001, has cut or refused to fund New Orleans flood control, and ignored scientists’ warnings of increased hurricanes as a result of global warming. —Jordan Flaherty, <em>Floodlines: Community and Resistance from Katrina to the Jena Six</em> (Haymarket Books, 2010)</p>
<p>Tufte’s (2001) excellent book on chart design warns against a common error. —Chicago University Press, <em>The Chicago Manual of Style</em> (University of Chicago Press, 2010)</p>
<p>Most of all, that earlier Silk would have prayed devoutly to the Outsider, though the Outsider had warned that he would send no aid. —Gene Wolfe, <em>Exodus from the Long Sun</em> (Tom Doherty Associates, 1997)</p>
<p>Prudence warned against it as well. —Gene Wolfe, <em>The Sword of the Lictor</em> (Arrow Books, 1991)</p>
<p>Her voice told him, very sweetly, that he was welcome to leave a message after the beep, but warned that she hardly ever listened to them and that it was much better to talk to her directly, only he couldn’t because she wasn’t in, so he’d best try again. —Douglas Adams, <em>The Long Dark Tea-Time of the Soul</em> (Simon <span class="amp">&</span> Schuster, 1991)</p>
<p>I remembered the previous owner had said a mechanic had told him the plate was hard to get on. That was why. The shop manual had warned about this, but like the others he was probably in too much of a hurry or he didn’t care. —Robert Maynard Pirsig, <em>Zen and the Art of Motorcycle Maintenance: An Inquiry Into Values</em> (HarperTorch, 2006)</p>
<p>So, if evidence is such an intangible thing after all, why am I warning against new ways of interpreting evidence? —Douglas Hofstadter, <em>Gödel, Escher, Bach: An Eternal Golden Braid</em> (Basic Books, 1979)</p>
</blockquote>
<h2 id="etymology"><a class="toclink" href="#etymology">Etymology</a></h2>
<p>The <a href="http://www.oed.com/view/Entry/225791"><em>Oxford English Dictionary</em></a> writes:</p>
<blockquote>
<p>Old English warenian, warnian, wearnian = Middle Low German warnen to warn, inform, Flemish (Kilian) †waernen to warn, put on one’s guard, Old High German warnôn, warnên to warn, refl. to provide oneself, to take precautions (Middle High German, modern German warnen to warn; the Swedish varna, Danish varne are probably < German) < Germanic *waranōjan (-ǣjan), < *war- to be cautious: see ware adj.</p>
<p>In Old English and in Continental Germanic this verb seems to have been to some extent confused with Germanic *warnōjan, cognate and synonymous with *warnjan to refuse, forbid, etc. (see warn v.2).</p>
</blockquote>
<p>Among the twenty-three senses the lists are seven marked <em>absol.</em> and two marked <em>intr.</em> (5.e. has both labels, <em>absol. or intr.</em>), but none of these is likely to be of interest here, for the newest citation is from 1900.</p>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/A_Comprehensive_Etymological_Dictionary_of_the_English_Language.png"
alt="Ernest Klein, A Comprehensive Etymological Dictionary of the English Language (Elsevier Pub. Co., 1967)"
/>
<p>Ernest Klein, <em>A Comprehensive Etymological Dictionary of the English Language</em> (Elsevier Pub. Co., 1967).</p>
</div>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/Skeat_Etymological_Dictionary.png"
alt="Walter William Skeat, An Etymological Dictionary of the English Language (Oxford University Press, 1924)"
/>
<p>Walter William Skeat, <em>An Etymological Dictionary of the English Language</em> (Oxford University Press, 1924).</p>
</div>
<h2 id="modern-english-dictionaries"><a class="toclink" href="#modern-english-dictionaries">Modern English Dictionaries</a></h2>
<p>The <a href="http://century-dictionary.com"><em>Century Dictionary</em></a> (1889–1891) listed <em>to warn</em> only as transitive (vol. 8, p. 6825), and so does <em>Webster’s Unabridged Dictionary</em> (1913), both of which can be searched at <a href="http://www.wordnik.com/words/warn">Wordnik</a>.</p>
<p>Some two decades later, the <em>Thorndike-Century Junior Dictionary</em> recognized the intransitive use already:</p>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/Thorndike-Century_Junior_Dictionary_1935.png"
alt="Edward Lee Thorndike, The Thorndike-Century Junior Dictionary (Scott Foresman / Company, 1935)"
/>
<p>Edward Lee Thorndike, <em>The Thorndike-Century Junior Dictionary</em> (Scott Foresman / Company, 1935).</p>
</div>
<p>But it does not seem to have made the cut in Webster’s 1950 New Twentieth Century Dictionary:</p>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/Webster’s_New_20th_Century_Dictionary_1950.png"
alt="Noah Webster and Harold Whitehall, Webster’s New Twentieth Century Dictionary of the English Language, Unabridged (The Publishers Guild, 1950)"
/>
<p>Noah Webster and Harold Whitehall, <em>Webster’s New Twentieth Century Dictionary of the English Language, Unabridged</em> (The Publishers Guild, 1950).</p>
</div>
<p>Webster’s New World Dictionary, however, included it:</p>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/Webster’s_New_World_Dictionary.png"
alt="Noah Webster, Webster’s New World Dictionary of the American Language, Encyclopedic Edition (World Publishing Company, 1951)"
/>
<p>Noah Webster, <em>Webster’s New World Dictionary of the American Language, Encyclopedic Edition</em> (World Publishing Company, 1951).</p>
</div>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/Webster’s_New_International_Dictionary_2nd_Ed.png"
alt="Noah Webster, New international Dictionary of the English Language (G. & C. Merriam Company, 1955)"
/>
<p>Noah Webster, <em>New international Dictionary of the English Language</em> (G. & C. Merriam Company, 1955).</p>
</div>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/Chambers’s_20th_Century_Dictionary_1959.png"
alt="William Geddie, Chambers’s Twentieth Century Dictionary (W. & R. Chambers, 1959)"
/>
<p>William Geddie, <em>Chambers’s Twentieth Century Dictionary</em> (W. & R. Chambers, 1959).</p>
</div>
<p>The British Advanced Learner’s dictionary omitted it in 1961:</p>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/Advanced_Learner’s_Dictionary_1961.png"
alt="Albert Sydney Hornby, E.V. Gatenby, and H. Wakefield, The Advanced Learner’s Dictionary of Current English (Oxford University Press, 1961)"
/>
<p>Albert Sydney Hornby, E.V. Gatenby, and H. Wakefield, <em>The Advanced Learner’s Dictionary of Current English</em> (Oxford University Press, 1961).</p>
</div>
<p>So did the rather obscure bilingual Dictionary of English Style:</p>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/Dictionary_of_English_Style_1961.png"
alt="Albrecht Reum, Dictionary of English Style (Weber, 1961)"
/>
<p>Albrecht Reum, <em>Dictionary of English Style</em> (Weber, 1961).</p>
</div>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/Webster’s_7th_New_Collegiate_Dictionary_1965.png"
alt="Noah Webster, Webster’s Seventh New Collegiate Dictionary (G. & C. Merriam Company, 1965)"
/>
<p>Noah Webster, <em>Webster’s Seventh New Collegiate Dictionary</em> (G. & C. Merriam Company, 1965).</p>
</div>
<p>The British <em>Fowler</em> usage dictionary recognized it in 1965:</p>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/Fowler’s_Modern_English_Usage_2nd_Ed.png"
alt="Fowler’s Modern English Usage: 2nd Edition (Clarendon Press, 1965)"
/>
<p><em>Fowler’s Modern English Usage: 2nd Edition</em> (Clarendon Press, 1965).</p>
</div>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/Fowler’s_Modern_English_Usage_2nd_Ed.png"
alt="Philip Babcock Gove, Webster’s Third New International Dictionary of the English Language, Unabridged (G. & C. Merriam Co., 1966)"
/>
<p>Philip Babcock Gove, <em>Webster’s Third New International Dictionary of the English Language, Unabridged</em> (G. & C. Merriam Co., 1966).</p>
</div>
<p>Even as late as 1970, the <em>American Funk <span class="amp">&</span> Wagnalls Standard Dictionary</em> (that is just its name) still omitted it:</p>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/Standard_Dictionary_of_the_English_Language_1970.png"
alt="Funk & Wagnalls Company, Funk & Wagnalls Standard Dictionary of the English Language (Funk & Wagnalls, New York, 1970)"
/>
<p>Funk & Wagnalls Company, <em>Funk & Wagnalls Standard Dictionary of the English Language</em> (Funk & Wagnalls, New York, 1970).</p>
</div>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/American_Heritage_Dictionary_of_the_English_Language_1971.png"
alt="William Morris, The American Heritage Dictionary of the English Language (American Heritage Publishing Company, 1971)"
/>
<p>William Morris, <em>The American Heritage Dictionary of the English Language</em> (American Heritage Publishing Company, 1971).</p>
</div>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/Random_House_Dictionary_of_the_English_Language_1973.png"
alt="Jess M. Stein, The Random House Dictionary of the English Language (Random House, 1973)"
/>
<p>Jess M. Stein, <em>The Random House Dictionary of the English Language</em> (Random House, 1973).</p>
</div>
<p>The <em>American Heritage Dictionary</em> with its 1979 edition, of course, continued to include it:</p>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/American_Heritage_Dictionary_New_College_Ed_1979.png"
alt="William Morris, The American Heritage Dictionary of the English Language (American Heritage Publishing Company, 1979)"
/>
<p>William Morris, <em>The American Heritage Dictionary of the English Language<em> (American Heritage Publishing Company, 1979).</p>
</div>
<p>The <em>Oxford Advanced Learner’s Dictionary</em> of 1980, however, still omitted it. Since this is the newest Advanced Learner’s Dictionary in my collection, let it be revealed that the current version, which is available online at <a href="http://www.oxfordadvancedlearnersdictionary.com/">oxfordadvancedlearnersdictionary.com</a>, lists transitive and intransitive use side by side.</p>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/Oxford_Advanced_Learner’s_Dictionary_1980.png"
alt="Anthony Paul Cowie, A. C. Gimson, and Albert Sydney Hornby, Oxford Advanced Learner’s Dictionary of Current English (Oxford University Press, 1980)"
/>
<p>Anthony Paul Cowie, A. C. Gimson, and Albert Sydney Hornby, <em>Oxford Advanced Learner’s Dictionary of Current English</em> (Oxford University Press, 1980).</p>
</div>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/Dictionary_of_Contemporary_English_1984.png"
alt="Paul Procter, Longman Dictionary of Contemporary English (Longman, 1984)"
/>
<p>Paul Procter, <em>Longman Dictionary of Contemporary English</em> (Longman, 1984).</p>
</div>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/The_Random_House_Dictionary_of_the_English_Language.png"
alt="Stuart Berg Flexner, The Random House Dictionary of the English Language (Random House, 1987)"
/>
<p>Stuart Berg Flexner, <em>The Random House Dictionary of the English Language</em> (Random House, 1987).</p>
</div>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/American_Heritage_Dictionary_3rd_Ed.png"
alt="Anne H. Soukhanov, The American Heritage Dictionary of the English Language (Houghton Mifflin, 1992)"
/>
<p>Anne H. Soukhanov, <em>The American Heritage Dictionary of the English Language</em> (Houghton Mifflin, 1992).</p>
</div>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/Collins_Cobuild_English_Dictionary.png"
alt="John Sinclair, Collins COBUILD English Dictionary (HarperCollins, 1995)"
/>
<p>John Sinclair, <em>Collins COBUILD English Dictionary</em> (HarperCollins, 1995).</p>
</div>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/Longman_Dictionary_of_Contemporary_English.png"
alt="Della Summers and Adam Gadsby, Longman Dictionary of Contemporary English (Longman, 1995)"
/>
<p>Della Summers and Adam Gadsby, <em>Longman Dictionary of Contemporary English</em> (Longman, 1995).</p>
</div>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/Cambridge_International_Dictionary_of_English.png"
alt="Paul Procter, Cambridge International Dictionary of English (Cambridge University Press, 1995)"
/>
<p>Paul Procter, <em>Cambridge International Dictionary of English</em> (Cambridge University Press, 1995).</p>
</div>
<p><em>Merriam-Webster’s Dictionary of English Usage</em>, freely available through <a href="http://books.google.com/books?id=2yJusP0vrdgC">Google Books</a>, says (p. 947):</p>
<blockquote>
<p>Use of warn as an intransitive verb is common and widespread, although it is actually a fairly recent development. It seems to have originated in American English in the early 20th century. American Dictionaries have recognized the intransitive <em>warn</em> as standard for many years, but several prominent British dictionaries continue to omit it, suggesting that it may occur less commonly in British than in American English. (On the other hand, Sir Ernest Gowers noted in his 1965 revision of Fowler that “intransitive use … is now common in journalism.”) Few commentators have warned against its use. Those who have would revise the preceding sentence so that <em>warned</em> has a direct object, making it something like “Few commentators have warned their readers against its use.” Such revision is unnecessary and may even be awkward or misleading in contexts where the warning is directed generally rather than to a specific individual or group.</p>
<p>“… which warns of the approach of spring” —Mary Austin, <em>Starry Adventure</em>, 1931</p>
<p>“But he warns against what he thinks might well turn out to be a fatal illusion” —Alain Locke, <em>Key Reporter</em>, Autumn 1951</p>
<p>“Often in the years that followed, Gottwald warned against the rising tide of Fascist parties” —<em>Current Biography 1948</em></p>
</blockquote>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/The_New_Fowler’s.png"
alt="Henry Watson Fowler and R. W. Burchfield, The New Fowler’s Modern English Usage, 3rd Edition (Oxford University Press, 1996)"
/>
<p>Henry Watson Fowler and R. W. Burchfield, <em>The New Fowler’s Modern English Usage</em>, 3rd Edition (Oxford University Press, 1996).</p>
</div>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/The_New_Oxford_American_Dictionary.png"
alt="Elizabeth Jewell and Frank R. Abate, The New Oxford American Dictionary (Oxford University Press, 2001)"
/>
<p>Elizabeth Jewell and Frank R. Abate, <em>The New Oxford American Dictionary</em> (Oxford University Press, 2001).</p>
</div>
<div class="lightbox-block align-center half-width">
<img
src="/images/intransitive-warn/dictionaries/New_Oxford_American_Dictionary_2005.png"
alt="Erin McKean, The New Oxford American Dictionary (Oxford University Press, 2005)"
/>
<p>Erin McKean, <em>The New Oxford American Dictionary</em> (Oxford University Press, 2005).</p>
</div>
<p>(<em>The American Heritage Dictionary</em> and the <em>Century Dictionary</em> and several more can be searched at <a href="http://www.wordnik.com/words/warn">Wordnik</a>.)</p>
<p>At the time that I write this, February 21, 2011, Wiktionary lists four citations for intransitive <em>to warn</em>, among them one from a 1526 translation of Bible verses Galatians , 9–10 by William Tyndale: “then Iames Cephas and Iohn … agreed with vs that we shuld preache amonge the Hethen and they amonge the Iewes: <em>warnynge</em> only that we shulde remember the poore.” <em>The Oxford English Dictionary</em> describes this usage as absolute.</p>
<p>The entry in the <a href="http://www.merriam-webster.com/dictionary/warn">Merriam-Webster Online Dictionary</a>, which cites one intransitive example (“‘This won’t be easy,’ he warned.”), looks similar to the one in the <em>American Heritage Dictionary</em>, but in <a href="http://www.learnersdictionary.com/search/warn"><em>Merriam-Webster’s Learner’s Dictionary</em></a> there are a few more examples, among them the following intransitive uses: “The company has warned (investors) that its profits are likely to be lower in the coming year,” “The book warns about/of the dangers of not getting enough exercise,” and “She warns against making changes too quickly.”</p>
<p>Originally, I wanted to close with a funny remark on an anagram of <em>warn</em>, but the only one I could find is <a href="http://www.oed.com/view/Entry/158711"><em>rawn</em></a>, a term for “the roe or ovaries of a female fish (or shellfish)”. If you can think of a good closing remark or if you have found errors in this document, please leave a comment on my blog. Thanks for your interest, and I hope you had a seizable amount of fun reading this.</p>
<script type="text/javascript">if (!document.getElementById('mathjaxscript_pelican_#%@#$@#')) {
var align = "center",
indent = "0em",
linebreak = "false";
if (false) {
align = (screen.width < 768) ? "left" : align;
indent = (screen.width < 768) ? "0em" : indent;
linebreak = (screen.width < 768) ? 'true' : linebreak;
}
var mathjaxscript = document.createElement('script');
mathjaxscript.id = 'mathjaxscript_pelican_#%@#$@#';
mathjaxscript.type = 'text/javascript';
mathjaxscript.src = 'https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.3/latest.js?config=TeX-AMS-MML_HTMLorMML';
var configscript = document.createElement('script');
configscript.type = 'text/x-mathjax-config';
configscript[(window.opera ? "innerHTML" : "text")] =
"MathJax.Hub.Config({" +
" config: ['MMLorHTML.js']," +
" TeX: { extensions: ['AMSmath.js','AMSsymbols.js','noErrors.js','noUndefined.js'], equationNumbers: { autoNumber: 'none' } }," +
" jax: ['input/TeX','input/MathML','output/HTML-CSS']," +
" extensions: ['tex2jax.js','mml2jax.js','MathMenu.js','MathZoom.js']," +
" displayAlign: '"+ align +"'," +
" displayIndent: '"+ indent +"'," +
" showMathMenu: true," +
" messageStyle: 'normal'," +
" tex2jax: { " +
" inlineMath: [ ['\\\\(','\\\\)'] ], " +
" displayMath: [ ['$$','$$'] ]," +
" processEscapes: true," +
" preview: 'TeX'," +
" }, " +
" 'HTML-CSS': { " +
" availableFonts: ['STIX', 'TeX']," +
" preferredFont: 'STIX'," +
" styles: { '.MathJax_Display, .MathJax .mo, .MathJax .mi, .MathJax .mn': {color: 'inherit ! important'} }," +
" linebreaks: { automatic: "+ linebreak +", width: '90% container' }," +
" }, " +
"}); " +
"if ('default' !== 'default') {" +
"MathJax.Hub.Register.StartupHook('HTML-CSS Jax Ready',function () {" +
"var VARIANT = MathJax.OutputJax['HTML-CSS'].FONTDATA.VARIANT;" +
"VARIANT['normal'].fonts.unshift('MathJax_default');" +
"VARIANT['bold'].fonts.unshift('MathJax_default-bold');" +
"VARIANT['italic'].fonts.unshift('MathJax_default-italic');" +
"VARIANT['-tex-mathit'].fonts.unshift('MathJax_default-italic');" +
"});" +
"MathJax.Hub.Register.StartupHook('SVG Jax Ready',function () {" +
"var VARIANT = MathJax.OutputJax.SVG.FONTDATA.VARIANT;" +
"VARIANT['normal'].fonts.unshift('MathJax_default');" +
"VARIANT['bold'].fonts.unshift('MathJax_default-bold');" +
"VARIANT['italic'].fonts.unshift('MathJax_default-italic');" +
"VARIANT['-tex-mathit'].fonts.unshift('MathJax_default-italic');" +
"});" +
"}";
(document.body || document.getElementsByTagName('head')[0]).appendChild(configscript);
(document.body || document.getElementsByTagName('head')[0]).appendChild(mathjaxscript);
}
</script>Late Bill2010-02-12T19:51:00+00:002010-02-12T19:51:00+00:00Dawn Dreschertag:impartial-priorities.org,2010-02-12:/late-bill.html<p>A liar’s prequel to Langston Hughes’s short story <em>Early Autumn</em>.</p><p>You have to know, dear reader, that Bill and I had been best friends for
years before all this happened. To honor this friendship I will at no
point in this account resort to falsehood. I will clarify why he did
what he did and I expect you, my reader, to trust me as you would trust
your own best friend.</p>
<p>Bill had always been a rather solitary person, but when his beloved
brother suddenly died he was truly alone. Luckily, at the same time I
joined his high school class and soon our congenial personalities and
our shared dream of studying law and becoming attorneys bonded us in
such a profound way that we met almost every afternoon at his place to
do homework and to debate recent legal opinions of supreme court judges.
It was I who taught him to forget his brother.</p>
<p>One morning, it was the year 19— and we had just graduated, a letter came
that he had been accepted into a college in Ohio. He had already gotten
an acceptance from his first choice college, but when he learned that I,
too, had received the same letter, his decision was clear. At the time,
even I deemed this coincidence auspicious; in retrospect it was ominous.</p>
<p>In college we shared one room. At first there was only one bed in it, so
we asked around whether it was possible to get a second bed. The answer
Bill usually got was a brisk “No,” which, to us, sounded almost British
in that Ohio accent. Finally, and with a few bottles of liquor, the
janitor could be bribed into providing a spare bed for me. It took the
three of us almost half an hour to carry it up the stairs along the main
hallway, through a few ogival arches and into the smaller corridor that
led to our room. Also, we had to forgo the luxury of the extra
table—besides the desk—to fit it in. It was worth all the trouble though,
for it was softer and less worn than the other bed. We sometimes had
bets to determine who would sleep in which.</p>
<p>The first two semesters went by quickly. Bill had performed well in his
exams and all the studying still left time for us to take long strolls.
Long they were necessarily for it took us some time to get out of the
city and to the more picturesque countryside or the suburbs. When he
lacked the time, Bill usually contented himself with shorter walks to
certain places in the vicinity. So it was on the day he met Mary. The
third semester had just begun and he still had classes later that day,
making a long stroll impractical. Instead he went to a wide hilly meadow
not far from the college. There was an old gate, or rather the remnants
of one. It must have had a roof in its days of glory, and of the wall
that had once stood to its sides, little was left, so you could just
walk around it. That day however Bill felt like walking through it,
imagining a roof on top of it and its structure sound and sturdy. Then
he fought his way through waist-high grass until he reached one of the
weathered rocks, where he sat down and began reading in a textbook. A
mere page or two later and only a few yards off, suddenly, but without
startling him, she revealed her presence: She lifted her head over the
tall grass and their eyes met for the first time. In a romantically
subdued register they talked for hours—Bill missed all his classes.</p>
<p>The economics department, where Mary was studying, was situated close to
our wing. This “our” however had already begun to crumble. We spent a
lot less time together since he had met Mary; I could not possibly be
around when he was with her. In fact I was unable. But I was saying that
her department had not been far off and so Bill met her frequently,
often serendipitously. At the same time the effort he had to put into
his studies increased significantly. No longer could he go on the
strolls he so enjoyed; they had been replaced by brisk walks from the
lecture hall to his room and back again. Accordingly, whenever he met
with Mary, he knew that it meant missing some lecture or deadline—a
drawback he accepted all to readily.</p>
<p>Since you, dear reader, probably never have met Mary, you have no idea
of her uncanny gift for empathy and her keen insight into Bill, into his
situation. She had noticed how he was pressed for time, how his
timetable had gotten so very tight and rigid and how his coffee
consumption had increased. Yet, when they talked about it, neither of
them was ready to cut back on their time together. The only solution
which seemed feasible was to plan their rendezvous, to meet to a
specific time at a specific place. From ten p.m. onward every evening
was Bill’s suggestion, and of course on weekends, when they could meet
by day. The place, they quickly decided, would be a room at the top of
one of the towers—they had discovered it a mere week earlier. On one
side, part of its wall had recently fallen in and the room seemed to be
used as a dump for decrepit chairs and tables. For the winter they would
have to find a different place, but during the remaining warm nights of
the year, when the silvery light of the moon drenched the room’s
desultory assembly of furniture, they relished every second spent there.
When the moon failed to join them, however, they usually left the then
subfusc and shadowy place to go somewhere else; often to that meadow
where they first met, or—especially when it was raining—just through the
vast university buildings on a quest to discover its obscurest corners
and innermost recesses.</p>
<p>Sleep became one of Bill’s foremost concerns for he slept enough only
during the weekend. Nevertheless, the moment he and Mary convened in
that room and they sat together on the floor with a perfect view of the
whole premises, he was instantaneously wide awake. It is hard to imagine
how they always had something to tell and learn from each other, be it
in words or in this tranquil silence of perfect rapport which is no less
eloquent. On certain nights, especially when the moon was nearly full,
they bolted the door to their storeroom and, with the moonlight
glistening in pearls on her silky skin, surely gave rise to some tales
of the tower being haunted. Many weeks went by like that.</p>
<p>Sometime after the first month, when Bill’s initial infatuation had
segued seamlessly into something he first more tentatively but soon
confidently described to himself as love, something new crept in with
it. Back in the days when he lived with his parents, he had always been
afraid someone would want to separate the two of us, sever our
friendship ties, take me away from him. To prevent this from happening
we had reached the agreement to see to it that our friendship remain a
secret as far as that was possible. Even from his parents. While in
school we had just never worked in groups together and had met only
clandestinely when no one was around; meeting in a similarly furtive
manner after school at his place had proven harder, at least when his
parents were home. Luckily it seemed family policy not to talk about
personal matters, so the issue never came up. In him, this shared fear
that we might lose each other if the wrong people learned about us,
transformed radically as his relationship with Mary progressed. I can
only hypothesize that when Mary, unwittingly, began to tear him away
from me, and we both had to go our separate ways in college, this fear
just lingered on, finding a new target in his love for her. Part of him
knew those anguishes to be unsubstantiated for he trusted Mary as he had
once trusted me, but as much as that part tried to shrug them off, his
feeling of inadequacy and the dread, someone more intrepid and valiant
might snatch Mary away from him just kept getting stronger. Also there
was a nagging suspicion that when this happened, Mary might not even be
aware of ita thought that failed to strike him as absurd. It grew until
it surrounded him like a wall, closed, eventually, even to me.</p>
<p>One Monday shortly after noon I came into our room startling Bill who
had just napped with his head resting on one of my volumes by Alfred
Swaine Taylor. Absently he stacked it onto the Robert Christison while
he turned around. (I needed those books for a course of mine. Purely
educational.) Unfortunately, the reason I had come was to relay to Bill
that a new project—having to do something with European stock markets and
different timezones—forced Mary to work every Monday and Tuesday night,
so that they would not be able to meet on those two days for the moment.
As much as he regretted that change, I was relieved to learn that still
he was glad of the extra sleep it got him.</p>
<p>In the following morning, Bill became aware of something disturbing.
True, he was always very tired during the day, despite the copious
quantities of coffee he drank, but usually his sleep was rather light,
nonetheless. The last few nights however, he had slept very deeply. So
deeply that today, not even the long and loud ringing of his alarm clock
had sufficed to wake him. Yet he felt tired and exhausted as if he had
spent the whole night studying. Had someone sedated him? And if so, why?
Also he wondered why I had not seen to it that he woke up—had I even been
back for the night? Moreover there were those dreams. In all three
nights he had had dreams somehow circling around Mary though she was
seldom actually visible. They were hazy and indiscriminate, but at the
same time close enough to waking life that they hardly faded during the
day. Almost like ordinary memories.</p>
<p>Sitting on his bed, reconstructing last night’s dream, two originally
far separate concepts in his brain suddenly snapped together: Whenever
he was with Mary, Bill never thought of me. He searched his memory for
any shred of a sentence with which he had ever mentioned me to Mary—and
came up with nothing. How is it possible that throughout the hundreds of
hours they had spent together he had never mentioned his one long-time
friend? From this point onward his attitude toward me grew even more
askance and he tried to channel his confused fears and suspicions by
making them explicit in a short story he wrote. Not wanting me to read
it, he kept it always in his briefcase, a measure which proved
ineffectual. His story, set in the future, paints a dismal picture of
Mary’s and his fate after Mary has suddenly been married away from him
by some anonymous perpetrator—me?</p>
<p>Over the course of the whole next week he slept only two or three hours
in total and then also very lightly again. As a result, his behavior
became increasingly erratic. One rainy afternoon two college staff
inspected our room. They asked Bill whether he really needed the extra
bed. After he implored them on his knees not to take it away, they,
looking somewhat vexed, acquiesced and left. It kept raining. It had
been raining a lot that month, so you should think that by then Bill and
Mary should have been experts in the architecture of the buildings. Far
from it. In fact that Thursday they found a hall, connecting two
sections of the basement, one which they had never seen before. Due to
construction work, several doors stood open which were usually barred,
and trails of cement dust led through them. Apart from its novelty to
them however, the interior was not especially thrilling. After a few
minutes of walking through that basement, Bill passed by a low door and,
pulling it open, looked on an assortment of dusty furniture and stacks
of curly, yellowed paper. At first he wanted to show this seemingly
forgotten old storeroom to Mary but she had already walked far ahead, so
he thought better of it and followed after her, quickening his pace.</p>
<p>Since the previous Tuesday, Bill had been drinking very little for fear
someone—especially I—who might have slipped some soporific drug in his
water or even his coffee the week before, could try it again. His
defense, of course, could only be to pay utmost attention to what he
drank. In the cafeteria he was very cautious indeed. First he checked
that his mug was absolutely clean, then he waited for five other
people—none of whom he knew—to fill their mugs from the pot, watching them
attentively, before he filled his. While he was eating—again he had been
very cautious in taking from the right bowls—thrice people he hardly knew
came up to his table. To each of them he made it perfectly clear how
attentive he was. Of course he let them eat at the table, after all
Bill’s is a very polite demeanor, but not for a second did he take his
eyes off them. They had to know that no one could put drugs in his food
without his noticing! They soon gave up and moved to another table. He won.</p>
<p>Despite all his efforts he had yet another night of that deep unrelaxing
sleep that left him wearier than the evening before. Again he had this
curious type of dream in surroundings he did not recognize. He
remembered a plant with conspicuous blue flowers seemingly in some sort
of greenhouse. This plant he did recognize. He had seen it in a book I
must have been so unwary to forget, open, on the desk. Suddenly it
occurred to him, and he scolded himself for not having thought of it
earlier, that his tormentor might even use injections to drug him. If
administered dexterously one may not even feel the pinch of the needle.
Without intending to, the man he envisioned holding the syringe in his
hand was me. Henceforth he was careful to swerve to the other side of
the hall whenever anyone came as menacingly close as a couple yards, a
distance he took care to maintain to all other people during lectures.
Right after one of those lectures he was walking along the main hallway
on the ground floor when a certain detail about it caught his heightened
attention: The hall was much longer than the day before. At that time he
thought that someone might indeed have extended it, after all there was
a lot of construction work going on around the building, later however
Bill was rather inclined to attribute the elongation, the perceived
elongation that is, to some sort of mirror magic at its ends. Seconds
later he stumbled and fell.</p>
<p>He had become used to these fits of dizziness; while he had lost a
distinct feeling of thirst, these fits still served as a reminder for
him to drink something. But now was not the time. No compromises! No
risk-taking! He cannot recall what had taken place after he fell, but
not even an hour later he was back at the cafeteria with a new plan.
First he talked to someone very loudly. Hence, if I was stealthily
spying, I was bound to notice him. Then he ordered a glass of water,
positioned himself in the center of the room, so to be in plain view of
his tormentor incognito, and drank. Ha! In fact he only pretended to
drink, no drop went past his lips. Having thus fooled his adversary, he
thought himself safe.</p>
<p>Then night approached. With considerable mental effort and lots of
coffee Bill kept himself in that state he called awake. For fear of
falling into it, he also kept away from the bed as if it were a canyon.
A stump of a candle shed its feeble light upon a pile of legal texts
which he knew for an inextricable labyrinth of paragraphs but only saw
as an army of blurry ants thronged together—in the semi-darkness
seemingly moving. Long already, he had been unable to concentrate on
anything. Whatever thought he tried to focus on, immediately dispersed
like a ball of spiderlings touched. The candle went out. He waited for
me. An vexing tune played in his mind, haunted him, played in piercingly
high keys, then higher and higher. His thoughts screamed against the
noise. He drove his fingernails into the wooden desk, as the darkness
was slowly sucking him in. Half an hour later he had become one with the
pitch black nothingness that filled the windowless room. He waited,
wondering where I was and who was with me. Ethereal shapes, black on
black, were leaping at him and in the repetitive carnage of inner voices
he was yanked down boundless cataracts of despair. His words became
insults to their meaning. Gushes of brackish water on seething grease.
Eyes open or closed, indistinct memories crept up to him, vanishing the
moment he noticed them. Then a picture appeared, like one from his
dreams, and was gone. It had shown Mary from behind, silhouetted against
the sky. But where? He tried to haul pieces of the picture back into his
consciousness but he had no control of his thoughts. They all slipped
away, then the thoughts that reminded him not to let them slip away,
slipped away also. Again and again in never-ending cycles he recalled
what he had tried to recall and lost it again. Then, much later, the
picture flickered past his inner eye once more, and without conscious
recognition of anything in particular he felt that its setting had to be
the lofty room where they had met so often. The feeling abode with him.
He clung to it as a shipwrecked sailor might to a chunk of wood. Still I
had not arrived; now Bill gave up waiting and, rediscovering his limbs,
made toward the door.</p>
<p>Every lightning flash flooded the deserted main hall with its icy
travesty of sunlight. While he was climbing the stairs, one hand on the
rail, the shadow of a cobweb was cast upon the wall. He reached the top,
the door was ajar, he stepped in and was alone. The missing part of the
wall had become a wall of water. As impenetrable for Bill as the
darkness of his room—and as mesmerizing. For minutes he stared into the
dark void, often a white void for a split-seconds. He remembered what he
usually saw from there and after several minutes his brain tricked him
into actually seeing parts of it: Water running down the weathered
granite of another building, trees, bent by the storm, and then the path
that led toward the meadow. His quest for me, for Mary and for
who-knew-what easily mingled with his memories of the meadow and his
rampant imagination drove him down all the staircases and out into the
overwhelming downpour. His clothes were drenched in a matter of seconds.</p>
<p>For the creeks running down his face, it was hard for him to keep his
eyes open. Yet he saw suspicious movements and raced after them. First
through muddy grass, then onto asphalt. Midnight was past. Where were we
hiding? He ran even faster. In the cold cones of the street lights the
raindrops became leaden bullets, pelleting his skin. After felt hours of
running and stumbling he reached the dilapidated gate that lead unto the
meadow. He sank to the ground. His whole body ached. Then a thought of
Mary renewed his resolution but failed to renew his strength. To escape
the muddy whirl around him, he crawled onto the remnants of masonry
adjacent to the gate and from this slightly elevated vantage point for
the first time beheld the meadow for what it really was: A cemetery.
Perhaps it was even the first cemetery the city had had. Today it lay in
ruins, overgrown and now largely flooded. Though he could not see it
from his position by the lychgate, he knew from his strolls over a year
ago that the new cemetery was only a few miles to the east. The boulders
everywhere had to be the vestiges of mausolea and tombstones. I grew
scared; I had to show myself.</p>
<p>“Confound it! Bill! You’re in the middle of a thunderstorm. You’ll get
us both killed.”</p>
<p>“What are you doing out here?” His question was accentuated by
lightning, very close. I shuttered.</p>
<p>“I’m worried about you. I know for a fact that you get little sleep and
now, moreover, you hardly drink anymore. You are my best friend, I have
to look after you.”</p>
<p>“Oh, yeah. Good job. You sedated me several times, didn’t you? What is
it you do while I sleep?”</p>
<p>“I didn’t. Never.” If I explained everything to him now, he would
believe me; he could not but acknowledge its veracity.</p>
<p>“I don’t care how you got me to sleep like that. What are your insidious
plans for Mary?”</p>
<p>I was torn between telling him everything at once, and trying to weasel
my way out of it and getting us back inside somehow. He grew impatient
with my hesitance and laboriously crawled upon a larger chunk of wall
remnants. With some effort he stood erect and cried toward the sky:</p>
<p>“You care about be? Then leave Mary alone or I’ll forfeit my life to
Zeus!” He was not joking.</p>
<p>I saw no other way: “Bill, it’s very simple. I am you. We share one
mind. If you’re dehydrated I’m feeling awful, too.”</p>
<p>“What the fuck are you talking about?” was Bill’s reply, yet I knew that
really he was only surprised not to be more surprised.</p>
<p>“Yours is also my body, Bill. And I won’t have it fried by lightning out
here; nor am I keen on getting hypothermia.”</p>
<p>“But Mary, she’s real, isn’t she?” He almost lost balance. “And that
project of hers, you made that up, didn’t you?”</p>
<p>“Yes, I might have lied about a few things, but it was for our both
good, for our friendship, the one she jeopardizes.”</p>
<p>“O, I know what you’re up to. Dare not! Not her! Not again!I’d rather
kill us both.”</p>
<p>Obviously he was delirious. I cannot remember what I replied but I saw
that he did not believe me at all. He surged with anger and perhaps that
was how I finally got my foot into the door. Bill passed out; I took over.</p>
<p>Gradually he woke up. Then this process reached a certain level and he
snapped awake. It was still very early. Naturally he had no notion of
how he had gotten back to his room, and when he had hung his wet clothes
to dry. To him, though, it was all crystal clear: His friend, nay,
adversary, was trying to take Mary away from him. He, Bill, had the
power to stop him; and he had an idea. Still in his bed clothes he ran
through the vacant halls of the law school, then down to the basement
where they were renovating. The site was deserted. While he stacked
bricks onto a barrow so high he could hardly control it, only thoughts
of Mary could briefly pierce his wild frenzy. Several barrowfuls of
bricks he brought thus through the low door into the forgotten storeroom
which he had discovered so recently. Then mortar and a container
accelerator. He filled several buckets with hot water from a near
restroom. Finally he purloined a selection of trowels, entered the room,
pulled the door shut, removed the handle and went to work. He was afraid
I might manage to stop him, so he hurried. A mere hour later the low
door was walled up.</p>
<p>I cannot say why it is exactly that I felt compelled to write all this.
Perhaps it is to vindicate poor Bill, in case someone finds my—our—mortal
remains one day. Yet still I have hope that my cries will be heard
sooner or later. Bill had stayed very focused for the whole day and the
following night before I was able to take control. The mortar, aided by
the accelerator, aided by the hot water had already proven impenetrable
at that point. I think I will hide these sheets around here
somewhere—hopefully, he will not find them.</p>