The article: [“Are you really in a race? The Cautionary Tales of Szilárd and Ellsberg”](https://forum.effectivealtruism.org/posts/cXBznkfoPJAjacFoT/are-you-really-in-a-race-the-cautionary-tales-of-szilard-and)

## Fundamental Justification

### What outcomes were to be expected before you started the project? What were unusually good and unusually bad possible outcomes? (Please avoid hindsight bias and take the interests of all sentient beings into account.)

My intention for this blog post was for it to be fairly clear and memorable, aimed at a general audience not just for an AI governance audience - especially perhaps a machine learning researcher who doesn't know much about history. The main takeaway I wanted wasn't for people to think "this is the most common/likely outcome" but rather to add a historic example to their repertoire that they can refer to. It was supposed to be a cautionary tale, a prompt to people to think not "all sprints are wrong" but rather "wait am I in an Ellsberg situation?" - and if so to have some general, sensible recommendations and questions to ask. 

### What actual outcomes are you aware of?

Several people have left comments indicating they found it useful, such as Carl Shulman, Rohin Shah, Pablo Stafforini and others. I have presented on it to GovAI and CHAI. Christian Ruhl contacted me, and we wrote a new version of the piece for a general non-EA audience, published in the Bulletin of the Atomic Scientists:
https://thebulletin.org/2022/07/why-policy-makers-should-beware-claims-of-new-arms-races

### Who can make a legitimate claim to a fraction of the impact, and have you talked to them?

In the post I acknowledged colleagues at CSER, CFI, GovAI and Rethink Priorities – especially Di Cooke, Matthijs Maas, Helen Toner, Alex Lintz and Markus Anderljung – for feedback.

### Who are the current owners of the impact and what fraction do they each own?

Haydn Belfield - 100%

## Procedural Questions

### What is your minimum valuation under which you’ll not sell any shares in your impact?

I am unsure.

### What would you have done had there been no chance to get retro funding? (This helps us assess our impact but has no effect on our evaluation of the certificate’s impact.)

I would have written this anyway - I found out about the impact markets idea afterwards.

### What can we improve about this process?

Some indication about justifications for different ranges for minimum valuation.


<p>The article: <a href="https://forum.effectivealtruism.org/posts/cXBznkfoPJAjacFoT/are-you-really-in-a-race-the-cautionary-tales-of-szilard-and">“Are you really in a race? The Cautionary Tales of Szilárd and Ellsberg”</a></p>
<h2 id="fundamental-justification">Fundamental Justification</h2>
<h3 id="what-outcomes-were-to-be-expected-before-you-started-the-project-what-were-unusually-good-and-unusually-bad-possible-outcomes-please-avoid-hindsight-bias-and-take-the-interests-of-all-sentient-beings-into-account">What outcomes were to be expected before you started the project? What were unusually good and unusually bad possible outcomes? (Please avoid hindsight bias and take the interests of all sentient beings into account.)</h3>
<p>My intention for this blog post was for it to be fairly clear and memorable, aimed at a general audience not just for an AI governance audience - especially perhaps a machine learning researcher who doesn't know much about history. The main takeaway I wanted wasn't for people to think "this is the most common/likely outcome" but rather to add a historic example to their repertoire that they can refer to. It was supposed to be a cautionary tale, a prompt to people to think not "all sprints are wrong" but rather "wait am I in an Ellsberg situation?" - and if so to have some general, sensible recommendations and questions to ask. </p>
<h3 id="what-actual-outcomes-are-you-aware-of">What actual outcomes are you aware of?</h3>
<p>Several people have left comments indicating they found it useful, such as Carl Shulman, Rohin Shah, Pablo Stafforini and others. I have presented on it to GovAI and CHAI. Christian Ruhl contacted me, and we wrote a new version of the piece for a general non-EA audience, published in the Bulletin of the Atomic Scientists:<br><a href="https://thebulletin.org/2022/07/why-policy-makers-should-beware-claims-of-new-arms-races">https://thebulletin.org/2022/07/why-policy-makers-should-beware-claims-of-new-arms-races</a></p>
<h3 id="who-can-make-a-legitimate-claim-to-a-fraction-of-the-impact-and-have-you-talked-to-them">Who can make a legitimate claim to a fraction of the impact, and have you talked to them?</h3>
<p>In the post I acknowledged colleagues at CSER, CFI, GovAI and Rethink Priorities – especially Di Cooke, Matthijs Maas, Helen Toner, Alex Lintz and Markus Anderljung – for feedback.</p>
<h3 id="who-are-the-current-owners-of-the-impact-and-what-fraction-do-they-each-own">Who are the current owners of the impact and what fraction do they each own?</h3>
<p>Haydn Belfield - 100%</p>
<h2 id="procedural-questions">Procedural Questions</h2>
<h3 id="what-is-your-minimum-valuation-under-which-youll-not-sell-any-shares-in-your-impact">What is your minimum valuation under which you’ll not sell any shares in your impact?</h3>
<p>I am unsure.</p>
<h3 id="what-would-you-have-done-had-there-been-no-chance-to-get-retro-funding-this-helps-us-assess-our-impact-but-has-no-effect-on-our-evaluation-of-the-certificates-impact">What would you have done had there been no chance to get retro funding? (This helps us assess our impact but has no effect on our evaluation of the certificate’s impact.)</h3>
<p>I would have written this anyway - I found out about the impact markets idea afterwards.</p>
<h3 id="what-can-we-improve-about-this-process">What can we improve about this process?</h3>
<p>Some indication about justifications for different ranges for minimum valuation.</p>
