Posts

Comments

Comment by james.lucassen on A proposal for a small inducement prize platform · 2021-06-08T16:09:33.025Z · EA · GW

I am pretty excited about the potential for this idea, but I am a bit concerned about the incentives it would create. For example, I'm not sure how much I would trust a bibliography, summary, or investigation produced via bounty. I would be worried about omissions that would conflict with the conclusions of the work, since it would be quite hard for even a paid arbitrator to check for such omissions without putting in a large amount of work. I think the reason this is not currently much of a concern is precisely because there is no external incentive to produce such works - as a result, you can pretty much assume that research on the Forum is done in good faith and is complete to the best of the author's ability.

Potential ways around this that come to mind:

  • Maybe linking user profiles on this platform to the EA Forum (kind of like the Alignment Forum and LessWrong sharing accounts) would provide sufficient trust in good intentions?
  • Maybe even without that, there's still such a strong self-selection effect anyway that we can still mostly rely on trust in good intentions?
  • Maybe this only slightly limits the scope of what the platform can be used for, and preserves most of its usefulness?
Comment by james.lucassen on What key facts do you find are compelling when talking about effective altruism? · 2021-04-19T18:10:38.814Z · EA · GW

If it costs $4000 to prevent a death from malaria, malaria deaths happen at age 20 on average, and life expectancy in Africa is 62 years, then the cost per lifetime saved is $0.0109/hour. 

If you make the average US income of $15.35/hour, this means that every marginal hour you work to donate can be expected to save 1,412 hours of life, if you take the very thoroughly researched, very scalable, low-risk baseline option. If you can only donate 10% of your income, then your leverage is reduced to a mere 141.2. Just by virtue of having been born in a developed country, every hour of your time can be converted to days or weeks of additional life for someone else.

While not as insanely huge as some of the figures making the argument for longtermism, I find this figure more shocking on a psychological level because it's so simple to calculate and yields such an unexpected result. This type of calculation is what first got me interested in Singer-style EA.