I think because of the flow of money into EA, it feels like some people have updated towards cost-effectiveness and counterfactual reasoning being less important than before.
I disagree with that view—I think that cost-effectiveness and counterfactual reasoning are exactly as important as they were before, the only change should be that our cost-effectiveness bar for funding things should be slightly lower. It remains true that small amounts of money from people in rich countries can dramatically raise someone’s income via GiveDirectly, save a life through AMF or improve someone’s subjective wellbeing via StrongMinds, so the obligation to be cost-effective remains very strong.
I think not enough effort seems to go into estimating and optimising the cost-effectiveness of the community building side of EA, probably in part because this is difficult to do, highly uncertain and thus prone to motivated reasoning.
But I think much more effort should go into estimating and optimising the cost-effectiveness of EA community building anyway. Some concrete examples I’d like to see—do we think we overspent / underspent on EAGxs and EAGs this year?
My thoughts on this:
I think because of the flow of money into EA, it feels like some people have updated towards cost-effectiveness and counterfactual reasoning being less important than before.
I disagree with that view—I think that cost-effectiveness and counterfactual reasoning are exactly as important as they were before, the only change should be that our cost-effectiveness bar for funding things should be slightly lower. It remains true that small amounts of money from people in rich countries can dramatically raise someone’s income via GiveDirectly, save a life through AMF or improve someone’s subjective wellbeing via StrongMinds, so the obligation to be cost-effective remains very strong.
I think not enough effort seems to go into estimating and optimising the cost-effectiveness of the community building side of EA, probably in part because this is difficult to do, highly uncertain and thus prone to motivated reasoning.
But I think much more effort should go into estimating and optimising the cost-effectiveness of EA community building anyway. Some concrete examples I’d like to see—do we think we overspent / underspent on EAGxs and EAGs this year?