I think most efforts to grow EA grow both Longtermist and non-Longtermist elements of EA.
Would it make sense for you to distinguish between growing longtermist and non-longtermist elements of EA? You prioritise longtermist interventions over animal welfare and global health ones.
I do agree though that the Longtermist effects of animal welfare are much bigger than near-term benefits.
Likewise for global health interventions? If so, animal welfare and global health interventions are also longtermist in the sense their longterm benefits are much larger than their nearterm benefits? In this case, it would be better to avoid terms like āLongtermist political donationsā, āother Longtermist donationsā, and āLongtermist careersā, which could refer to animal welfare and global health interventions? If I thought the longterm effects of the vast majority of interventions were much larger than their nearterm effects, I would simply refer to the areas of the interventions instead of highlighting they are longtermist.
Yeah, itās rough and thereās not a perfect method.
You did not make any comparison of the longterm benefits of different areas? At this point, I am just looking for one comparison, not a perfect one. Why not defaulting to the basic intuition that the benefits over the next few decades are a good proxy for the total benefits, which suggests global health interventions are more cost-effective than ones aiming to decrease the risk of global catastrophes?
I wouldnāt be that extreme [GiveWellās top charities being 2*10^-30 times as cost-effective as some longtermist interventions], but I have some credence that something like that extreme of a tradeoff is right.
I doubt GiveWellās top charities are 2*10^-30 times as cost-effective as some longtermist interventions.
Donāt think we need one for those sorts of things, just like we donāt need a quantitative model suggesting, say, the odds of Vance being the next president are around 30%.
There is huge disagreement about the risk of extinction due to AI. Below are the predictions from superforecasters made in the Existential Risk Persuasion Tournament (XPT). There is nothing like that disagreement about the probability of Vance being the next president of the United States (US) assuming he is one of the 2 final candidates, which makes huge deviations from 50 % extremely unlikely.
I think weāve talked elsewhere about why I donāt buy that soil nematodes etc make us totally clueless about which animal welfare interventions are good.
The last point I made in my comment was not about this. I meant that global health interventions may be much more cost-effective than animal welfare interventions even neglecting effects on soil invertebrates (like nematodes).
Would it make sense for you to distinguish between growing longtermist and non-longtermist elements of EA? You prioritise longtermist interventions over animal welfare and global health ones.
Likewise for global health interventions? If so, animal welfare and global health interventions are also longtermist in the sense their longterm benefits are much larger than their nearterm benefits? In this case, it would be better to avoid terms like āLongtermist political donationsā, āother Longtermist donationsā, and āLongtermist careersā, which could refer to animal welfare and global health interventions? If I thought the longterm effects of the vast majority of interventions were much larger than their nearterm effects, I would simply refer to the areas of the interventions instead of highlighting they are longtermist.
You did not make any comparison of the longterm benefits of different areas? At this point, I am just looking for one comparison, not a perfect one. Why not defaulting to the basic intuition that the benefits over the next few decades are a good proxy for the total benefits, which suggests global health interventions are more cost-effective than ones aiming to decrease the risk of global catastrophes?
I doubt GiveWellās top charities are 2*10^-30 times as cost-effective as some longtermist interventions.
There is huge disagreement about the risk of extinction due to AI. Below are the predictions from superforecasters made in the Existential Risk Persuasion Tournament (XPT). There is nothing like that disagreement about the probability of Vance being the next president of the United States (US) assuming he is one of the 2 final candidates, which makes huge deviations from 50 % extremely unlikely.
The last point I made in my comment was not about this. I meant that global health interventions may be much more cost-effective than animal welfare interventions even neglecting effects on soil invertebrates (like nematodes).