“EA” doesn’t have a talent gap. Different causes have different gaps.

There’s been a lot of dis­cus­sion and dis­agree­ment over whether EA has a tal­ent or a money gap. Some peo­ple have been say­ing there’s not that large of a fund­ing gap any­more and that peo­ple should be us­ing their tal­ent di­rectly in­stead. On the other hand, oth­ers have been say­ing that there definitely still is a fund­ing gap.

I think both par­ties are right, and the rea­son for the mi­s­un­der­stand­ing is that we have been refer­ring to the en­tire EA move­ment in­stead of break­ing it down by cause area. In this blog post I do so and demon­strate why we’re like the blind men touch­ing differ­ent parts of the elephant, and how if we put all of it to­gether, we’ll be able to make much bet­ter de­ci­sions.

Poverty

  • Ta­lent gap—Small (~10 peo­ple)

  • Money gap—Large (~$86 mil­lion = ~1720 peo­ple do­ing E2G)

An­i­mal rights

  • Ta­lent gap—Large (~100+ peo­ple)

  • Money gap—Mixed (de­pends on agree­ment with Lewis Bol­lard)

Ar­tifi­cial intelligence

  • Ta­lent gap—Mid­dle (~50 peo­ple)

  • Money gap—Small (most pro­jects are very well funded)

Meta or­ga­ni­za­tions (that fall out­side of one of the above ar­eas.)

  • Ta­lent gap—Small (~20 peo­ple)

  • Money gap—Mixed (de­pends on agree­ment with Nick Beck­stead)

I am not ex­tremely con­fi­dent on all these num­bers (par­tic­u­larly the size of the AI tal­ent gap), but I am con­fi­dent of the broader claim that the gaps are differ­ent be­tween cause ar­eas, and we would all benefit from mak­ing that dis­tinc­tion in pub­lic dis­course. I am happy to up­date these as peo­ple make good ar­gu­ments for them in the com­ments. Below I’ll go into fur­ther de­tails of how I came to these es­ti­mates.

Poverty tal­ent gap

In my ex­pe­rience, poverty or­ga­ni­za­tions gen­er­ally hire out­side of the EA move­ment for many roles. There are still small gaps for some poverty or­ga­ni­za­tions hiring man­age­ment and lead­er­ship roles from the EA pool (~4). There are also some gaps in op­er­a­tional tal­ent (~2). A part of this gap also comes from the pos­si­bil­ity of found­ing more effec­tive poverty char­i­ties (~4), such as a to­bacco tax­a­tion or con­di­tional cash trans­fer char­ity, like what has been done with Char­ity Science Health and For­tify Health.

Poverty money gap

The gap for money in poverty is huge, even when only look­ing at char­i­ties sig­nifi­cantly stronger than Give Directly, whose gap is very large and ar­guably vir­tu­ally un­limited. The gap is close to $100 mil­lion af­ter Good Ven­tures funds its por­tion. There is also rea­son to ex­pect this gap to grow with re­cent changes in Good Ven­ture’s fund­ing plans and a strong group of in­cu­ba­tion char­i­ties in GiveWell’s sys­tem. This gap only grows if you think there are strong op­por­tu­ni­ties in poverty out­side of GiveWell’s list. As­sum­ing donat­ing 50% of a $100,000 salary, it would eas­ily take 1,720 peo­ple do­ing E2G to fill this gap. And that is not even in­clud­ing new Givewell in­cu­bated/​recom­mended char­i­ties!

An­i­mal rights tal­ent gap

The tal­ent gap for an­i­mal rights is very large. Many AR or­ga­ni­za­tions are hiring and try­ing to grow as fast as pos­si­ble. There is also con­sid­er­able scope for en­trepreneur­ship and found­ing new and effec­tive an­i­mal rights or­ga­ni­za­tions. The an­i­mal rights com­mu­nity as a whole is very small and the num­ber of EAs in the move­ment is even more limited.

An­i­mal rights money gap

His­tor­i­cally an­i­mal rights has been chron­i­cally ham­pered by in­suffi­cient fund­ing across the move­ment. How­ever the en­trance of Open Phil to the area has cre­ated a very differ­ent situ­a­tion. I now cat­e­go­rize the fund­ing gaps as mixed. The fund­ing is fairly cen­tral­ized be­tween Open Phil and the AR Funds be­ing run by the same per­son (Lewis), which con­trols nearly 50% of all fund­ing in AR. If you have strong agree­ment with Lewis about the pri­ori­ties in the area, I would say the fund­ing gap is small. How­ever, if you have very differ­ent views, then the fund­ing gap could be seen as large.

Ar­tifi­cial in­tel­li­gence tal­ent gap

The tal­ent gap for Ar­tifi­cial in­tel­li­gence is mid­dling, with many or­ga­ni­za­tions in the field in need of re­searchers as well as some gaps in meta-or­ga­ni­za­tions fo­cus­ing on meta-re­search. There are also sig­nifi­cant gaps in op­er­a­tional tal­ent to help the sup­port struc­tures of these or­ga­ni­za­tions.

Ar­tifi­cial in­tel­li­gence money gap

The money gap for AI or­ga­ni­za­tions seems very small, with even large fun­ders be­ing turned away from many pro­jects. Many or­ga­ni­za­tions have very large amounts of fund­ing, and given the re­cent changes in pub­lic­ity, much like an­i­mal rights, AI went from be­ing chron­i­cally un­der­funded to well funded in al­most all ar­eas. Fur­ther­more, due to the fairly wide spread of fun­ders, even peo­ple with more unique per­spec­tives on AI will find it hard to find good gaps.

Meta or­ga­ni­za­tions tal­ent gap

Im­por­tantly in this sec­tion, I mostly con­sider meta or­ga­ni­za­tions that do not fall un­der an­other cause area. For ex­am­ple, ACE would fall un­der an­i­mal rights, not un­der meta. The tal­ent gap for these or­ga­ni­za­tions gen­er­ally seems small, with some posted roles in lead­er­ship (~7), op­er­a­tions (~3), re­search (~3) and other gen­eral roles (~3) across or­ga­ni­za­tions. There seems to be some scope for found­ing new char­i­ties as well (~4).

Meta or­ga­ni­za­tions money gap

Much like an­i­mal rights, there’s a lot of cen­tral­iza­tion of fund­ing with a hand­ful of fun­ders con­trol­ling a very high per­centage of to­tal fund­ing. Like in an­i­mal rights, there is one per­son who con­trols the EA funds on meta-or­ga­ni­za­tions and is the lead in­ves­ti­ga­tor for Open Phil. Thus I think an EA’s per­spec­tives on fund­ing gaps will largely de­pend on how well their views al­ign with Nick Beck­stead’s. This gap can range from very small to mod­er­ate sized (low mil­lions) de­pend­ing on how broadly you define meta-or­ga­ni­za­tions.

Over­all, as you can see, the tal­ent and money gaps vary largely de­pend­ing on the cause. If you think poverty is the high­est im­pact area, earn­ing to give is a very good choice. On the other hand, if you think an­i­mal rights is the best, figur­ing out how to best give your tal­ents might be a bet­ter way for­ward. If you agree with Lewis, that is. Re­gard­less of what cause you think is high­est pri­or­ity and what you think the gaps truly are, break­ing them down by cause area will help ev­ery­body make bet­ter de­ci­sions.