EA grants available to individuals (crosspost from LessWrong)

(This is a cross­post from LessWrong; here’s the origi­nal)

I’m con­sid­er­ing ap­ply­ing for some kind of a grant from the effec­tive al­tru­ism com­mu­nity. A quick sketch of the speci­fics is here. Rae­mon replied there with a list of pos­si­bil­ities. In this post, I’ll look into each of those pos­si­bil­ities, to make this pro­cess eas­ier for who­ever comes next. In the or­der Rae­mon gave them, those are:

OpenPhil

“Can I ap­ply for a grant?
In gen­eral, we ex­pect to iden­tify most giv­ing op­por­tu­ni­ties via proac­tive search­ing and net­work­ing. We ex­pect to fund very few pro­pos­als that come to us via un­so­lic­ited con­tact. As such, we have no for­mal pro­cess for ac­cept­ing such pro­pos­als and may not re­spond to in­quiries. If you would like to sug­gest that we con­sider a grant — whether for your pro­ject or some­one else’s — please con­tact us. ”

This looks like a case where it’s at least par­tially about “who you know”. I do in fact have some con­tacts I could ap­proach in this re­gard, and I may do so as this search pro­ceeds.But this does seem like a bias that it would be good to try to re­duce. I un­der­stand that there are se­ri­ous failure modes for “too open” as well as “too closed”, but based on the above I think it cur­rently tilts to­wards the lat­ter. Per­haps a pub­li­cly-an­nounced pro­cess for com­mu­nity vet­ting? I sus­pect there are peo­ple who are qual­ified and will­ing to help sort the slush-pile that would cre­ate.

CEA (Cen­ter for Effec­tive Altru­ism)

Ap­pli­ca­tions for the cur­rent round have now closed. If you’d like to be no­tified when ap­pli­ca­tions next open, please sub­mit your con­tact in­for­ma­tion through this form.
The goal of Effec­tive Altru­ism Grants is to en­able peo­ple to work on pro­jects that will con­tribute to solv­ing some of the world’s most im­por­tant prob­lems. We are ex­cited to fund pro­jects that di­rectly con­tribute to helping oth­ers, as well as pro­jects that will en­able in­di­vi­d­u­als to gain the skills needed to do so.
...
CEA only funds pro­jects that fur­ther its char­i­ta­ble ob­jects.[1] How­ever, we wel­come ap­pli­ca­tions that may be of in­ter­est to our part­ners who are also look­ing to sup­port promis­ing pro­jects. Where ap­pro­pri­ate, we have some­times passed ap­pli­ca­tions along to those part­ners.

This would seem to be a dead end for my pur­poses in two re­gards. First, ap­pli­ca­tions are not cur­rently open, and it’s not clear when they will be. And sec­ond, this ap­pears to fo­cus on pro­jects with im­me­di­ate benefits, and not meta-level ba­sic re­search like what I pro­pose.

BERI (Berkeley Ex­is­ten­tial Risks Ini­ti­a­tive) in­di­vi­d­ual grants

BERI’s In­di­vi­d­ual Grants pro­gram fo­cuses on mak­ing grants to in­di­vi­d­u­als or teams of in­di­vi­d­u­als, rather than to or­ga­ni­za­tions. There are sev­eral types of in­di­vi­d­ual grants pro­grams that BERI ex­pects to run, such as:
In­di­vi­d­ual Pro­ject Grants are awarded to in­di­vi­d­u­als to carry out pro­jects di­rectly in ser­vice of BERI’s mis­sion.
In­di­vi­d­ual Level-Up Grants are awarded to in­di­vi­d­u­als to carry out pro­jects or in­ves­ti­ga­tions to im­prove the skills and knowl­edge of the grantee, with hopes that they will carry out valuable work for BERI’s mis­sion in the fu­ture.
What is the pro­cess for ob­tain­ing an in­di­vi­d­ual grant from BERI?
Typ­i­cally, BERI will host “rounds” for its var­i­ous in­di­vi­d­ual grants pro­grams. De­tails about how to ap­ply will be in the an­nounce­ment of the round.… If you would like to be no­tified when BERI is run­ning one of the above grants rounds, please send an email to in­di­vi­d­ual-grants@ex­is­tence.org not­ing which type of grant round you are in­ter­ested in.

Another dead end, at the mo­ment, as ap­pli­ca­tions are not open.

EA funds

There are 4 funds (Global Devel­op­ment, An­i­mal Welfare, Long-Term Fu­ture, and Effec­tive Altru­ism Meta). Of these 4, only Long-Term Fu­ture ap­pears to have a pro­cess for in­di­vi­d­ual grant ap­pli­ca­tions, linked from its main page. (Luck­ily for me, that’s the best fit for my plan any­way.)

We are par­tic­u­larly in­ter­ested in small teams and in­di­vi­d­u­als that are try­ing to get pro­jects off the ground, or that need less money than ex­ist­ing grant-mak­ing in­sti­tu­tions are likely to give out (i.e. less than ~$100k, but more than $10k). Here are a few ex­am­ples of pro­ject types that we’re open to fund­ing an in­di­vi­d­ual or group for (note that this list is not ex­haus­tive):
+ To spend a few months (per­haps dur­ing the sum­mer) to re­search an open prob­lem in AI al­ign­ment or AI strat­egy and pro­duce a few blog posts or videos on their ideas + To spend a few months build­ing a web app with the po­ten­tial to solve an op­er­a­tions bot­tle­neck at x-risk or­gani­sa­tions + To spend a few months up-skil­ling in a field to pre­pare for fu­ture work (e.g. microe­co­nomics, func­tional pro­gram­ming, etc). + To spend a year test­ing an idea that has the po­ten­tial to be built into an org.

This is definitely the most promis­ing for my pur­poses. I will be ap­ply­ing with them in the near fu­ture.

Conclusions

I’m look­ing for funds in the $10K-$100K range for a short-term pro­ject that would prob­a­bly fall through the gaps of tra­di­tional fund­ing mechanisms — an in­di­vi­d­ual ba­sic re­search pro­ject. It seems the EA com­mu­nity is try­ing to ad­dress this is­sue of fund­ing this kind of pro­ject in a way that has fewer ar­bi­trary gaps while still hav­ing rigor­ous stan­dards. Nev­er­the­less, I think that the land­scape I sur­veyed above is still frag­mented in ar­bi­trary ways, and wor­thy pro­jects are still prob­a­bly fal­ling through the gaps.

Rae­mon sug­gested in a com­ment on my ear­lier post that “some­thing I’m hop­ing can hap­pen some­time soon is for those grant­mak­ing bod­ies to build more com­mon in­fras­truc­ture so ap­ply­ing for mul­ti­ple grants isn’t so much du­pli­cated effort and the pro­cess is eas­ier to nav­i­gate, but I think that’ll be awhile”. I think that such “com­mon in­fras­truc­ture” would help a more-unified triage pro­cess so that the best pro­pos­als wouldn’t fall through the cracks. I think this benefit would be even greater than the ones Rae­mon men­tioned (less du­pli­cated effort and eas­ier nav­i­ga­tion). I un­der­stand that this re­fac­tor­ing takes time and work and prob­a­bly won’t be ready in time for my own pro­posal.