It'd be useful to know which grantmakers deserve credit for spotting great opportunities early. What are the best examples of grants that didn't seem very promising to the EA community at the time, but have now turned out to be much more exciting than expected?

New Answer
Ask Related Question
New Comment

9 Answers sorted by

I haven't looked closely, but from a fairly-but-not-completely uninformed perspective, Tim's allocation of part of his donor lottery winnings to the Czech Association for Effective Altruism looks prescient and potentially unusually counterfactually impactful.

I think credit should also go to those in the community who applied for and successfully won grants for projects that were not their own. Making sure a larger EA donor was aware of a particular project is one way to beat (in a very collaborative sense) the larger grant-makers. It is basically doing active grant making but at low risk and low personal ability to fund.

I know one example which is Chris Chambers work on Registered Reports. The giving opportunity was picked up on / created by Hauke Hildebrandt in a Lets Fund report. The application to fund it was made by Jacob Hilton (who had no connection to the issue but just though someone should do an application) and the grant was given by the EA Infrastructure Fund (details here).

If others have more examples of this would be great to hear.

re: Chris Chambers: Registered Reports advocacy - I think this could be a big deal if this publication format becomes the standard in the sciences:

Roughly $80,000 (including the EA funds grant) was raised to support Professor Chris Chambers’ to advocate for “Registered Reports” which improves the way research is done across all hypothesis-driven science. More than 200 journals have now adapted this publication format including Nature Communications (e.g. for epidemiology papers) and PLoS ONE[20], the second largest scientific journal and by one measure the most important scientific journal.[21] Some of this was likely due to our grantees advocacy.[22] Recently, Chambers said that “some of the most useful and flexible funding I've received has been donated by hundreds of generous members of the public (& small orgs) via our Lets-Fund.Org supported [crowdfunding campaign]”.[23] 

There are also signs of adoption of the Registered Reports format in high-impact areas such as development economics,[24] health economics,[25],[26] climate change,[27] and catastrophic biological bisks[28]. One recent study provides initial evidence of higher research quality... (read more)

Off the top of my head, I'm aware of some donors supporting the following organizations a few months before large funders did:

  • Global Challenges Project (formerly Student Career Team)
  • Rethink Priorities' longtermist and meta teams
  • maybe Czech EA
  • maybe AI Impacts
  • Buying laptops or providing travel funding for promising individuals

I think some donors may just have gotten lucky. But I do think there's a small number of donors (maybe 10-30% of those who attempt?) who managed to 'beat' the broader funding ecosystem.

I spent two minutes making a list of projects that grantmakers donors funded that large funders (e.g. EA Funds) deliberately didn't fund. The list ended up being roughly twice as long as the above one (8 orgs + some smaller things).

I think the framing of good grantmaking as "spotting great opportunities early" is precisely how EA gets beat.

Fast Grants seems to have been hugely impactful for a fairly small amount of money, the trick is that the grantees weren't even asking, there was no institution to give no, and no cost-effectiveness estimate to run. It's a somewhat more entrepreneurial approach to grantmaking. It's not that EA thought it wasn't very promising, it's that EA didn't even see the opportunity.

I think it's worth noting that a ton of OpenPhil's portfolio would score really poorly along conventional EA metrics. They argue as much in this piece. So of course the community collectively gets credit because OpenPhil identifies as EA, but it's worth noting that their "hits based giving" approach divers substantially from more conventional EA-style (quantitative QALY/cost-effectiveness) analysis and asking what that should mean for the movement more generally.

The coronavirus Fast Grants were great, but their competitive advantage seems to have been that they were they first (and fastest) people to move in a crisis.

The overall Emergent Ventures idea is interesting and worth exploring (I say, while running a copy of it), but has it had proven cost-effective impact yet? I haven't been following the people involved but I don't remember MR formally following up.

2AppliedDivinityStudies8mo
There's a list of winners here [https://newscience.org/emergent-ventures-winners/], but I'm not sure how you would judge counterfactual impact. With a lot of these, it's difficult to demonstrate that the grantee would have been unable to do their work without the grant. At the very least, I think Alexey was fairly poor when he received the grant and would have had to get a day job otherwise.

So of course the community collectively gets credit because OpenPhil identifies as EA, but it's worth noting that their "hits based giving" approach divers substantially from more conventional EA-style (quantitative QALY/cost-effectiveness) analysis and asking what that should mean for the movement more generally.

My impression is that most major EA funding bodies, bar Givewell, are mostly following a hits based giving approach nowadays. Eg EA Funds are pretty explicit about this. I definitely agree with the underlying point about weaknesses of traditional EA methods, but I'm not sure this implies a deep question for the movement, vs a question that's already fairly internalised

4Linch8mo
The Global Health and Dev side of Open Phil still gives a lot to less hits-based giving.
2AppliedDivinityStudies8mo
Sure, but outside of OpenPhil, GiveWell is the vast majority of EA spending right? Not a grant-making organization, but as another example, the Rethink Priorities report on Charter Cities [https://forum.effectivealtruism.org/posts/EpaSZWQkAy9apupoD/intervention-report-charter-cities] seemed fairly "traditional EA" style analysis.
2Neel Nanda8mo
Sure. But I think the story there was that Open Phil intentionally split off to pursue this much more aggressive approach, and GiveWell is more traditional charity focused/requires high standards of evidence. And I think having prominent orgs doing each strategy is actually pretty great? They just fit into different niches

I think Fast Grants may not be great on a longtermist worldview (though it might still be good in terms of capacity-building, hmm), and there are few competent EA grantmakers with a neartermist, human-centric worldview.

I think it's worth noting that a ton of OpenPhil's portfolio would score really poorly along conventional EA metrics. They argue as much in this piece.

To be clear, to the extent your claim is true, giving money to things that ex ante have a lower cost-effectiveness than Givewell top charities + have low information value is more of a strike against Open Phil than it is against the idea of using cost-effectiveness analysis. 

4 years ago Our World In Data had problems getting funded after the Gates Foundation didn‘t extend their support. Apparently it worked out in the end (probably not due to the 50€ I chipped in back then, though).

From Max Roser‘s AMA:

We have some support from readers who donated to our work and that could keep us going for some time. Buy as I was mentioning elsewhere here, we have no longer term funding for the time after mid-November, but we are talking to two possible funders. But if these things shouldn't work out it would be a disaster for us yes.

I wrote OpenPhil about it back then but it didn‘t fall into any of their cause areas unfortunately.

This is really interesting, does anyone know which funder(s) stepped in and saved the day?

4MaxRa8mo
Good question. Their website has a list of possible candidates:
6NunoSempere8mo
This is missing the meaty part of the story. Our World in Data was the org that Dominic Cummings got in trouble for trying to throw £500k [https://www.theguardian.com/politics/2021/jun/22/cummings-told-officials-to-bypass-procedures-on-530k-grant-to-data-team-leak-reveals] at it, because he thought that their work was essential in the middle of the pandemic. Eventually it seems like they declined the grant and got funding through the "Department of Health and Social Care" instead. See Dominic Cummings' side of the story here [https://dominiccummings.substack.com/p/the-uks-procurement-system-and-judicial] .

Grants for UK policy work focused on the long-term. In particular the APPG for Future Generations and CLTR.

 

The bigger EA donors have typically been (and still are) incredibly sceptical of funding policy work and there has been almost no* funding in this space for smaller new projects. (the SAF Fund might change that, although funding rounds are infrequent). Big donors tend to be sceptical due to:

  • Concern about the reputational risk of this work.
  • Lack of expertise in donor organisations to vet such projects.

 

So, out of my personal donations one I am particularly proud of is probably offering initial funding to get the APPG for Future Generations growing and encouraging the people running it to hire and expand.  (I don’t know the full funding for these orgs but I believe, as well as my donation, that they were initially funded by donors who were enthusiastic about policy but outside the big EA grantmakers.)

 

I believe (but I am bias) that this was a great call and that UK policy work has gone shockingly well. The APPG for Future Generations and CLTR have lead to policy wins feeding into to the UK government making resilience a priority, improving the UK's ability to manage unexpected extreme risks, improving the UK's preparedness for future pandemics, and improving how UK policy makers consider the long-term.

(It is also notable that there has not been reputational problems and some conversations suggests that the risk-conscious approach taken by here may actually have reduced some reputational risks.)

 

*** Disclaimer: I work for the APPG. Views all my own but expect bias. ***

 * This area does seem strangely neglected still.  For example based on grant pay-outs and feedback received I believe that:

  • Open Phil does not fund new longtermist-adjacent policy advocacy projects – with the exception of CSET (a project run by a very senior political figure). Although they do give to existing policy projects such as existing large biosecurity organisations.
  • The Long-Term Future Fund receives numerous applications for policy  projects (I know of a few and am sure there have been more)  but does not appear to have funded
... (read more)
9weeatquince8mo
I think policy is a domain where there are still many opportunities for spotting and funding projects early, and having more impact than donating to the EA Funds. As mentioned it may be very hard for the EA Funds to do this kind of work. That said I think this only applies if if you have (or know and trust someone) with relevant expertise. There are risks and I think plenty of the projects I have come across I thought, this should not to be funded. (I had planned to write a whole post on this and on how to do active grant-making well as a small donor – not sure if I will have time but maybe)
5Neel Nanda8mo
I would love to read this post (especially any insights that might transfer to someone with AI Safety expertise, but not much in other areas of EA!). Do you think there's much value in small donors giving to areas they don't know much about? Especially in areas with potential high downside risk like policy. Eg, is the average value of the marginal "not fully funded" policy project obviously positive or negative?
2weeatquince8mo
Hey Neel. It doesn't add much (still not had time for a top level post) but in case helpful a bit more of my reasoning is set out here [https://forum.effectivealtruism.org/posts/jkMgbiXcR4MDNFs3z/where-are-you-donating-in-2021-and-why?commentId=JWCEGcisffgngqTTw] .

Early EA community building in London.

Back in 2015 there were no funds available for local community building, no funds from CEA or community grants or EA Funds etc. There was general scepticism about if it was worth funding such local work.

Kudos and many thanks go to Alex Gordon-Brown for initially suggesting that he could fund a local community organiser and to Kit Harris for initial funding and to many others in the London community for additional contributions.

This seems to have been a great call (although I might be bias as it was my job). Certainly since then there has been a massive scale up in resources for local community organisers all around the world. There has been a focus on community and in-person outreach as a particularly effective way to spread EA ideas without the message becoming garbled (i.e. it is more high fidelity). And of course the London community grew a lot. But it took individual donors actively pushing for this to happen, and funding it when others were sceptical, in order to lead the way. 

If you count seed funding as part of an incubation program I think Charity Entrepreneurship (CE) is probably the absolute king of "spotting great opportunities early".

Their earliest global poverty charities, Fortify Health and Suvita, incubated and seed funded by CE, went on to get GiveWell Incubation grants. Fish Welfare Initiative and Animal Advocacy Careers were incubated and seed funded by CE and have been funded by the EA Animal Welfare Fund. Happier Lives Institute was incubated and seed funded by CE and has now been funded by the Infrastructure Fund. I expect there are other examples and newer CE charities will do just as well.

If donors want to spot great opportunities early, being on the CE seed funding network (or even just donating to CE) seems a good way to do it. (If interested get in touch).

*** Disclaimer: I now work for CE. Views all my own but expect bias. ***
 

Not sure if this counts, but many small donors supported MIRI while GiveWell/OpenPhil were very skeptical.

Do you think MIRI at that time was exciting? Do you think other people should think that? (Genuinely asking, and not even necessarily from a MIRI-skeptical position. It seems possible that MIRI at the time was pretty unproductive and unpromising, and also that MIRI at different times was better, and that funding didn't necessarily help that transition take place). 

2Larks8mo
Yes at the time I did; this [https://forum.effectivealtruism.org/posts/nSot23sAjoZRgaEwa/2016-ai-risk-literature-review-and-charity-comparison#MIRI] was written just after OpenPhil's first grant, but it reflects an opinion I had held for many years by that point.