Announcing Manifund Regrants

Thanks, Ozzie! I expect we'll be chatting more about working together. Some point-by-point replies:

I imagine some of this information should eventually be private to donors. Like, the medical expenses one.

Yeah, idk, this is one of the areas I think I'm idiosyncratic; I think in an ideal world that almost no information would be kept private (maybe excepting infohazards like bomb construction techniques.) I think this attitude has been incredibly good for Manifold's growth and development, and am patterning my approach to Manifund based on this.

In any case, right now we're one small player amidst a variety of EA funders, so I'm not very worried about our radical transparency stance. I do also think there's a good chance we change courses (maybe 30% in the next 12 months?) based on grantee/regrantor/donor feedback.

I'd want to eventually see Slack/Discord channels for each regrantor and their donors, or some similar setup.

Interesting idea! My best guess is that our specific donor for Round 1 doesn't have the time/interest to do this, but if other donors are looking for a more hands-on role, we'd be open to that as well!

I imagine some regranters would eventually work in teams

This sounds promising. Right now we're in a pretty proto-mvp stage, trying to catch up to what other funders have figured out. I do think collaboration across grantmakers is generally good; though caveat that teamwork does impose coordination costs, and for small (<$20k?) grants I'd lean towards moving faster.

As money amounts increase, I'd like to see regranters getting paid. It's tough work.

We went back and forth on this. Right now all our regrantors are volunteers, while FF paid theirs. I think getting the specific compensation scheme right is tricky (pay by grant? by hour? by % commission?) and I suspect FF's may have had bad incentives + optics, so we opted to launch first and revisit this later.

(to be continued)

Part 2:
 

one full-time person who just gathers information about teams/people, and relays that to regranters.

Open to this as well, if there are specific individuals you think would do this well as a fulltime job! (Though, in this model I'm not quite sure what the role of the regrantor is -- just to scout out opportunities?)

As I wrote about here, I'm generally a lot more enthusiastic about supporting sizeable organizations than tiny ones. I'd hope that this could be a good project to fund projects within sizeable organizations.

Yes, we'd love for individual projects to apply through Manifund regranting, because we think we could pay for those in a pretty lightweight manner. I like RP's model of having individual projects apply for funding themselves.

On the larger point of large vs small orgs, I'm personally someone who thrived moving from Google => Streamlit (series B startup) => founding Manifold. I agree EA is pretty strange, and think we could benefit from overhauling the ecosystem to be more like the tech scene, which seems to be much better than EA at executing at their objectives. This could be either management practices at high-growth startups like Stripe (see High Growth Handbook) applied to the ecosystem as a whole; or instituting a more efficient venture ecosystem like YC for seed-stage orgs.

I want to see more attention on reforming/improving the core aspects/community/bureaucracy of EA. These grantmakers seem very AI safety focused. 

I agree with you on the importance of organizational reform, though it's extra unclear if that kind of thing is addressable by a regranting program (vs someone enacting change from within OpenPhil). Perhaps we'll ourselves address this if/when Manifund itself represents a significant chunk of EA funding, but that seems like a version 2 problem while we're currently at v0.

Ideally it could be possible to have ratings/reviewers of how the regranters are to work with. 

Yeah! One of my pet ideas is "Yelp for EA orgs"; useful for individual regrantors, but also for EA grantmakers as a whole? Could integrate into our platform if we end up becoming a directory for all EA projects and funders... (also a v1/v2 problem, though regrantor reviews might be v0)

I'm not very excited by Impact Certificates. More "traditional" grantmaking seems much better.

Agree to disagree for now! I have written up what I think are outstanding issues with impact certs, but I'm still bullish on the idea (esp with a large eg AI Safety yearly prize set up). Would love to do a podcast/debate on this sometime!

One obvious failure mode is that regranters might not actually spend much of their money. It might be difficult to get good groups to apply. This is not easy work. 

Yes, this is one of my top concerns with our regranting program so far; we've issued budgets to regrantors a few weeks ago and so far grantmaking pace is slower than expected. I do think this public launch will help though, as it solicits more applications for them to look over.

Thank you again for the extensive feedback!

Haha, thanks for bringing this up. One correction, Rachel and I are married (as of last month).

A quick background on this is that around February, Scott Alexander of Astral Codex Ten asked Manifold to set up an impact market to be able to run the ACX Forecasting Minigrants round (which is the site you see now at https://manifund.org). At the time, our existing team on Manifold were already occupied, and I had seen Rachel's work on various programming projects such as openbook.fyi. After careful consideration, and checking in with both Scott and Manifold for Charity's board of advisors, I decided to bring her on for a 6-week consulting engagement, which we've since renewed and turned into a fulltime offer.

Obviously, we recognize the potential conflicts of interest and didn't make this decision lightly. My best judgement is that Rachel has done fantastically in this position so far, comparable to eg what I would expect a new grad at Google. (If you're technical, I invite you to judge her commits on our open source repository).

The $50k regrantor budget that both her and I have are primarily to allow us to dogfood our own site. For the two of us to build a useful product for regrantors to use, it's important that we have on-the-ground experience of making regrants ourselves. You're also welcome to evaluate specifically the two grants she's recommended so far (to Rachel Freedman and the Donations List Website)!

First want to say that I was also pretty uncomfortable about this, and initially told Austin I didn't want to do it—you’re right that I am not qualified to be a grant maker, at least in the sense that I would not be hired as one in another context. I don’t deserve whatever status that role happens to bestow upon me, and I don’t particularly want the power.

That said, me being a regrantor has made the UX much better, since I’m basically the sole person pushing code and dogfooding is so powerful: it’s changed the prompting questions on write-ups, the way projects in need of more funding are displayed, the way funding targets are specified, and lots of other tiny things that were a bit uglier or higher friction or simply broken before. Less concretely, my model of what it feels like to write a publicly visible grant writeup, to search for giving opportunities, to select grant amounts, etc. has become more vivid, which I expect to be strategically useful going forward. And it's only been a few weeks so far.

So, I agree it’s bad optically (I agree-voted your comment), but ultimately think this was a good call. Especially because the counterfactual of having not given me the $50k is not that it would be going to some better-evaluated grants than the ones I’ve made, but that it would be sitting in a bank account, and (obviously) I think the grants I’ve made are better than that.

I think this is an important point and it's good you brought it up. But I think you were unnecessarily harsh, and you could have done the same good with less harshness. 

A version I would have preferred:

How were grantors chosen, and in particular why was Rachel -- included? She seems young, without relevant accomplishments, and has no experience grant making. I believe she's dating the Manifold founder, which makes this a question of conflict of interest.

This leaves open the option you're missing something, and that this could be remedied. You can always become harsh later if you're not given a satisfactory answer (it looks like you feel you were).

 FWIW I don't feel fully satisfied by Austin's answer but I also think the person with standing to object is the anonymous donor. One reason I'm excited about this infrastructure is that it allows people to accumulate track records and for funders to choose who to delegate to. 

Depends who chose the regrantors? If the two humans most behind Manifund chose them, I agree it's odd/bad that they chose themselves (independent of their relationship). If donors chose them that seems surprising but fine.

Edit: oh they say "We chose regrantors." Hmm. I think choosing themselves is a mistake.

Indeed, Qualy has graciously agreed to serve as a regrantor for Manifund. Excited to see what they recommend!

Yes, we should be able to do this! Let us know if you have a specific PhD or study in mind :)

Thanks, we hope so too! (To be clear, we also have a lot of respect for centralized grantmaking orgs and the work they do; and have received funding through some in EA such as LTFF and SFF.)

I think one of the difficulties here is that Austin and his colleagues are doing this through their personal and professional networks, just like anyone starting a project. Unless the Manifold team were to make some sort of a blinded application process (which has it's own costs and burdens), I don't really see a way around the obstacle of recruiting mainly from people you know. An organization can change this overtime. But despite how much I dislike how much harder it is to get funded if you're not well-connected in the Bay Area community, not relying on connections at the start strikes me as very challenging.

I'll also share my opinion that Austin has the right response here: he is open to suggestions and willing to consider them.

Hi Gideon, I'm a regranter and I live in Latin America. This has basically no effect on what I choose to fund.

I also care a lot more about animals than most (maybe all) the regranters. 

The thing you want to select for in a group of regranters is the diversity of knowledge of opportunities and some intellectual diversity of approaches to longtermism. It seems like Manifund did a decent job here? I think i was selected for my forecasting on Manifold but there are others who work in pandemic preparedness and a lot of people work in different ways no AI safety (which is sort of what the donor wanted).

We're happy to consider more diverse regrantors -- if you have specific candidates in mind, please send them this launch post, or make an intro to us (austin@manifund.org)!

Thanks for the feedback! Geographic diversity in particular does seem pretty important for getting the most out of regranting—much dissatisfaction about the current funding situation comes from how much harder it is to get funded if you're not well-connected in the Bay Area community.

I'm disappointed that we currently don't even have much UK representation, since that's the other EA hub. This is largely because we are based in the SF so are better connected here. As Austin said, happy to hear suggestions for people connected in other areas who could better surface new opportunities!

Without claiming that the other dimensions of diversity aren't important, I see geographic and cause area diversity as the most important dimensions.

Regranters are much more likely to grant to people they either know personally or through their networks. They are also much more likely to grant to projects they understand.

If I had to, I would guess that there is maybe one Republican leaning regrantor in there. Do you think we should actively recruit more Trump supporting regrantors, as this would also increase diversity? It also looks like almost all of the regrantors have university degrees. Do you think there should be more regrantors with no formal education? Approximately 14% of the world adult population is illiterate after all. 

Which of the humanities should people be recruited from?