I am an earlyish crypto investor who has accumulated enough to be a mid-sized grantmaker, and I intend to donate most of my money over the next 5-10 years to try and increase the chances that humanity has a wonderful future. My best guess is that this is mostly decided by whether we pass the test of AI alignment, so that’s my primary focus.

AI alignment has lots of money flowing into it, with some major organizations not running fundraisers, Zvi characterizing SFF as having “too much money”, OpenPhil expanding its grantmaking for the cause, FTX setting themselves up as another major grantmaker, and ACX reporting the LTFF’s position as:

what actually happened was that the Long Term Future Fund approached me and said “we will fund every single good AI-related proposal you get, just hand them to us, you don’t have to worry about it”

So the challenge is to find high-value funding opportunities in a crowded space.

One option would be to trust that the LTFF or whichever organization I pick will do something useful with the money, and I think this is a perfectly valid default choice. However, I suspect that as the major grantmakers are well-funded, I have a specific comparative advantage over them in allocating my funds: I have much more time per unit money to assess, advise, and mentor my grantees. It helps that I have enough of an inside view of what kinds of things might be valuable that I have some hope of noticing gold when I strike it. Additionally, I can approach people who would not normally apply to a fund.

What is my grantmaking strategy?

First, I decided what parts of the cause to focus on. I’m most interested in supporting alignment infrastructure, because I feel relatively more qualified to judge the effectiveness of interventions to improve the funnel which takes in people who don’t know about alignment in one end, takes them through increasing levels of involvement, and (when successful) ends with people who make notable contributions. I’m also excited about funding frugal people to study or do research which seems potentially promising to my inside view.

Next, I increased my surface area with places which might have good giving opportunities by involving myself with many parts of the movement. This includes Rob Miles’s Discord, AI Safety Support’s Slack, in-person communities, EleutherAI, and the LW/EA investing Discord, where there are high concentrations of relevant people, and exploring my non-LW social networks for promising people. I also fund myself to spend most of my time helping out with projects, advising people, and learning about what it takes to build things.

Then, I put out feelers towards people who are either already doing valuable work unfunded or appear to have the potential and drive to do so if they were freed of financial constraints. This generally involves getting to know them well enough that I have a decent picture of their skills, motivation structure, and life circumstances. I put some thought into the kind of work I would be most excited to see them do, then discuss this with them and offer them a ~1 year grant (usually $14k-25k, so far) as a trial. I also keep an eye open for larger projects which I might be able to kickstart.

When an impact certificate market comes into being (several promising signs on the horizon!), I intend to sell the impact of funding the successful projects and use the proceeds to continue grantmaking for longer.

Alongside sharing my models of how to grantmake in this area for others in a similar position and getting advice on it, the secondary purpose of this post is to pre-register my intent to sell impact in order to strengthen the connection between future people buying my impact and my current decisions. I’ll likely make another post in two or three years with a menu of impact purchases, once it’s more clear which ones produced something of value.

I have donated about $40,000 in the past year, and committed around $200,000 over the next two years using this strategy. I welcome comments, questions, and advice on improving it.

61

New Comment
5 comments, sorted by Click to highlight new comments since: Today at 6:11 AM

Nice! Here's my own strategy for grant making. Context: I gave away ~$120k to AI safety stuff in 2021. My source of funding is somewhat sporadic, so for example I don't expect to give much this year, but do think it's likely that in 2023 I'll have a similar or larger opportunity to make grants.

  • I'm focused on hits. So I'm willing to donate to stuff that might not succeed because the hits matter.
  • Another way to say this is I think of grants as investments rather than purchases.
  • I looked for opportunities that seem especially illegible or hard to fund or early stage and uncertain.
  • I think of this as angle-grating.
  • To that end, I also don't intend to make a lot of follow up donations. Either I got a hit and something takes off (and is able to get funding elsewhere to continue) or it flops and we move on the the next thing and try again.
  • That said, I can still provide mentorship, etc. if that's helpful to grantees beyond money. And I'd be happy to grant to the same person again for something sufficiently new and different.
  • Another way I think of it: I'm trying to scrape up the potential hits that aren't good enough risks for larger granting orgs to take on who have larger reputational risks.

Hopefully seeing how I'm approaching roughly the same problem is helpful in devising your own strategy.

A lot of that makes sense, other than I don't understand the angle-grating part?

Lol, double typo. "angel granting"

Thanks for helping in our quest to save the world! :) Would you mind linking the Discord servers in question, assuming that they’re open invitation? I wasn’t aware we had an investing-centric server, and would love to join it. Also, a question from the perspective of someone who might benefit from grant money (fairly small financial needs with the potential for a lot of free time)—what's the best way to reach out about that sort of thing?

Rob's is invite-only right now, but I have an invite and sent you it. I'm adding links to the two public ones to the main post, the investing one is private and I don't have any invites, but they're considering switching to an open invite next week so I'll get back to you.

For how to get on the radar of grantmakers, I think applying to microgrants programs like ACX+ or LTFF is a reasonable start (one of my grantees was from there ACX+). Another idea which should open up soon with impact certificates is just going ahead and doing something of value then selling your impact. Talking to AI Safety Support is also a near-universally good thing to do for people who want to help. I'd also be open to hearing more about your situation in DM.

New to LessWrong?