Leo’s experimental microgranting program
0 proposals
0 active projects
$0
Original description here
This is a microgranting program with minimal bureaucracy. You spend an hour writing a short description of what you want to do, and if I like it I’ll give you $10k to make it happen. If you’ve ever hesitated to apply to any grant because you didn’t want to deal with a complicated process, or you’re worried you’re too junior, then this is the grant for you.
I want to fund projects to make AGI go well for all of humanity. If you have a cool idea that only helps indirectly, you should lean towards submitting it anyways. I don’t care whether it has AI vibes, I only care whether it actually makes the future go well. I want unusual ambitious ideas - things that others might be hesitant to fund. I want things that are big if true, rather than incremental improvements. For the purpose of this program, I’m ruling thinkers in not out.
Big grantmaking programs spend a lot of time per grant. The goal of this program is to make many small grants. Just like startups fail, most projects I fund will crash and burn.
Important note: This program is an experiment. I’m doing this because I’m interested in figuring out whether grantmaking is something I want to do more of in the future, and to see if this is a viable model for grantmaking. This program may change dramatically, or cease to exist at any moment if I get bored of grantmaking - I’ll update this doc if that happens, but any grants sitting in the pipeline might never even end up being considered.
Spend about an hour filling out the form below. Use up to 1000 words, though don’t fill up the word limit needlessly. Write clearly. I want to give you money; make my job easy!
Don’t bullshit and exaggerate and add buzzwords. I know that strategy works on HR people, but I’ve been doing research for a decade and seen many resumes. I know the game of inflating accomplishments. I can tell. It only hurts your chances. Be genuine.
Every letter of your application should be generated by your brain and hand typed. If your application is AI written or edited it will be automatically rejected.
You don’t have to have done alignment stuff in the past. If you haven’t, provide examples of cool things you’ve done in other areas. Focus on substance over credentials. I don’t care where you went to college, or whether your paper is in arxiv or neurips.
Feel free to submit multiple applications if you have multiple ideas.
Grants from this program should not be cited as evidence of endorsement when applying to other things, or as a shiny trophy on your linkedin/resume/job application, etc. (Why? There’s an epistemic cascade where grantmakers defer to other grantmakers, and so people seek easy credentials and leverage them to get more prestigious credentials. I don’t want to deal with these social dynamics.)
I may recuse myself from grants if I feel I have a conflict of interest. In that case, I will try to forward it to someone I trust to be good at making the decision.
Send applications to this form. Do not apply for microgrants through any other means - do not text/email me your application, do not approach me irl to pitch an idea. There is no guarantee of how long I will take to reply, if ever.
Once you’ve spent your grant money, you can apply for another one, with a progress update. Your project doesn’t have to be a success to apply for more money! Solid negative results or even a good explanation of what you learned is good.
If you are a grantmaker, like this program, and want to give me money to regrant, please let me know. By default I will be giving my own money but I will be able to give more grants if I have more money.
If you know people who would benefit from a microgrant, please refer them and have them note somewhere in their application that you referred them, along with your contact info. I am paying a $100 bounty per referral that leads to an approved microgrant.