The new influx of philanthropic capital that Nan Ransohoff wrote of (possibly ~$37–100B per year) invites us to think about an important Q: how do you get philanthropic incentives to be closer to market efficient?
Entrepreneurs are often making a big bet that success=generational wealth, failure=not much. Can philanthropy replicate this with prizes? Weird structures of equity in hedge funds that depend on the future going well? Would love to hear people’s incentive structure ideas.
In a recent thread I reported the alleged existence of a “philanthropic overhang” ultimately sourced to @Ryan Kidd, the existence of which was then disputed by Anthropic’s @Zac Hatfield-Dodds. So, does this anticipated influx of philanthropy even exist? I would still like to see the steelman version of both sides of this debate-that-hasn’t-happened.
As for optimizing philanthropic structure, my own situation is sufficiently remote from the sedate world of intergenerational deliberation that I care about entirely different things. For me it’s more like: hey philanthropist, do you hear a voice from over that cliff? That’s me. There’s a few of us stuck on a ledge down here. If you threw down a few $10k, I could probably get one of these ledge dwellers to safety, and able to play her role in the few final moments we have before AI takeover is complete. But if that’s too much to ask, maybe there are some Human Intelligence Tasks we can do for you, for a few hundred or thousand dollars? There aren’t many job vacancies out here on the ledge, and sometimes it gets windy...
The claim I said was false is about the present: “The field [of AI safety] is swamped with billions of undeployed dollars”. This is entirely consistent with substantial growth of funding in the near future!
Please be careful when describing what I’ve said; loose paraphrases etc have been a problem for me before and are a strong disincentive from any public engagement (for me and for others).
The new influx of philanthropic capital that Nan Ransohoff wrote of (possibly ~$37–100B per year) invites us to think about an important Q: how do you get philanthropic incentives to be closer to market efficient?
Entrepreneurs are often making a big bet that success=generational wealth, failure=not much. Can philanthropy replicate this with prizes? Weird structures of equity in hedge funds that depend on the future going well? Would love to hear people’s incentive structure ideas.
In a recent thread I reported the alleged existence of a “philanthropic overhang” ultimately sourced to @Ryan Kidd, the existence of which was then disputed by Anthropic’s @Zac Hatfield-Dodds. So, does this anticipated influx of philanthropy even exist? I would still like to see the steelman version of both sides of this debate-that-hasn’t-happened.
As for optimizing philanthropic structure, my own situation is sufficiently remote from the sedate world of intergenerational deliberation that I care about entirely different things. For me it’s more like: hey philanthropist, do you hear a voice from over that cliff? That’s me. There’s a few of us stuck on a ledge down here. If you threw down a few $10k, I could probably get one of these ledge dwellers to safety, and able to play her role in the few final moments we have before AI takeover is complete. But if that’s too much to ask, maybe there are some Human Intelligence Tasks we can do for you, for a few hundred or thousand dollars? There aren’t many job vacancies out here on the ledge, and sometimes it gets windy...The claim I said was false is about the present: “The field [of AI safety] is swamped with billions of undeployed dollars”. This is entirely consistent with substantial growth of funding in the near future!
Please be careful when describing what I’ve said; loose paraphrases etc have been a problem for me before and are a strong disincentive from any public engagement (for me and for others).
The justification for a philanthropic overhang is:
Existing large funders 2-3xing their annual budgets this year;
The pledged commitments from Anthropic founders and employee DAFs.;
The posts like these from cG staff;
The OpenAI Foundation pledging $1B in 2026;
Nan Ransohoff’s calculation that “AI will likely add ~$37-100B per year in intended philanthropic spend in the near future.”