Comments
Coefficient Giving just gave GiveWell $1 billion. Where should other donors give now?
Linkpost for my Substack piece, lightly adapted. Disagreement very welcome.
What happened
Coefficient Giving (cG) announced a $1 billion gift to GiveWell on 23rd July. This increases a previous commitment of $175m for 2026. cG say this could be a one-off surge, but it has implications for other donors either way. Both organisations have lowered their funding thresholds this year (GiveWell from 8x their benchmark to 6x, Coefficient from 2,000x their benchmark to 1,000x). Even if GiveWell hit their aim of grantmaking that will "approach or even exceed $1 billion this year", cG's gift is roughly the size of everything GiveWell expects to grant this year.
They don't specifically mention AI wealth, but I think it's fairly clear what they mean when they say "much of the future funding we're anticipating depends on the valuations of some extremely volatile assets." They also say that a belief in the potential of AI to transform global health affects their decision. In essence, AI wealth and short AI timelines are now setting budgets for bednets, vitamin-A distribution and vaccination incentives. This is the first major example of how the Funding Anthropalypse will reshape EA.
What should donors do - my take
The answer depends on size. For donors giving thousands, not much has changed: 6x GiveWell's benchmark is still some of the best marginal cost-effectiveness a normal donor can buy, and GiveWell has committed to tell donors if it thinks their money should go elsewhere. If you are giving thousands each year, I would stay as you are.
For donors giving millions, a lot has changed. The highest impact move is probably not adding to GiveWell's margin, but building other capacity that will help to absorb next year's likely windfall. cG concede that "the marginal cost-effectiveness of GiveWell-directed funding will be lower than it has been in the past couple of years", and that donors redirecting marginal donations would be, to some degree, "perfectly rational".
Finally, this does nothing for power concentration - if anything, it makes it worse. cG and GiveWell are the two largest players in EA, and they are not independent: the two largest funders are now exposed to the same volatile source at the same time. I expect the gift itself to be an outstanding grant at the object level, funding highly cost-effective programmes that do an enormous amount of good for very little money. But this excellent grant does nothing to address the fragile architecture of the space.
Disclosure: I have a conflict of interest in some of what I recommend in the full piece, as an operator in this space myself via the Mid-stage Global Health Fund. Discount accordingly.
A few things I'd be interested in hearing about:
