Note: I’m writing this in a personal capacity; all views are my own
A couple I’m close with intend to donate at least half of their assets over their lifetime. Currently, they are most excited about GiveWell, pretty excited about animal welfare opportunities, and not excited about AI safety giving. They have a preference for giving a pretty similar amount every year.
By many EA’s lights, this couple’s preferences are far out of line with their comparative advantage as donors in the EA movement. Their comparative advantages include that:
- They are two US citizens, each with the ability to give the maximum in hard dollar political contributions to a number of congressional candidates.
- They have access to money now. My understanding is that certain AI safety groups would trade multiple dollars two years from now for a dollar today. This suggests that this couple’s preference to give gradually may have less impact than frontloading donations (e.g, through a DAF where they can carry forward their tax deductions over five years)
- They do not work at AI companies. This means that organizations like METR can accept funding from this couple without the conflict of interest that comes from taking money from the companies they aspire to hold accountable for dangerous capabilities progress.
- They file jointly with a low marginal tax rate (they are essentially retired). This means it may make more sense for them to give to 501c4 or hard-dollar political opportunities, since their tax savings from tax-deductible gifts is far below that of other EA donors.
They cannot “make deals” with other donors to swap donations. This would be illegal in the case of political giving, and may be against the spirit of AI safety groups’ policies around impartiality in other cases. If these deals were written in an enforceable contract, they may also violate Donative Intent.
However, if EA donors embraced the spirit of giving based on comparative advantage, we would grow the total impact the EA community could achieve by the lights of all worldviews. If instead of $X to GiveWell, this couple saw their donations as Y% of the impact of the whole community of EA donors, their same “Y%” would represent a greater quantity of total impact in a world where every EA donor gave based on comparative advantage.
Giving is a deeply personal choice. It is wonderful that this couple is excited about effective giving and so generous with their resources. I write this post not to claim that certain causes are better than others, but to share a reflection that comparative advantages are real. Right now, donors in positions like this couple have high-leverage opportunities that may exceed the cost-effectiveness of what institutions or AI company staff can give next year by a factor of 10 or more. To learn more, you can check out this blurb on an AI safety donation opportunity recommended by Coefficient Giving staff and sign up to receive opportunities from Eric Neyman.
The obvious first thing to ask is have they considered donating to the operations costs of an effective giving fundraiser team such as One for the World ( https://1fortheworld.org/ ) as a means of raising more money for the GiveWell Top Charities than they themselves would be able to donate?
Getting people to cross-cause prioritise is difficult. There's much that can be obtained by considered within-cause prioritisation.