Currently it seems like each grantmaker is (on average) responsible for ~$10m/y. One question I think about sometimes: how will # of grantmakers scale as more $ go towards AI safety funding? If funding is eg 3x’ing year-over-year, it’s unclear whether we’re currently training up that number of grantmakers.
My vibes-based sense is that at least currently, the amount of philanthropic capital that could go towards AI safety projects is growing quite a bit faster than the number of grantmakers. I’m pretty worried about this.
Taking a look at CG:
Per this, the number of GCR program staff at CG only grew 2x from 2019 to 2022.
Quickly looking at archives of the team list from EOTY 2022 and EOTY 2025, it looks like the growth rate of program staff over that period was roughly 2 to 2.5x.
Another question might be: what is a good ratio of # of grantmakers to # of direct work? I’d ballpark there to be ~1000 fulltime AIS direct workers; does a 20:1 ratio seem high, low, or just right?
I think the ratio framing is a bit tricky and depends on a lot of other variables (for instance, how mature the field is, how many promising ideas are floating around the memesphere, how good AIs are at doing direct work, how much philanthropic capital there is, etc). The other thing is that the number of direct workers is itself downstream of grantmaker capacity.
My vibes-based sense is that at least currently, the amount of philanthropic capital that could go towards AI safety projects is growing quite a bit faster than the number of grantmakers. I’m pretty worried about this.
Taking a look at CG:
Per this, the number of GCR program staff at CG only grew 2x from 2019 to 2022.
Quickly looking at archives of the team list from EOTY 2022 and EOTY 2025, it looks like the growth rate of program staff over that period was roughly 2 to 2.5x.
I think the ratio framing is a bit tricky and depends on a lot of other variables (for instance, how mature the field is, how many promising ideas are floating around the memesphere, how good AIs are at doing direct work, how much philanthropic capital there is, etc). The other thing is that the number of direct workers is itself downstream of grantmaker capacity.