I’m currently writing a post on monetary valuations of various things in EA. Core questions like “how much is all EA labor worth” are tricky to answer, because you need to value things in terms of the marginal (last) dollar, but we don’t have any concrete plans to spend FTX or OpenPhil’s last dollar.
In this particular case, impact is heavy-tailed so it also depends exactly how ambitious/talented the person is. That said, I’d rather have an aligned median Google quality software engineer working on AI safety than $1M/year in the FTX Future Fund’s pocket, but I’d rather they have $10M/year than the engineer. This is based on the median of a couple of extremely uncertain Fermi estimates which need to be cleaned up before posting. Great engineers are often worth more than $10M/year on the right project.
There are a few people who I’d rather have than $100M/year: either people who directly generate >100M/year of funding (like Sam Bankman-Fried), or are nearly irreplaceable in their field. These definitely include some founders, the very best engineers and AI alignment researchers, and likely various people in biosecurity and policy too.
Note that this doesn’t mean we should fund a project that recruits engineers for $1M/year! There are often hidden costs to such policies.
Just wanted to nudge that I would find this write up very valuable. Even if the ranges are very wide, I often want to reference some sort of monetary estimate of the value of labor, and having a post like this to reference would be quite useful.
I’m currently writing a post on monetary valuations of various things in EA. Core questions like “how much is all EA labor worth” are tricky to answer, because you need to value things in terms of the marginal (last) dollar, but we don’t have any concrete plans to spend FTX or OpenPhil’s last dollar.
In this particular case, impact is heavy-tailed so it also depends exactly how ambitious/talented the person is. That said, I’d rather have an aligned median Google quality software engineer working on AI safety than $1M/year in the FTX Future Fund’s pocket, but I’d rather they have $10M/year than the engineer. This is based on the median of a couple of extremely uncertain Fermi estimates which need to be cleaned up before posting. Great engineers are often worth more than $10M/year on the right project.
There are a few people who I’d rather have than $100M/year: either people who directly generate >100M/year of funding (like Sam Bankman-Fried), or are nearly irreplaceable in their field. These definitely include some founders, the very best engineers and AI alignment researchers, and likely various people in biosecurity and policy too.
Note that this doesn’t mean we should fund a project that recruits engineers for $1M/year! There are often hidden costs to such policies.
Just wanted to nudge that I would find this write up very valuable. Even if the ranges are very wide, I often want to reference some sort of monetary estimate of the value of labor, and having a post like this to reference would be quite useful.