Thanks for writing and sharing this. I feel confused about the scale on some of these. For example:
it now somewhat less appears that a competent group of EAs could expect to do well in arbitrary industries… “How exceptional are EAs, compared to other social groups?”: 5⁄10 → 3⁄10
Pre-FTX collapse, an EA was the richest self-made person in the world under 30 after having switched to an industry they had no experience in (crypto), and this is evaluated as 5⁄10 ability to do well in arbitrary industries? What is the scale here? What are social groups that are 10/10? Is the claim that Sam was 10⁄10 but the rest of us were 1⁄10 so it averages out to 5/10? But weren’t something like >0.1% of EAs billionaires at that point? Surely that’s at like the 99th percentile of billionaire density for social groups?
And we are currently 3/10? The fact that most EA’s have the right to work in an OECD country seems like it automatically should put us at at least the 80th percentile or something?
Maybe this is pedantic, but this lack of precision on the scale makes it hard for me to interpret the strength of these updates. E.g. if Owen is making the claim that EA’s have below-average ability amongst humans to accomplish arbitrary business goals (which is my naïve interpretation of the 3⁄10 rating) then this seems like a really strong and surprising claim.
(Note: I’m more sympathetic to the claim that EAs do things that are impactful in absolute value but have a negative sign, so the overall net impact is lower. But I understand Owen to be talking about a different thing here.)
I read the endpoints as 0 is no extraordinary abilities at all (i.e., similiar to most groups with similiar characteristics like education and geography) and 10 as a group of Nobel prize winners or something.
Thanks for writing and sharing this. I feel confused about the scale on some of these. For example:
Pre-FTX collapse, an EA was the richest self-made person in the world under 30 after having switched to an industry they had no experience in (crypto), and this is evaluated as 5⁄10 ability to do well in arbitrary industries? What is the scale here? What are social groups that are 10/10? Is the claim that Sam was 10⁄10 but the rest of us were 1⁄10 so it averages out to 5/10? But weren’t something like >0.1% of EAs billionaires at that point? Surely that’s at like the 99th percentile of billionaire density for social groups?
And we are currently 3/10? The fact that most EA’s have the right to work in an OECD country seems like it automatically should put us at at least the 80th percentile or something?
Maybe this is pedantic, but this lack of precision on the scale makes it hard for me to interpret the strength of these updates. E.g. if Owen is making the claim that EA’s have below-average ability amongst humans to accomplish arbitrary business goals (which is my naïve interpretation of the 3⁄10 rating) then this seems like a really strong and surprising claim.
(Note: I’m more sympathetic to the claim that EAs do things that are impactful in absolute value but have a negative sign, so the overall net impact is lower. But I understand Owen to be talking about a different thing here.)
I read the endpoints as 0 is no extraordinary abilities at all (i.e., similiar to most groups with similiar characteristics like education and geography) and 10 as a group of Nobel prize winners or something.