Perhaps there are other useful factors in addition to student population that we can look at, such as “% of students majoring in non-profit management, environmental studies, etc.” as a rough proxy for the level of “proto-EA-ness” in a student population.
I agree that those specific examples are probably not great proxies, but things related to machine learning or other important technologies might be good—especially in conjunction with information about the competitiveness of the programs.
Anecdotally, the most engaged young EAs I’ve come across are usually heavily into liberal arts (a lot of philo and history majors) or CS majors. I’d find it difficult to make meaningful major-specific strategies off of that.
Hmm at the (strong) danger of selecting too much on traits that I think I or my friends are likely to have, rather than predictors of actual expected impact:
enjoyment of rationalist-y writing
vegetarianism or other predictors of moral choices, particularly at an unusually young age
though I think an increasingly high percentage of young liberals are vegetarians these days, so it’s probably a weaker signal
preference for utilitarianism-like thinking
preference for analytic philosophy in general
high general cognitive ability
International Math (or Physics, or Informatics etc) Olympiad gold/silver medalists, or other signifiers of outlier ability
which universities have (impact- or engagement- weighted) EA alumni
evidence of wanting to/being able to think from the perspective of others, e.g. debate/philosophy club
I thought Joseph was trying to identify proxies that you could easily measure at the university level, e.g., by collecting stats from university webpages. Some of the proxies you mention—e.g., “enjoyment of rationalist-y writing”—don’t seem to fit this goal, as they seem more targeted to the individual (or are just not already collected and reported at the university level). Granted, some such as IMO medalists and EA alumni representation do seem amenable to such analysis, although the latter seems like it would be a very laggy-metric.
Something like “existence of and participation in speech & debate clubs and related extracurriculars (e.g., ethics bowl, Model UN, Mock Trial)” seems like it would be worth looking into as a candidate.
I expect this to be a pretty poor proxy fwiw.
I agree that those specific examples are probably not great proxies, but things related to machine learning or other important technologies might be good—especially in conjunction with information about the competitiveness of the programs.
Yeah there’s way too many variables.
Anecdotally, the most engaged young EAs I’ve come across are usually heavily into liberal arts (a lot of philo and history majors) or CS majors. I’d find it difficult to make meaningful major-specific strategies off of that.
I agree that it is pretty sloppy/rough. Can you share any suggestions for better proxies?
Hmm at the (strong) danger of selecting too much on traits that I think I or my friends are likely to have, rather than predictors of actual expected impact:
enjoyment of rationalist-y writing
vegetarianism or other predictors of moral choices, particularly at an unusually young age
though I think an increasingly high percentage of young liberals are vegetarians these days, so it’s probably a weaker signal
preference for utilitarianism-like thinking
preference for analytic philosophy in general
high general cognitive ability
International Math (or Physics, or Informatics etc) Olympiad gold/silver medalists, or other signifiers of outlier ability
which universities have (impact- or engagement- weighted) EA alumni
evidence of wanting to/being able to think from the perspective of others, e.g. debate/philosophy club
I thought Joseph was trying to identify proxies that you could easily measure at the university level, e.g., by collecting stats from university webpages. Some of the proxies you mention—e.g., “enjoyment of rationalist-y writing”—don’t seem to fit this goal, as they seem more targeted to the individual (or are just not already collected and reported at the university level). Granted, some such as IMO medalists and EA alumni representation do seem amenable to such analysis, although the latter seems like it would be a very laggy-metric.
Something like “existence of and participation in speech & debate clubs and related extracurriculars (e.g., ethics bowl, Model UN, Mock Trial)” seems like it would be worth looking into as a candidate.