I think this is a good question and there are a few answers to it.
One is that many of these jobs only look like they check the âimproving the worldâ box if you have fairly unusual views. There arenât many people in the world for whom e.g. âdoing research to prevent future AI systems from killing us allâ tracks as an altruistic activity. Itâs interesting to look at this (somewhat old) estimate of how many EAs even exist.
Another is that many of the roles discussed here arenât research-y roles (e.g. the biosecurity projects require entrepreneurship, not research).
Another is that the type of research involved (when the roles are in fact research roles) is often difficult, messy, and unrewarding. AI alignment, for instance, is a pre-paradigmatic field. The problem statement has no formal definition. The objects of study (broadly superhuman AI systems) donât yet exist and therefore canât be experimented upon. Out of all possible research that could be done in academia, âexpected tractabilityâ is a large factor in determining what questions people try to tackle. But when youâre filtering strongly for impact as EA is, you can no longer select strongly for tractability. So itâs much more likely that things will be a confusing muddle that itâs difficult to make clear progress on.
Honestly for me itâs probably at the âalmost too good to be trueâ level of surprisingness (but to be clear it actually is true!). I think itâs a brilliant community /â ecosystem (though of course thereâs always room for improvement).
I agree that you probably generally need unusual views to find the goals of these jobs/âprojects compelling (and maybe also to be a good job applicant in many cases?). That seems like a high bar to me, and I think itâs a big factor here.
I also agree that not all roles are research roles, although I donât know how much this weakens the surprisingness because some people probably donât find research roles appealing but do find e.g. project management appealing. (Also I do feel like most research is pretty tough one way or another, whether or not itâs âEAâ research.)
I guess thereâs also the âdownsidesâ I mentioned in the post. One that particularly comes to mind is that there still arenât a ton of great EA jobs to just slot into, and the ones that exist often seem to be very over-subscribed. Partly depends on your existing profile of skills of course :).
I think this is a good question and there are a few answers to it.
One is that many of these jobs only look like they check the âimproving the worldâ box if you have fairly unusual views. There arenât many people in the world for whom e.g. âdoing research to prevent future AI systems from killing us allâ tracks as an altruistic activity. Itâs interesting to look at this (somewhat old) estimate of how many EAs even exist.
Another is that many of the roles discussed here arenât research-y roles (e.g. the biosecurity projects require entrepreneurship, not research).
Another is that the type of research involved (when the roles are in fact research roles) is often difficult, messy, and unrewarding. AI alignment, for instance, is a pre-paradigmatic field. The problem statement has no formal definition. The objects of study (broadly superhuman AI systems) donât yet exist and therefore canât be experimented upon. Out of all possible research that could be done in academia, âexpected tractabilityâ is a large factor in determining what questions people try to tackle. But when youâre filtering strongly for impact as EA is, you can no longer select strongly for tractability. So itâs much more likely that things will be a confusing muddle that itâs difficult to make clear progress on.
Some quick thoughts on this from me:
Honestly for me itâs probably at the âalmost too good to be trueâ level of surprisingness (but to be clear it actually is true!). I think itâs a brilliant community /â ecosystem (though of course thereâs always room for improvement).
I agree that you probably generally need unusual views to find the goals of these jobs/âprojects compelling (and maybe also to be a good job applicant in many cases?). That seems like a high bar to me, and I think itâs a big factor here.
I also agree that not all roles are research roles, although I donât know how much this weakens the surprisingness because some people probably donât find research roles appealing but do find e.g. project management appealing. (Also I do feel like most research is pretty tough one way or another, whether or not itâs âEAâ research.)
I guess thereâs also the âdownsidesâ I mentioned in the post. One that particularly comes to mind is that there still arenât a ton of great EA jobs to just slot into, and the ones that exist often seem to be very over-subscribed. Partly depends on your existing profile of skills of course :).