tl;dr: I like the post. One thing that I think it gets wrong is that if “picking up trash in the park is a fine EA project to cut our teeth on”, then that is a sorry state for EA to find itself in.
I think that a thing that this post gets wrong is that EA seems to be particularly prone to generating bycatch, and although there are solutions at the individual level, I’d also appreciate having solutions at higher levels of organization. For example, the solution to “you find yourself writing not-so-valuable blogposts” is probably to ask a mentor to recommend you valuable blog posts to write.
One proposal to do that was to build what this post calls a “hierarchical networked structure” in which people have people to ask about which blog posts or research directions would be valuable, and Aaron Gertler is there to offer editing, and further along the way, EA groups have mentors which have an idea of which EA jobs are particularly valuable to apply to, and which are particularly likely to generate disillusionment, and EA group mentors themselves have someone to ask advice to, and so on. This to some extent already exists; I imagine that this post is valuable enough to get sent on the EA newsletter, which means that involved members in their respective countries will read it and maybe propagate its ideas. But there is way to go.
Another solution in that space would be to have a forecasting-based decentralized systems, where essentially the same thing happens (e.g., good blog posts to write or small projects to do get recommended, career hopes get calibrated, etc.), but which I imagine could be particularly scalable.
We can also look at past movements in history. In particular, General Semantics also had this same problem, and a while ago I speculated that this lead to its doom. Note also that religions don’t have the problem of bycatch at all.
That’s good feedback and a complementary point of view! I wanted to check on this part:
“I think that a thing that this post gets wrong is that EA seems to be particularly prone to generating bycatch, and although there are solutions at the individual level, I’d also appreciate having solutions at higher levels of organization.”
Are you saying that you think EA is not particularly prone to generating bycatch? Or that it is, but it’s a problem that needs higher-level solutions?
Yeah, that’s not my proudest sentence. I meant the former, that it is particularly prone to generating bycatch, and hence it would benefit from higher level solutions. In your post, you try to solve this at the level of the little fish, but addressing that at the fisherman level strikes me as a better (though complimentary) idea.
tl;dr: I like the post. One thing that I think it gets wrong is that if “picking up trash in the park is a fine EA project to cut our teeth on”, then that is a sorry state for EA to find itself in.
I think that a thing that this post gets wrong is that EA seems to be particularly prone to generating bycatch, and although there are solutions at the individual level, I’d also appreciate having solutions at higher levels of organization. For example, the solution to “you find yourself writing not-so-valuable blogposts” is probably to ask a mentor to recommend you valuable blog posts to write.
One proposal to do that was to build what this post calls a “hierarchical networked structure” in which people have people to ask about which blog posts or research directions would be valuable, and Aaron Gertler is there to offer editing, and further along the way, EA groups have mentors which have an idea of which EA jobs are particularly valuable to apply to, and which are particularly likely to generate disillusionment, and EA group mentors themselves have someone to ask advice to, and so on. This to some extent already exists; I imagine that this post is valuable enough to get sent on the EA newsletter, which means that involved members in their respective countries will read it and maybe propagate its ideas. But there is way to go.
Another solution in that space would be to have a forecasting-based decentralized systems, where essentially the same thing happens (e.g., good blog posts to write or small projects to do get recommended, career hopes get calibrated, etc.), but which I imagine could be particularly scalable.
We can also look at past movements in history. In particular, General Semantics also had this same problem, and a while ago I speculated that this lead to its doom. Note also that religions don’t have the problem of bycatch at all.
That’s good feedback and a complementary point of view! I wanted to check on this part:
“I think that a thing that this post gets wrong is that EA seems to be particularly prone to generating bycatch, and although there are solutions at the individual level, I’d also appreciate having solutions at higher levels of organization.”
Are you saying that you think EA is not particularly prone to generating bycatch? Or that it is, but it’s a problem that needs higher-level solutions?
Yeah, that’s not my proudest sentence. I meant the former, that it is particularly prone to generating bycatch, and hence it would benefit from higher level solutions. In your post, you try to solve this at the level of the little fish, but addressing that at the fisherman level strikes me as a better (though complimentary) idea.