Thanks for writing this up! I broadly agree with your points and I think it’s an important topic. One factor that I see as pushing in the other direction is that many EA-affiliated orgs are small, new, and/or have uncertain futures. Investing in the design of a higher-quality recruitment process makes the most sense when you’re confident that you’ll get to reap the benefits of that process across many hires and many years.
If you have a six-month runway during which you plan to make one or two hires, it’s less cost-effective to spend dozens of hours improving your general hiring process (as opposed to simply trying to make the right individual hires). It’s also less feasible to gather statistically meaningful data by running experiments on your (tiny) number of top candidates; even if you’re willing to invest massive resources, there just isn’t enough data yet to support most types of experiments.
I enjoyed the challenge of recruiting for the EA-friendly organization I led (the Center for AI Policy), and I would be happy to consider full-time recruiting roles, but I wanted to point out these challenges. It’s not just that people aren’t aware of or interested in better recruiting; it’s also that sometimes their organizations aren’t large or permanent enough to justify large investments in recruiting skill.
I definitely agree with this challenge — I also wonder if this is part of the reason many of the people who I have found to be most thoughtful about recruiting in the field founded or ran small or new organizations — they had to recruit under different constraints (e.g. offering less job stability, less name recognition, etc), and had to be more creative to get talented people in.
Though I understand the challenge and have this experience (see my other comment) I do see three related cases where it might be even more critical for a start-up or uncertain initiative to hire (and fire) well.
First, hiring in some novel organizations in nascent cause areas might have an effect on the development of a whole field. If there are only a few players, making a wrong hire could change the reputation of the whole space and slow down or reverse progress for the cause. E.g. If you’re in wild animal welfare and provide credibility through employment to someone using unscientific methods or who is vocal publicly about highly controversial solutions to wild animal suffering, that could be disastrous for the development of wild animal welfare science. I could see similar risks in EA community building.
Conversely, if you hire the best, it could become easier to attract more high talent, accelerating the path to impact.
Second, in interventions where the reward of success is enormous but the risk of no or negative impact large, it seems critical to hire the best you can get.
And third, if you’re starting an org that’s trying something new, and you hire someone with average expertise, skills, or drive, or there’s a mismatch between competencies needed and offered, your endeavor might fail. This might lead you, and outsiders, to incorrectly believe the whole intervention isn’t tractable.
But, if you have short timelines and just need warm bodies who can be easily replaced, I’d likely also invest less in recruitment.
But yeah, I mostly relate to the frustrations of the impossibility of doing some of these ideal recruitment practices.
Thanks for writing this up! I broadly agree with your points and I think it’s an important topic. One factor that I see as pushing in the other direction is that many EA-affiliated orgs are small, new, and/or have uncertain futures. Investing in the design of a higher-quality recruitment process makes the most sense when you’re confident that you’ll get to reap the benefits of that process across many hires and many years.
If you have a six-month runway during which you plan to make one or two hires, it’s less cost-effective to spend dozens of hours improving your general hiring process (as opposed to simply trying to make the right individual hires). It’s also less feasible to gather statistically meaningful data by running experiments on your (tiny) number of top candidates; even if you’re willing to invest massive resources, there just isn’t enough data yet to support most types of experiments.
I enjoyed the challenge of recruiting for the EA-friendly organization I led (the Center for AI Policy), and I would be happy to consider full-time recruiting roles, but I wanted to point out these challenges. It’s not just that people aren’t aware of or interested in better recruiting; it’s also that sometimes their organizations aren’t large or permanent enough to justify large investments in recruiting skill.
I definitely agree with this challenge — I also wonder if this is part of the reason many of the people who I have found to be most thoughtful about recruiting in the field founded or ran small or new organizations — they had to recruit under different constraints (e.g. offering less job stability, less name recognition, etc), and had to be more creative to get talented people in.
Though I understand the challenge and have this experience (see my other comment) I do see three related cases where it might be even more critical for a start-up or uncertain initiative to hire (and fire) well.
First, hiring in some novel organizations in nascent cause areas might have an effect on the development of a whole field. If there are only a few players, making a wrong hire could change the reputation of the whole space and slow down or reverse progress for the cause. E.g. If you’re in wild animal welfare and provide credibility through employment to someone using unscientific methods or who is vocal publicly about highly controversial solutions to wild animal suffering, that could be disastrous for the development of wild animal welfare science. I could see similar risks in EA community building.
Conversely, if you hire the best, it could become easier to attract more high talent, accelerating the path to impact.
Second, in interventions where the reward of success is enormous but the risk of no or negative impact large, it seems critical to hire the best you can get.
And third, if you’re starting an org that’s trying something new, and you hire someone with average expertise, skills, or drive, or there’s a mismatch between competencies needed and offered, your endeavor might fail. This might lead you, and outsiders, to incorrectly believe the whole intervention isn’t tractable.
But, if you have short timelines and just need warm bodies who can be easily replaced, I’d likely also invest less in recruitment.
But yeah, I mostly relate to the frustrations of the impossibility of doing some of these ideal recruitment practices.