Since I am both mid-career and EA, maybe I can say a little about this even if I can’t give a full answer.
I was concerned about existential risk due to AI prior to the start of my career (heck, prior to going to college, and this was in 2000), but for a variety of reasons I failed to do much directly about this. I got distracted by life, had to get a job to deal with more pressing needs, and spent several years just trying to get along without putting much effort into AI safety.
Then a couple of years ago my life got better, I had more slack, and I used that slack to start working on AI safety as a “hobby”. So far this has proven pretty successful: I’ve published some things, had many interesting conversations with people who are also doing direct work on AI safety (part or full time), and helped influence research directions and progress.
I don’t know what this will turn into, but the hobby model is worth considering as a way to transition mid-career: get interested in and start working on something you care about, and eventually maybe transition to doing that work full time. Plus you’ll be somewhat unique in that you’ll be carrying forward all your existing career capital that others in your chosen space likely won’t have.
The downside of this approach is that it requires you have enough time and energy to do it. To make progress here it may be necessary to take a less demanding job to creating that time and energy or give up other commitments.
Definitely interested to see what others suggest or have tried.
This is similar to what I have been thinking, as I’ve had “Learn Mandarin” down as a goal for four years now. My idea being to do a “hobby” at first and then shift. But your point is you need the time and energy, and I’ve never been able to break out of my currently demanding job to start to pick it up.
I think this would suggest identifying concrete steps that can be taken as a “hobby” but produce something functional that can be signaled in the career marketplace. Certificates and the like.
Since I am both mid-career and EA, maybe I can say a little about this even if I can’t give a full answer.
I was concerned about existential risk due to AI prior to the start of my career (heck, prior to going to college, and this was in 2000), but for a variety of reasons I failed to do much directly about this. I got distracted by life, had to get a job to deal with more pressing needs, and spent several years just trying to get along without putting much effort into AI safety.
Then a couple of years ago my life got better, I had more slack, and I used that slack to start working on AI safety as a “hobby”. So far this has proven pretty successful: I’ve published some things, had many interesting conversations with people who are also doing direct work on AI safety (part or full time), and helped influence research directions and progress.
I don’t know what this will turn into, but the hobby model is worth considering as a way to transition mid-career: get interested in and start working on something you care about, and eventually maybe transition to doing that work full time. Plus you’ll be somewhat unique in that you’ll be carrying forward all your existing career capital that others in your chosen space likely won’t have.
The downside of this approach is that it requires you have enough time and energy to do it. To make progress here it may be necessary to take a less demanding job to creating that time and energy or give up other commitments.
Definitely interested to see what others suggest or have tried.
This is similar to what I have been thinking, as I’ve had “Learn Mandarin” down as a goal for four years now. My idea being to do a “hobby” at first and then shift. But your point is you need the time and energy, and I’ve never been able to break out of my currently demanding job to start to pick it up.
I think this would suggest identifying concrete steps that can be taken as a “hobby” but produce something functional that can be signaled in the career marketplace. Certificates and the like.