Figuring out if you can make good progress in AI safety, and whether you’re a good fit, is a hard problem – don’t expect to solve it overnight. That said, there are a number of good ways to find out your fit for AI safety:

Read about the work of people working on AI safety

One important consideration is the difference between technical AI safety work, and work on AI policy. Work in each of these fields will likely fit quite different people.

For technical AI research, a set of interviews with AI safety researchers was recently released. For AI policy or strategy, 80,000 hours has some good information on what careers here can look like broadly.

AGI Ssafety Ffrom fundamentals is a course designed to introduce many concepts in this area to people with only a basic statistical and mathematical background. Courses on both technical research and governance are available; doing some or all of the readings from this course could give you a feel for whether you would enjoy doing AI safety research.

Talk to people working on AI safety

To talk to people working on AI safety in person, you could attend an Effective Altruism conference and book one-to-one meetings with a number of AI researchers to ask them what they like and dislike about their work.

There are also a number of Slack groups available where you could speak to researchers, or other people looking to test their fit for alignment.

You could also check if there’s a group near you focussed on AI safety or Effective Altruism.

Do adjacent work

If you now have a pretty good idea of what working in AI safety looks like, but are still not sure it’s a good fit, you have two great options. One is to do similar work; work on policy for something else, or do research in other challenging, pre-paradigmatic fields.

Do internships in AI safety, or get funded for a short grant

Finally, the best way to get information about your fit for AI safety work is to work on AI safety! It’s not a lifetime commitment; there are a number of internships available, and a number of grantmakers will support short forays into research for promising candidates.