NLResearch Mentor, AI Safety
PRISMRemote
Why this is on the board
You spend around 14 hours, spread over 16 weeks. Those hours fall exactly at the moments when four early-career researchers in AI safety — research into how to keep powerful AI systems safe — set their course: which question they ask, which method they choose and what they do with their first results. The field seems to be short above all of experienced mentors, and someone who completes a good first project is more likely to stay in this work.
This board is in beta and this is our judgement. If you think we have it wrong, tell us. Give feedback · How we reason about it
About the role
Guide four AI safety researchers through a 16-week project from research question to conference submission with nine focused mentor sessions. - Lead a team of four fellows through research design, methodology, and experimental work in AI safety domains like interpretability or alignment. - Attend nine strategic mentor sessions (~14 hours total) at key decision points: idea pitch, methodology review, baseline results, and final submission. - Receive co-authorship on the team's conference paper…
Want to understand this first?
Start with our free course: about 100 minutes in six short units, at your own pace, with no sign-up. To go further, apply for the EA Fellowship: five weeks of reading and discussion in a small group. Also free, but selective.