Courses at ETH Zurich that are useful for working on AI safety, ranked by the community. Only upvote courses you have taken. To upvote, give the course's GitHub issue a 👍 (you need a GitHub account). Took one? Comment on its issue with which parts are relevant to AI safety and why. We may work your comments into the description.
+ Suggest a courseETH has no course on AI safety itself. This independent programme run by Zurich AI Safety fills the gap with small-group discussions of curated readings, covering alignment, RLHF, scalable oversight, robustness, interpretability and technical governance.
Do at least one research project during your studies. Find people who work on AI safety directly and ask to work with them, or think about how you could give the work of a lab you already know an AI safety twist.