AI might change the world as we know it.

let’s ensure those changes are positive.

Apply to our Intro to ML Safety Fellowship, by March 8th Friday 11:59pm PT here:

We think positively shaping transformative AI is worth working on because:

  1. Experts predict that we will likely see human-level AI within our lifetimes
  2. Getting future AI systems to share human values looks like it could be an extremely difficult technical problem.
  3. Moreover, even with a technical solution, coordinating corporate and government actors to safely deploy these systems is itself a very thorny policy issue.

AI Alignment is a problem of immense stakes:

  1. Succeeding in aligning powerful AI systems could usher in a new world of human flourishing, a world free from poverty, oppression, and suffering.
  2. Failure to solve this problem appears to pose a major risk of human extinction or the disempowerment of humanity.
  3. Despite these stakes, this problem is extremely neglected, with only a few hundred people working on it full time. That’s why we need you!

We think that reducing risks from advanced artificial intelligence may be one of the most important problems of our time. We also think it’s a highly interesting and exciting problem, with open opportunities for many more researchers to make progress on it.

CAIA exposes undergraduate and graduate students exploring research relevant to reducing risks from advanced AI by running a quarter-long introductory fellowship group on AI safety for technical machine learning track.

We are welcoming eager folks to join our leadership team! If you would like to join, fill up the interest form, and write regarding your interest to

Join our Discord after filling our interest form!

Why care about AI Safety?

In recent years, we’ve seen AI exceed our expectations in a wide variety of domains — including playing Go, composing human-like text, writing code, and modeling protein folding. It may not be long until we create AI systems that are much more capable than humans at solving most cognitive problems.

What does work in AI Safety look like?

“AI safety” refers to efforts to prevent artificial intelligence from causing harm. Possible harms from powerful future AI systems pursuing goals in conflict with human flourishing, because such systems could pose an existential risk to humanity.