We are hiring people who want to work on the hardest unsolved problem in AI alignment. If you believe the alignment problem extends far beyond language, we would like to talk.
We are based in San Francisco. We work in-person.
Small team, high ownership. We are a small team where everyone works on the hardest problems directly. There are no layers between you and the work that matters. You will have more responsibility and more impact than at any larger organization.
Research with production urgency. We publish and we ship. Our work advances the field and serves real design partners at frontier labs. If you want to do research that matters immediately, not in two years, this is the place.
First-principles thinking. The verification problem for visual AI does not have established solutions. We start from the structure of the problem, not from what existing pipelines can be adapted. This requires intellectual honesty and comfort with ambiguity.