Before you write to us.

This page is not a job listing. It is a description of how we work and what we care about, so that you can decide whether you would be at home here before you decide whether to apply.

We are building towards Engineering Superintelligence — AI that can design, simulate, and build complex machines and physical systems, trained in the world's most demanding engineering environments. It is a long, unglamorous, deeply technical problem. It requires people who know how the physical world actually works and people who know how learning systems actually work, in the same room, arguing well.

• From first principles. We start with the problem and the physics, not with the tool or the template.
• Truth over comfort. We say what the evidence supports, including what remains uncertain. We change our minds when it is warranted, and we expect the same of each other.
• Ownership. Small, senior teams. Each person carries a whole problem, not a slice of one.
• Craft. What we build must be reflected in how we build. The quality of our work shows in the code, the model, the experiment, and the argument. Each deserves care.
• Stewardship. The systems we train will help design machines that carry people and power cities. Intelligence at that level must be stewarded, not unleashed — isolated, traceable, and in human hands.
• Care. The work is hard. We look after one another while doing it.

We believe the intelligence we build will reflect the people who build it — their rigor, and their character. So alongside exceptional capability, we look for heart, kindness, compassion, and love. Not as words on a wall, but as how you treat the people you work with when the work is hard, how you handle being wrong, and why you want to do this at all.

We are hiring AI researchers; aerospace, mechanical, and materials engineers; scientific machine learning experts; and computer scientists, in Menlo Park, San Francisco, Boston, London, Zurich, and Paris. We call the fit between a person and this work resonance, and it is the first thing we look for.

Send us your résumé and three examples of your best work — things you built, proved, shipped, or discovered — with a short note on why you want to do this. For collaborative work, tell us what you contributed; for work you cannot share, a description is enough.

We read everything ourselves and reply to everyone we would like to meet.