Figure AI: general-purpose humanoids, built for real work
Humanoid robots were mostly lab demos and viral videos. Moonshot bet they were closer to useful work than the consensus thought.
The Bet
Figure AI was founded in 2022 to build general-purpose humanoid robots. At the time, humanoids were widely seen as a research curiosity: impressive in short clips, impractical in real workplaces, and many years away from paying for themselves.
Moonshot’s bet was that several curves were converging at once: better actuators and batteries, cheaper hardware, and rapidly improving AI that could help robots understand and act in messy environments.
The world is built for the human form. A robot shaped like a person can work in spaces that already exist.
Moonshot’s View
We believed the world is built for the human form, from stairs and shelves to tools and doorways. A robot shaped like a person can work in spaces that already exist, instead of requiring every factory and warehouse to be redesigned.
We also believed labour shortages in physical work would make the problem urgent, not optional. When demand is that large, the first credible general-purpose robots do not need to be perfect to be valuable.
Room for Disagreement
The case against is strong: humanoids are mechanically complex, expensive to build and hard to make reliable. Specialised robots designed for a single task are often cheaper and better at that task.
We agree that single-purpose robots will remain important. Our disagreement is about generality: we think a robot that can learn many tasks will eventually beat a fleet of machines that can each do only one.
Notable
Figure AI, in brief.
- Sector
- Robotics
- Founded
- 2022
- Founder
- Brett Adcock
- Moonshot backed
- Early
- The contrarian call
- General-purpose humanoids were closer to useful than they looked
- Money raised
- $1.75B to date
- Company value
- $39B
- Moonshot 25
- No. 25 of 25 overall
Source: money raised and company value from Caplight, as of Nov 17, 2026. View source
Pitch Moonshot
Building something the crowd doesn’t believe in yet?
Tell us in about two minutes. No warm intro needed, and every pitch is read by the team.