You found the arcade.
Three small games, one for each pillar of our investment thesis. Each takes about five minutes. Each makes the same argument we make to founders and LPs, just with more forklifts.
Unit n.
Program a humanoid one motion at a time, from the factory cell to the care home. Every instruction adds to unit cost, and the parts get cheaper every year. Hit par and you win.
We would rather own the company that makes a capability affordable than the one that makes it first.
Handshake.
Bob has a new neural implant and a short list of things he wants to do again. You're the decoder between his brain and the machine. First a cursor, then a text, a coffee cup, the mailbox, and finally the payer.
We underwrite signal fidelity that exists today, and pathways buyers or payers will fund.
Ten Thousand Worlds.
Run thousands of simulated worlds, turn every failure into training data, and pick the infrastructure that lets you scale from 9 worlds to 10,000. Then your robot meets the real world. Watch for forklifts.
Physical-world models cannot learn from real-world data alone.
Play them in order and you'll notice something.
The humanoid you program in Unit n and Bob from Handshake come back in Ten Thousand Worlds, as the robots you're training. That's not just a callback. A company in one of our domains is almost always a customer of the other two, which is why we invest across all three.