If you want a multibillion $ valuation then it's not about the robots, it's about modeling the physical world. Of course, having robots is going to help with that!
Robots are the exhaust, not the engine. The real compounding asset is the world model: perception + prediction + causal “what-if” planning that generalises across tasks and environments. Hardware still matters, but mainly as a data flywheel and a reality-check against simulation shortcuts—valuation follows whoever owns the fastest loop from real-world interaction → model improvement → deployment.
Robots are the exhaust, not the engine. The real compounding asset is the world model: perception + prediction + causal “what-if” planning that generalises across tasks and environments. Hardware still matters, but mainly as a data flywheel and a reality-check against simulation shortcuts—valuation follows whoever owns the fastest loop from real-world interaction → model improvement → deployment.