Physical AI moves into tractors, not just robots

Bucher's argument rests on the distinction between humanoid robots designed for human environments and the need for autonomy in off-road industrial work. He emphasizes that the physics of off-road work doesn't change with better AI—machines still need to mow, spray, dig, and haul. The value lies in embedding intelligence into those machines, allowing them to operate with less dependence on skilled operators.

The labor crisis is a key driver. With U.S. farmers averaging nearly 60 years old and construction facing a need for hundreds of thousands of new skilled workers in 2026, the labor gap is structural, not cyclical. Bucher suggests that OEMs—the established equipment manufacturers—are the ones with the reach to scale autonomy, but they likely can't build AI capabilities overnight. Hence, the proposed partnership model: AI companies provide the "brain," while OEMs provide the "brawn."

Bucher also addresses potential concerns about reliability. He notes that critical safety decisions must happen on board the machine, with low-latency processing, rather than relying on cloud connectivity. This points to a future where machines are increasingly self-sufficient in perception and decision-making, even as they work alongside humans.