Stupid thesis considering PRC already installs more industrial robots than RoW combined. There's some validity to PRC tendency to arbitrage cheap labour to defer automation when economical... i.e. why pay for expensive SaaS when you can just throw cheaper brains at problem. But the second something cheaper than bodies, PRC tend to adopt more ruthlessly than west, and with automation replacing actual projected blue collar shortage, absolutely smooth brained hot take to think west will lead in robots let alone making robots.

Unitree makes like 5 digit numbers chassis with high iteration, they use people because these are baby numbers where human iteration maximizes productivity, not lack of capability. They do it because they have this gear that west doesn't. Meanwhile PRC indigenizing all tooling to the highest end of precision. They're not going to use western robots for much longer. Indigenous replacement rate is basically J takeoff, even if we completely ignore that ABB/Kuka/fanuc all have massive assembling and R&D footprint in PRC.

Or take Tesla, one of the more competent US builders... ctrl+c, ctrl+v their model to PRC, a few rapid iteration cycles later, PRC plants > western plants (EU/US). Even SpaceX, forget payload/tempo/operational, PRC industry builds like 2-3x more first stages than US space program including SpaceX, i.e. when they figure out reusable they can put up 2-3x size reusable fleet... and that's basically them barely trying. The industrial lesson by now should be whatever West builds, PRC likely end up doing better, faster, cheaper, frequently pick all 3.

This argument generic copium by now - muh western imagination gap can somehow compete with PRC industrial gap... and robots that build robots is slightly more marketable and feasible than gray goo... until you realize one of PRC's core capability is just doing what west did better anyway.