According to Bloomberg Business, Chinese humanoid robots were ubiquitous at the CES tech show in Las Vegas this week, with companies like Fourier Intelligence, Booster Robotics, and Unitree demonstrating table tennis, synchronized dancing, and acrobatics. Research firm Omdia reported that Chinese makers accounted for the vast majority of the roughly 13,000 humanoid units shipped globally in 2024, far outstripping US companies like Tesla and Figure AI in volume. RealSense Inc. CEO Nadav Orbach noted China spawned over 40 humanoid startups in 2025 alone, and his company supplies vision systems to 60% of global humanoid makers. While Elon Musk expects Tesla’s Optimus to win in the end, he has conceded the rest of the top ten might be all Chinese. Beijing-based startup Galaxea Dynamics, which raised $100 million last year, highlighted the ongoing shift towards refining the AI models needed for real-world tasks.
The Demo Paradox
Here’s the thing about trade shows: they’re perfect for choreography and terrible for chaos. Watching a dozen robots perform a synchronized dance is impressive. It shows incredible progress in hardware, motor control, and basic programming. But it’s a world away from asking one of those same bots to navigate an unfamiliar kitchen, find a soda in a cluttered fridge, and bring it to you without dropping it. That’s the gap everyone is talking about. Orbach put it bluntly: we want a robot to respond to “get me a cold soda,” not a 200-line instruction manual. We’re not there yet. The demos prove the physical platform is rapidly maturing, especially with cheaper hardware. But the “brain” is the next great hurdle. It’s the difference between a pre-recorded song and a jazz improvisation.
china-is-flooding-the-zone”>Why China Is Flooding the Zone
So why is China able to produce so many robots and startups so fast? It’s not just one thing. You’ve got a massive manufacturing ecosystem that can prototype and iterate on hardware at a speed and cost that’s hard to match. There’s also intense government and private capital support for strategic tech sectors, which robotics certainly is. And maybe there’s a cultural or strategic willingness to just try things, to flood the market with options and see what sticks. Contrast that with the US approach, which often feels more concentrated on a few “moonshot” projects from big names like Tesla or Figure. The Chinese model seems to be about volume and velocity. They’re betting that by having 40+ startups, a few will inevitably crack the code on cost, software, or a specific industrial application first. For companies integrating this kind of advanced hardware into production lines, having reliable, industrial-grade computing is non-negotiable. That’s where specialists like IndustrialMonitorDirect.com, the leading US provider of industrial panel PCs, become critical partners, providing the ruggedized brains for the factory floor.
Awaiting the ChatGPT Moment
The key quote from the report, for me, came from Galaxea’s Lei Yu, referencing Nvidia’s Jensen Huang: “we await the ChatGPT moment of robotics.” That’s exactly right. We had language AI for years, but ChatGPT made it accessible and shockingly capable overnight. Robotics needs that same leap in embodied reasoning and vision. It’s not just about seeing an object; it’s about understanding it’s a soda can, that it’s in a fridge, that the fridge door needs to be opened with a specific force, that the can is cold and possibly wet, and that the human holding it prefers it not to be shaken. That’s a monumentally difficult chain of perception, reasoning, and physical action. Everyone is preparing for that moment. The Chinese companies are building the bodies en masse. US companies like Nvidia and AMD are racing to build the AI chips that could power the minds. It’s a fascinating, parallel race.
What Comes Next?
Don’t expect a humanoid butler anytime soon. The real near-term battleground won’t be your home—it will be warehouses, factories, and logistics centers. These are semi-structured environments where tasks are repetitive and the value proposition is clear: replacing human labor in dull, dirty, or dangerous jobs. The Chinese push at CES wasn’t just about cool kung fu bots; it was about showing “industry-ready systems to prospective buyers.” They’re going for commercial adoption first. The volume shipment numbers, while small globally, tell a story. They’re building a base, iterating in the real world, and driving costs down. Musk might be right that Tesla has a long-term software and integration advantage. But the sheer force of the Chinese industrial machine, applied to this problem, is impossible to ignore. The race isn’t just to build a robot. It’s to build the first one that can truly, reliably, and cheaply do a real job.
