Looking beyond the robots’ footwork, China’s demonstration reveals a widening gap between spectacle and strategy — a gap that America, for all its robotics talent, is in real danger of falling into.

We are building impressive robots. We are not building the right ones.

Walk into any major robotics demo in the U.S. and you’ll observe fluid movements, precise manipulation, and maybe even a backflip. The most advanced Boston Dynamics robots can pick up and carry large objects that would risk injuring any human worker. On performance alone, we look competitive.

The problem is that performance in controlled settings is all we’re measuring. A recent Stanford report found that robots scoring nearly 90% success rates in controlled simulations succeed at just 12% of real household tasks. This gap between demo and deployment is not a rounding error — it is the whole problem. The U.S. is optimizing its humanoid robots for a sprint and calling it a marathon strategy.

Take the case of Figure AI’s 02 model. It logged 1,250 hours at BMW’s Spartanburg plant and moved 90,000+ components. By current metrics, it was a success. Look closer and the robot did one task: picking up sheet metal parts and placing them on a welding fixture for ten months straight. A mid-sized manufacturer — one already running automated systems — cannot justify thousands of dollars in investment for a machine that does one thing.

Successful one-off robot deployments obscure the real question: is this investment worth it at scale?

What NASA Taught Me About Brittle Machines

At NASA, decades of designing humanoid robots for environments that don’t forgive narrow thinking revealed that the machines that failed were the ones built for a single scenario. The ones that succeeded could multitask and be reprogrammed– for deployment in different settings. The arm built for the Space Shuttle, for example, was designed to position an astronaut who would catch and later release a satellite. It turned out the robot was better at making the catch itself — but positioning astronauts proved useful for other tasks, like repairing the Hubble Space Telescope.

The United States continues to build humanoids that are extraordinary in the conditions they were trained for and brittle everywhere else. Right now, in most factories, several humans generate better ROI than one humanoid robot.

We humans might not be the strongest or fastest, but we make up for it with adaptability. A single warehouse associate can pick up orders, restock shelves, flag a safety issue, and reroute around a spill — all before lunch. This fluid task-switching is the core of human labor value. If humanoid robots are supposed to take over U.S. factory floors, they need to be designed to be even more flexible than us.

The Policy Environment Isn’t Ready Either

To get there, we must first prime the policy environment. American manufacturers — particularly those in the mid-to-enterprise range that form the backbone of U.S. industrial output — have almost no structured pathway to adopt humanoid robotics at scale. Large manufacturers like BMW can absorb a ten-month single-task pilot as an R&D line item. A mid-sized auto supplier or contract manufacturer cannot.

The absence of a proper federal incentive structure will prolong the stagnation. Current federal R&D tax credits reward robotics discovery, not deployment. A manufacturer that spends $800,000 integrating a humanoid system gets essentially the same tax credit as one that buys a new forklift.

Even with $2.5 billion in venture capital already poured into robotics, private investment alone cannot spur effective deployment. Instead, the nation needs a distinct, robotics-focused “manufacturing deployment” tax incentive, stackable with the existing R&D credit — one that rewards those who make robots functional in real factories by offsetting integration costs, workforce transition expenses, and process redesign work. The U.S. can also expand the Manufacturing Extension Partnership, which already offers specialized consulting to small and mid-sized manufacturers, to provide humanoid deployment concierges at relatively low federal cost. Finally, NIST, working with NASA and others, should establish humanoid interoperability standards so manufacturers can combine multiple robotic systems safely.

What the Right Deployment Actually Looks Like

U.S. factories will also have to transform. Most industrial workflows were designed around human flexibility, improvisation, and self-directedness. Robots — even adaptable ones — will require something different: fleet-based tasking similar to how rideshares route cars, clear safety parameters for mixed human-robot environments, and new protocols for humanoids to interact with single-purpose machines.

What both the U.S. and China still misunderstand is the assumption that humanoids will replace jobs wholesale. Rather, they’ll generate value by filling the gaps that current automation can’t reach — the “in-between” work that’s too variable for a fixed conveyor system and too repetitive to justify a skilled employee. A mid-sized company will appreciate being able to shift humanoids into operating single-purpose machines like loading a washer, without having to replace the washer itself.

Moving materials between workstations, restocking inventory in busy warehouses, tending machines built for human interaction, and conducting inspections in dangerous and confined spaces aren’t glamorous use cases. But they represent operational problems that adaptable humanoids can begin to address within this decade.

America has the talent, the capital, and the industrial base to lead this transition. We are currently optimizing for the wrong outcomes and ignoring the policies that could enable real deployment.

The country that defines “good enough to deploy at scale” will set the terms for global manufacturing for decades. Right now, that country is not the United States.

It doesn’t have to stay that way.

The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.