Tech • AI • Robotics • Game
Their lead is real: Chinese robots vs. Tesla!
7/10TL;DR
China used the World Humanoid Robot Games in Beijing to showcase rapid progress in humanoid robotics, blending technical demonstrations, industrial ambition and a clear soft power message.
KEY POINTS
A tightly staged national showcase
Beijing hosted the first widely publicized edition of the World Humanoid Robot Games, gathering more than 2,000 robots, 51 events and five days of competition. The program mixed sprints, athletics, football and industrial-style tasks, while a polished opening ceremony featured roughly 100 robots marching with an orchestra. The event’s design emphasized scale, discipline and technological confidence.
Headline performances and viral images
State media highlighted a robot named Lightning, presented as running 100 meters faster than Usain Bolt, with an official time cited around 9.3 seconds. Other machines dominated medal tables, including H1 by Unitree, which won four gold medals in athletics. At the same time, some of the most shared images showed robots crashing into barriers or requiring stretcher-style removal, underscoring both spectacle and technical fragility.
A competition dominated by Chinese manufacturers
No European robots were reported among the competitors, reinforcing the event’s role as a domestic industrial showcase. The number of humanoid robot models displayed appears to have risen sharply from the previous edition, from roughly 500 to more than 2,000, signaling a deliberate effort to demonstrate manufacturing depth. The message was that China intends to dominate humanoid robotics much as it has pushed aggressively in electric vehicles and other hardware sectors.
The market is running ahead of current revenues
Investor excitement around humanoid robotics is surging. Unitree recently entered the market in Shanghai, with reports of a first-session jump of 629% before closing around 460% higher. Yet estimates cited for the company’s humanoid business suggest only about 5,000 units sold per year and revenue near $200 million, far below valuations driven by long-term expectations. Morgan Stanley has projected the humanoid robot market could reach $7.5 trillion by 2050.
Today’s main use case is still show business
Despite the futuristic imagery, the most proven commercial application remains entertainment, events and short-term corporate demonstrations. Humanoid robots reliably attract crowds, photographs and media attention, but they still struggle with many routine tasks such as folding clothes or handling cluttered home environments. In that sense, the show itself is already a business model, even before broad industrial deployment arrives.
Useful work remains harder than running fast
Sprinting in a straight line is impressive, but dexterous manipulation is a much higher bar. The key bottleneck is often the hand problem: plugging in a USB-C cable, closing a zipper, pouring tea or folding fabric requires far more precise sensing and control. Companies are investing heavily in robotic hands, and some developers argue a humanoid is best understood as a computer whose most important peripherals are its hands.
Industrial deployment may come before domestic helpers
The near-term path is expected to run through warehouses, factories and structured workplaces rather than private homes. Some estimates place the first industrially viable deployments within two to three years, helped by advances in AI, simulation and teleoperation-based training. Many systems are still being taught through repeated human demonstrations captured by cameras and fed into training pipelines.
China leads in hardware while the US pushes software and pilots
The strategic split is becoming clearer. China is seen as strongest in hardware, including motors, manufacturing scale and supply chains. In the United States, companies such as Figure, 1X and Agility are focusing on workplace pilots, while Tesla has again delayed visibility on Optimus V3 and has not provided meaningful production figures. Even so, American firms remain active in manipulation and AI training systems.
Power and autonomy are still major limits
Battery life remains a stubborn constraint. Current top humanoids are reported at roughly two hours of autonomy, still far from what would be needed for broad continuous use. Moving an 80 to 100 kilogram body consumes substantial energy, and robotic efficiency remains vastly inferior to the human body, which operates on remarkably low power by comparison.
CONCLUSION
The Beijing games showed that humanoid robotics is advancing quickly, but they also revealed how much of the sector still depends on choreography, investor belief and national strategy. The real race is no longer just about legs or speed, but about hands, batteries, AI control systems and industrial scale.
Explain this