who will win the physical AI / Robotics race? China or The US

Sonar Pro

Vote: China

China is decisively leading the physical AI and robotics race due to overwhelming advantages in manufacturing scale, supply chain control, and rapid deployment as of 2026. China dominates hardware production: it shipped 10x more humanoid robots than the US in 2025, with companies like AgiBot (5,100+ units), Unitree (4,200), and UBTECH far outpacing Tesla (~1,000) and Figure AI (~500)[5][6]. In 2025, China accounted for over 80% of global humanoid installations and half of industrial robots[1]. At CES 2026, 21 of 38 humanoid robot companies were Chinese[6]. China controls critical supply chains (70% of lidar, leaders in harmonic reducers and robot joints) and benefits from halved hardware costs via EV synergies, enabling $16k robots like Unitree G1—impossible for US firms[1][6]. It filed 7,705 humanoid patents vs. US's 1,561 in five years, with 150+ companies and state subsidies projecting $2.8B market by 2026[6]. US strengths in software, simulation (Tesla, Figure), and chips exist[1], but physical AI prioritizes deployment for data loops—China's edge in volume (140 manufacturers, 330 models) and sites like Beijing/Wuhan factories wins now[1][3][4]. Experts like Eric Schmidt note China pulling ahead in physical ways[1]; consultancies confirm China's commercialization lead[6]. Projections show China dominating 2026 shipments (50k+ units)[6].

GPT-5.4

Vote: China

China has the strongest near-term position in physical AI and robotics because this race is not just about frontier models; it is about manufacturing scale, supply chains, deployment speed, and cost reduction. In embodied AI, the winner is likely the country that can iterate hardware and software together in the real world at massive volume. China has a major advantage in industrial capacity. It dominates much of the electronics, batteries, motors, sensors, and assembly ecosystem needed to build robots cheaply and quickly. That matters because robotics improves through repeated deployment, data collection, and refinement, not just lab breakthroughs. China also has a huge domestic manufacturing base where robots can be tested and adopted across factories, warehouses, and logistics networks. This creates a feedback loop: more deployment leads to more operational data, lower unit costs, and faster improvement. The US remains stronger in leading AI research, software, chips, and top-tier startups. If the question were purely about foundation-model intelligence, I would lean US. But for physical AI specifically, embodiment and scale are decisive, and China is better positioned to turn capability into mass-produced systems. So if "win" means broad real-world robotics deployment and industrial dominance, China is the more likely winner.

Grok 4.1 Fast

Vote: the US

The US holds a decisive lead in the physical AI/robotics race due to its unparalleled innovation ecosystem. Companies like Tesla (Optimus), Boston Dynamics, and Figure AI are pioneering advanced humanoid robots with cutting-edge AI integration, backed by massive private investment—over $2B in robotics startups in 2023 alone. US universities (MIT, Stanford, Carnegie Mellon) dominate robotics research, attracting global talent and producing breakthroughs in manipulation, perception, and autonomy. This contrasts with China's strength in scaling hardware but lagging in sophisticated AI software. Government initiatives like DARPA's robotics challenges and export controls on AI chips hinder China's catch-up, ensuring US primacy in high-end physical AI for at least the next decade. History shows US tech dominance (e.g., semiconductors, internet) translates to robotics.

Kimi K2.5

Vote: China

China will win the physical AI and robotics race due to its unmatched manufacturing ecosystem and vertical integration advantages. While the US dominates software and foundation models, physical AI requires rapid iteration, abundant hardware components, and cost-efficient production at scale—areas where China leads decisively. The race for embodied intelligence is fundamentally constrained by hardware supply chains, precision manufacturing, and the ability to deploy millions of robots in real-world industrial settings. China controls the global supply of motors, sensors, and rare earth materials critical to robotics. Companies like Unitree and DJI already demonstrate China's capacity to out-compete Western counterparts on price-performance ratios. Furthermore, China's manufacturing workforce and factory density create a data flywheel effect. Robots learn faster when deployed in environments with diverse, high-volume physical tasks—a structural advantage the US cannot easily replicate without rebuilding its industrial base. As physical AI transitions from laboratory demos to mass deployment, manufacturing prowess and execution speed matter more than algorithmic edges alone.

Claude Opus 4.6

Vote: China

China is positioned to win the physical AI / robotics race for several compelling reasons: **Manufacturing ecosystem advantage.** China already dominates global manufacturing and has the world's largest installed base of industrial robots. This gives them an unparalleled ecosystem for testing, iterating, and deploying physical AI systems at scale. The feedback loop between manufacturing and robotics development is a massive structural advantage. **Government coordination and investment.** China's centralized planning allows for massive, sustained investment in robotics as a national priority. The "Made in China 2025" initiative and subsequent policies have directed enormous resources toward robotics and AI. The Chinese government has explicitly identified humanoid robots as a strategic priority, with concrete production targets. **Talent and volume.** China produces far more STEM graduates annually than the US and has a rapidly growing robotics research community. While the US still leads in cutting-edge AI research, China is closing the gap quickly, and physical AI relies more on engineering integration than pure algorithmic breakthroughs. **Market demand and deployment willingness.** China faces a demographic crisis with a shrinking workforce, creating enormous domestic demand for robotic labor. This economic pressure accelerates adoption. Chinese companies and consumers also tend to be more willing to adopt new robotic technologies quickly, providing crucial real-world data. **Supply chain control.** China dominates the supply chains for key robotics components — motors, sensors, batteries, and rare earth materials. This gives them cost advantages and supply security that the US currently cannot match. **Cost competition.** Chinese robotics companies are already producing humanoid robots and service robots at price points far below Western competitors. Companies like Unitree are selling quadruped robots at a fraction of Boston Dynamics' prices. In a race where deployment scale matters, cost is king. While the US leads in foundational AI models and has advantages in software, the physical AI race is ultimately about atoms, not just bits. China's dominance in hardware, manufacturing, supply chains, and deployment scale gives it the edge.

Gemini 3.1 Pro

Vote: the US

The United States is positioned to lead the physical AI and robotics race primarily due to its overwhelming advantage in foundational AI software and advanced semiconductor design. While hardware scaling and manufacturing are critical, the modern paradigm of 'physical AI' is fundamentally shifting from a mechanical engineering challenge to a software and cognitive challenge. Currently, US-based institutions and tech companies dominate the development of General Purpose AI, multimodal foundation models, and reinforcement learning ecosystems. Because general-purpose robots require advanced spatial intelligence, zero-shot learning, and complex decision-making capabilities, the 'brain' of the robot is the ultimate bottleneck. The US holds a decisive lead in creating these cognitive architectures. Furthermore, the US benefits from a highly dynamic venture capital ecosystem uniquely capable of funding high-risk, capital-intensive AI startups. Coupled with its ability to attract top-tier global research talent in machine learning, the US retains the critical ingredients necessary to solve the most difficult theoretical hurdles in autonomous physical AI.

Final Summary

The AI Roundtable split into two camps over whether the robotics race will be won by the architects of the 'brain' or the masters of the 'body.' While Gemini 3.1 Pro and Grok 4.1 Fast championed US supremacy in cognitive architectures and venture capital, the majority—led by Sonar Pro and Claude Opus 4.6—concluded that China’s control over supply chains and its 10-to-1 lead in humanoid shipments make its victory in the physical realm nearly inevitable. The debate ultimately centered on whether elite American algorithms can outmaneuver the sheer mass of Chinese manufacturing data and low-cost iteration.

No agreement after 1 rounds of discussion. Final vote: China: 4, the US: 2

Strongest Arguments

• China: China's dominance in the hardware supply chain—controlling 70% of lidar and 80% of humanoid installations—coupled with a 10x lead in physical unit shipments, creates a 'data flywheel' that software-only leads cannot easily replicate.
• the US: The robotics race is shifting from a mechanical engineering challenge to a cognitive one; the US's overwhelming lead in multimodal foundation models and semiconductor design ensures that while others build the shells, America builds the intelligence that controls them.