Kai-Fu Lee recently delivered a sharp critique of America's AI hardware capabilities. He contends that the U.S. is already lagging in this race, primarily due to shortcomings in manufacturing and supply chains.


Lee points out that U.S. attempts to restrict China through export controls have had limited effect. China is accelerating self-sufficiency in critical areas like GPUs, while American policies are inadvertently pushing China to break through technological barriers faster.

He emphasizes that hardware is not just about design but also manufacturing. With companies like TSMC concentrating production capacity in Asia, rebuilding U.S. supply chains would require significant time and massive investment. Currently, China holds advantages in infrastructure and production capacity planning.

Lee's assessment is grounded in real-world industry observations. He argues that America's overemphasis on short-term sanctions has neglected long-term competitiveness building. The AI hardware race ultimately comes down to supply chains and manufacturing efficiency—areas where China's investments are already paying off.

While this competition is far from over, the landscape is shifting. Unless the U.S. adjusts its strategy, it risks losing further ground. Lee's advice is straightforward: less political maneuvering, more pragmatic cooperation.