The Trump administration has begun loosening semiconductor export controls on China, allowing Nvidia to sell advanced AI chips like the H20 to Chinese buyers in exchange for a cut of the revenue. It’s a sharp reversal from the Biden-era strategy of choking off China’s access to cutting-edge technology entirely, and it raises a question that has no clean answer: is it better to sell your adversary the weapons of the future, or to force them to build their own?
The semiconductor supply chain is the most strategically consequential chokepoint in the global economy. Virtually all of the world’s most advanced chips are fabricated by a single company, TSMC, on an island that China claims as its own. The United States doesn’t make the chips, but it controls the tools and software needed to design and manufacture them. That leverage is real, but it’s not unlimited, and how aggressively to use it is the central debate.
The hawkish case for tight export controls is straightforward: deny China the technology it needs to build next-generation AI and weapons systems. But China isn’t standing still. Chinese firms are innovating around the restrictions. And the tighter the controls, the more reason Beijing has to pour resources into building a fully independent supply chain, which is exactly what Washington says it wants to prevent.
We discuss the Cold War precedent of CoCom, why the lessons of export controls are less encouraging than you’d think, how Taiwan became the most important island in the world, and whether the real chip war is one the United States is fighting with itself.
