Hans Tung Links AI Dominance to Compute Infrastructure Scale
Notable Capital managing partner Hans Tung argues that scaling data center compute capacity is more critical to AI leadership than software models or engineering talent.
Notable Capital managing partner Hans Tung asserts that geopolitical dominance in the artificial intelligence era depends on physical infrastructure rather than software models or human talent. Comparing modern data centers to 19th-century railroads and steel ships, Tung argues that the ability to scale compute capacity is the primary driver of success in the AI race.
Tung notes that the United States currently holds a lead in data center compute capacity, with at least 47 gigawatts compared to 24 gigawatts in China. However, he warns that China has demonstrated superior execution in large-scale infrastructure projects, citing its achievements in shipbuilding and high-speed rail.
Tung maintains that the U.S. must successfully execute its own historic building boom to maintain its competitive edge. He argues that the availability of compute is the deciding factor for victory, stating that without sufficient infrastructure, having better engineers or thinkers is irrelevant.