The global artificial intelligence (AI) race is no longer just about training bigger models or building more computers. The real industrial contest now centers on applying AI to the physical world—into factories, ports, farms and transportation. This is the domain of “physical AI,” where intelligence is embedded directly into production systems.
As Rob Atkinson and Sejin Kim write in The Korea Times, the United States and China are already competing along clear strategic lines. The United States is pairing frontier innovation—foundational AI models, next-generation semiconductors and reshored advanced manufacturing—with national security priorities. China, by contrast, is aiming for full automation at scale. Its “dark factory” model—lights-out, fully roboticized production—illustrates a strategy built on speed, volume and centralized industrial coordination—but while still keeping low wages.
Korea sits in a different position. It cannot match the U.S. on foundational model innovation or China on manufacturing scale. But it holds a unique combination of strengths that neither possesses together: semiconductor fabrication, precision manufacturing, and world-class industrial robotics deployment. These are the core inputs required for physical AI.
Read the column.
