On the assembly lines at GE Appliances’ 1.1mn sq ft cooker plant in La Fayette, Georgia, humans and robots work side by side. Large automated arms flip stoves upside down so workers can wire their bases, producing one unit every 15 seconds. Bill Good, head of manufacturing at the Chinese-owned company, says AI vision systems don’t get tired or miss defects, unlike human inspectors.

He is developing a live digital simulation of all plant processes to diagnose and fix issues in real time. Nvidia founder Jensen Huang predicts every industrial company will become a robotics company as AI expands into the physical realm. Elon Musk is investing tens of billions in Tesla’s shift to robotaxis and humanoids.

Deepu Talla, Nvidia’s vice-president of robotics and edge AI, expects physical AI-related revenues to grow from $10bn to $100bn over the next decade, citing that 90% of global actions occur in the physical world. Chris Miller, an economic historian at Tufts University, argues the US must apply AI more effectively than China to regain manufacturing supremacy. The Pentagon sees physical AI as vital to revitalizing the defence industrial base amid munitions shortages and competition from China.

Defence contractors are embedding AI in factories, while start-ups offer 3D printing and gecko-like robots for submarine hull inspections. Labour unions warn the trend risks devastating blue-collar workers as automation absorbs productivity gains. Ed Wytkind, senior adviser to the AFL-CIO Tech Institute, cautions against dangerous, de-skilling AI uses and insists workers—not tech billionaires—must shape the future of jobs.

In Cherokee, Alabama, Hadrian, a physical AI start-up founded by Australian Chris Power, is building a facility on a former railcar site to automate welding, fabrication, and machine tasks using AI coordination.