What is Physical AI (embodied intelligence)? Why Huang calls it a $40T market
AI is stepping out of the screen into the physical world: embodied intelligence and humanoid robots. Jensen Huang calls humanoid robots a ~$40T market and robotics Nvidia’s second growth curve after AI; SoftBank’s Son calls physical AI the birthplace of the next trillion-dollar company. We view this newest theme through the investor-consensus lens.
What “Physical AI” actually means
For two years, AI mostly lived on screens — generating text, images, code. “Physical AI” (embodied intelligence) gives AI a body: to perceive, act, and do work in the real world, most visibly through humanoid robots. Jensen Huang keeps repeating that humanoids map to a ~$40T labor-automation market — Nvidia’s second-biggest growth opportunity after AI.
The “sim-to-scale” value chain
The chain has layers: compute and the “brain” (Nvidia’s Isaac/GR00T platform, plus Halos, a full-stack robotics safety system launched June 2026); vehicles and manufacturing scale (Tesla Optimus, Hyundai/Boston Dynamics, Xiaomi, UBTECH); and the still-private pure-play leaders (Figure, Unitree, 1X, Apptronik, Agility). For public-market investors, Nvidia is the smoothest layer — the compute base for both AI and robots.
The consensus lens on the newest theme
The physical-AI story is seductive, but the pure-play leaders are mostly private and valuation runs ahead of deployment. That’s exactly where the consensus lens helps: rather than chase the hottest robot name, see which **listed** exposures the legends actually back with real money — Nvidia is the layer several of them hold across styles. The site now breaks out “Physical AI & Robotics” as its own layer; see its bull/risk and key names on the AI market page.
Sources: 24/7 Wall St. — Huang: humanoid robots a $40T market · CNBC — Humanoid robots touted as next AI investment opportunity · KraneShares — Humanoid Robotics in 2026