Unitree Founder Predicts Breakthroughs in Physical AI and Embodied Intelligence
At the 2026 World Physical AI Conference, Unitree founder Wang Xingxing predicted that physical AI and embodied intelligence could reach their 'ChatGPT moment' within the next 1 to 3 years, driven by future model iterations like GPT-6 and GPT-7. This perspective from a leading robotics figure highlights the rapid convergence of generative AI and physical hardware, suggesting that commercial robotics applications could soon see widespread adoption. Wang noted that the core challenge in crossing this threshold is aligning AI model inputs and outputs with real-world robots, and he expects commercial service deployments to outpace home and industrial adoption.
## BACKGROUND
Physical AI refers to autonomous systems that perceive, reason, and execute physical actions in the real 3D world by processing multi-modal data. Embodied intelligence combines physical hardware with AI, enabling a robot to interact dynamically with its environment.