Nvidia CEO Jensen Huang recently reflected on what he would study if he were a 22-year-old entering university in 2025 — and surprisingly, it’s not what he originally pursued. In an interview during his Beijing visit, Huang said he would choose the physical sciences over the software or electrical engineering path he took decades ago. He explained that today’s technological evolution demands a deeper understanding of the physical world, which he believes will shape the next wave of artificial intelligence.
Huang introduced the idea of “Physical AI”, a concept that moves beyond generative capabilities like text creation or image recognition. According to him, AI has already advanced through key stages — from perception (recognizing images and voice) to generation (creating content), and now to a more advanced phase he calls Reasoning AI. This new phase is characterized by AI systems that can solve unfamiliar problems, recognize unseen conditions, and make decisions in real time.
This evolution, Huang believes, leads to “Agentic AI” — digital assistants that can operate independently like human workers. But to enable such AI to function in the real, physical world, he argues we must better understand fundamental forces like friction, inertia, and cause and effect. Only then can machines safely manipulate real-world environments.
“When you take that physical AI and you put it into a physical object called a robot, you get robotics,” Huang said. With rising global labor shortages and expanding factory infrastructures, he views intelligent robotics as an essential part of the future workforce.
Huang’s perspective comes as Nvidia — the company he co-founded in 1993 — becomes the world’s most valuable company. With its chips powering today’s top AI systems, Huang’s vision signals a shift in how AI will integrate into both virtual and physical aspects of life going forward.
