From factory floors to office spaces: Physical AI is ‘set to be enormous’
Good morning. Artificial intelligence is moving beyond the virtual world and into the physical one — and this shift is referred to as physical AI.
What does this mean? Unlike AI that operates only on computers or digital data, it allows machines to sense, think, and act in the real world. Physical AI was a key topic of discussion during a on Monday at CES 2026 in Las Vegas.
Andrew Nusca, editorial director of ’s Brainstorm series and author of the newsletter, hosted a fireside chat with President and CEO Cristiano Amon on the subject. Amon stated that he agreed with CEO Jensen Huang’s prediction that physical AI will be the next major wave of artificial intelligence.
“This is going to be massive,” Amon said. Physical AI is rooted in real-time sensor data. “You train on what you see, what you sense, what you do,” he explained. This enables robots and self-driving cars, for example, to handle complex tasks, adapt to changing environments, and make split-second decisions as they move and interact with the world around them.
Physical AI is emerging in the automotive sector, Amon continued. When you think about assisted driving and autonomous driving, those are physical AI problems, he explained. Sensors and cameras observe everything around the vehicle and guide it on where to go and when, he noted. Qualcomm has transformed into a major automotive technology provider, positioning itself as a key player in the industry’s shift toward “software-defined vehicles.”
Qualcomm’s move into automotive was driven by the company’s recognition of growing computing needs within vehicles — especially for processors used in assisted driving, Amon explained. “You can’t put a server in the trunk of a car,” he said. It uses too much power and drains the car’s energy, he added.
To meet these demands, Amon said, Qualcomm focused on creating power-efficient semiconductor designs that can handle heavy computing loads. Leveraging its experience in consumer electronics, the company can integrate cameras, sensors, and connectivity features into a single silicon chip — a design well-suited for modern vehicles.
“The same factors that made Qualcomm successful in automotive will make us successful in robotics,” Amon said. The conversation around physical AI, he added, naturally extends to robotics — and he believes its impact will go far beyond humanoid robots. At CES, Qualcomm announced a full suite of .
A by Deloitte explores how AI and robotics are converging. “Robots powered by physical AI are no longer limited to research labs or factory floors,” the report states. “They’re inspecting power grids, assisting in surgery, navigating city streets, and working alongside humans in warehouses.”
According to Deloitte’s research, the industry, regulators, and potential adopters are working to remove barriers that slow large-scale deployment. As organizations overcome these challenges, AI-enabled robots are expected to move from niche applications to mainstream adoption.
Sheryl Estrada