The biggest opportunity in physical AI may not be the robots themselves—it could be the mountains of video they create. Speaking on Seagate Technology Holdings PLC‘s (NASDAQ:STX) fiscal fourth-quarter earnings call, CEO Dave Mosley said investors are focusing on humanoid robots and autonomous vehicles, but the real story is the data behind them.

“People focus very much on the end product—the robot, if you will—but I actually think it’s a lot more of a data play,” he said. “In particular, the physical AI stuff is largely more about video.”

Why Video Could Be Physical AI’s Biggest Asset

According to Mosley, robots are essentially networks of cameras and sensors that continuously generate unstructured video. That data is uploaded to local or cloud data centers, where AI models use it to build “world models” that help machines understand and interact with the physical world.

The more robots operate, the more video they create—and the more historical footage companies need to retain and reuse for future training.

For Seagate, that means physical AI could become a powerful long-term driver of storage demand. Rather than viewing robotics as simply another compute story, the company sees it as a data explosion that will require increasingly large amounts of high-capacity storage to preserve years of video used to train the next generation of AI systems.

Image: Shutterstock