Fei-Fei Li urges technologists to better communicate AI benefits

Fei-Fei Li has a message for her fellow AI researchers: stop scaring people, stop overpromising, and start talking about what this technology actually does. The Stanford professor and co-founder of World Labs used a keynote panel at the Ai4 2026 conference in Las Vegas to argue that the AI community’s communication problem has become a risk factor in its own right. Sharing the stage with Geoffrey Hinton and Andrew Ng, two of the most recognizable names in machine learning, Li made the case that the pendulum swings between utopian hype and apocalyptic dread are doing more damage than most technologists realize. Public opposition fueled by bad information, she argued, could slow adoption in areas where AI stands to do genuine good. The messaging problem Li’s critique cuts in both directions. On one side, you have the Silicon Valley tendency to frame every new model release as a civilization-defining leap. On the other, you have doomsday warnings about superintelligent systems that treat science fiction as a policy document. As co-director of Stanford’s Human-Centered AI institute, she has built much of her academic reputation around the idea that AI discourse needs to be rooted in ev...