When Fei-Fei Li was a Princeton undergraduate plotting her future in computer vision research, she had plenty of options for graduate school. The lure of Caltech, ultimately, came down to the Institute’s easy interdisciplinarity—and its people.

“I chose Caltech because of my co-advisers Pietro Perona, an AI professor, and Christof Koch, a neuroscientist,” she says. “I had the rare opportunity to work at the intersection of cognitive neuroscience and computer science. That was very enticing for me.”

Her efforts as a scientist at Stanford have pushed the field of artificial intelligence forward. Li’s signature contribution is ImageNet, a massive database she launched in 2009 with millions of labeled images. The repository constitutes a major resource behind the ever-expanding capabilities of computers to discern what they “see.”

“I was very interested in cracking the Holy Grail problem of computer vision, which is object recognition,” Li says. “Looking back, I’m pleased to see that so many people have found ImageNet helpful, including researchers in my own lab.”

She remembers her time at Caltech as filled with invigorating conversations with her fellow trainees and her mentors, whom she describes as “wonderful scientists full of imagination and very fun personalities.” Sample topics included Bayesian statistical models for object recognition, the abstract art of Mondrian, and the origin of Nutella.

“I have such fond memories of the intellectual energy and the people,” she says. “The labs were simple spaces, but I had the chance to be with scientists from around the world. It was always full of ideas and discussions and energy.”

These days, Li’s investigations explore how computer systems can be deployed to reduce medical errors. At the same time, she’s using her platform as a leading voice in AI research for good.

The educational nonprofit she co-founded, AI4ALL, works to increase diversity in the field. And the Stanford Institute for Human-Centered AI, an interdisciplinary research hub she co-founded in 2019, engages the social sciences and humanities to ensure that people remain at the center of the conversation about AI.

“AI has the potential to radically transform every industry and every society,” Li says. “As a first-generation AI researcher, it’s simultaneously a privilege and a responsibility to help guide AI technology and application to benefit all humans’ lives, not just a privileged few.”

— Wayne Lewis