Fei-Fei Li
The computer scientist who built ImageNet, the dataset that kicked off the deep learning era, and now runs World Labs.
Fei-Fei Li is a Stanford professor and one of the reasons modern AI works at all. In 2009 she built ImageNet, a collection of millions of labelled photographs, on the unfashionable theory that models were failing because they had too little data rather than because the algorithms were wrong. She turned out to be right, and the competition run on that dataset produced the breakthrough results that started the deep learning boom.
She led AI at Google Cloud from 2017 to 2018 and co-founded World Labs in 2024. Her current argument is that language models flatten the world into a line of words and therefore never really learn where anything is, which is a problem for anything that has to move through a real room.