Fei-Fei Li

Fei-Fei Li is the scientist who taught machines to see—and then spent her career asking what the seeing was for. In the mid-2000s, when computer vision was stuck behind algorithms that collapsed the moment they met the messy variety of the real world, Li made a heretical diagnosis: the bottleneck was not the algorithm but the data. She set out to build a dataset so large and so comprehensively labeled that it would approximate the visual abundance a child takes for granted, and she assembled ImageNet across years of effort, ultimately distributing the labeling work across tens of thousands of crowdsourced annotators worldwide. In 2012, a neural network trained on her data obliterated the competition in the ImageNet Large Scale Visual Recognition Challenge, and the deep-learning revolution began. But the most interesting thing about Li is not the breakthrough; it is what she did afterward. Rather than retire into eminence, she turned the same intellectual seriousness toward the harder question: not what the machine can