Dr. Fei Fei Li

Dr. Fei-Fei Li is known as the “godmother of AI”: creator of ImageNet (2006–2012), former director of Stanford’s Artificial Intelligence Laboratory (SAIL), former chief AI scientist at Google Cloud, co-director of Stanford’s Human-Centered AI Institute (HAI), and co-founder of World Labs. Entered AI at Caltech in 2000; has been at the centre of every major AI inflection point since.

Key ideas

• ImageNet as the data breakthrough. The missing ingredient for AI was not better algorithms but large-scale labelled data reflecting lived perceptual experience. 15M images, 22,000 concepts, open-sourced. Led to AlexNet (2012) and the modern deep learning era.
• Spatial intelligence as the next frontier. Language is one slice of intelligence. World models — AI systems that create, reason in, and interact with 3D spatial environments — are the necessary complement to language models.
• Bitter lesson is harder in robotics. The training-data / objective alignment that made LLM scaling straightforward breaks in robotics: robots need actions in 3D worlds, but training data doesn’t contain 3D-grounded actions.
• Human-centered AI. “There’s nothing artificial about AI.” Technology is a double-edged sword; human dignity and agency must anchor development, deployment, and governance.
• AGI is a marketing term. Scientifically undefined and inconsistently used; AI (the original question: can machines think and act as humans?) is the correct north star.

Appearances

See also

• World Models
• Large Language Models
• Pretraining
• Scaling Laws
• Bitter Lesson