Fei-Fei Li: From Princeton Researcher to ImageNet, AI4ALL, and World Labs

Fei-Fei Li is the Chinese-American computer scientist whose work on the ImageNet benchmark dataset, developed at Stanford University in the late 2000s, contributed substantively to the conditions that made the contemporary deep-learning era possible. Li is the Sequoia Capital Professor of Computer Science at Stanford University, the founding co-director of the Stanford Institute for Human-Centered AI (HAI), and the co-founder of World Labs, a spatial-intelligence company launched in 2024. She is also the co-founder of the AI-education non-profit AI4ALL, served as Vice President and Chief Scientist of AI/ML at Google Cloud from 2017 to 2018, and published a memoir, The Worlds I See, in 2023.

Early Life in Chengdu and the Family’s Move to New Jersey

Fei-Fei Li was born in Beijing in 1976 and raised primarily in Chengdu, in southwestern China. She has written and spoken extensively — including in her 2023 memoir The Worlds I See — about a household that emphasised education even through periods of economic difficulty. The family emigrated to the United States in 1992, when Li was sixteen, settling in Parsippany, New Jersey.

The family’s early years in New Jersey involved substantial economic difficulty: her parents took a series of low-paying service jobs, and Li herself worked in a Chinese-restaurant kitchen and as a dry-cleaning attendant while attending Parsippany High School. She has cited the experience in subsequent interviews as a formative period that shaped her view of education as a practical tool rather than a credentialing exercise.

Princeton Undergraduate and the Early Caltech Research Years

Despite the difficulty of the family’s circumstances, Li was admitted to Princeton University on a full scholarship, completing a Bachelor of Arts in physics in 1999. Her undergraduate research focused on physics — but she has spoken about a growing interest in computational approaches to perception that drew her toward computer-vision research.

She subsequently pursued a PhD at the California Institute of Technology in electrical engineering, supervised by Pietro Perona and Christof Koch, completing her doctorate in 2005. Her doctoral work focused on visual recognition and the statistical learning of object categories — research that established the technical grounding for the ImageNet project that followed.

Stanford and the Building of ImageNet (2007–2009)

Li joined the Stanford University Department of Computer Science in 2009, after a brief assistant-professor appointment at Princeton. The signature project of her early Stanford years was ImageNet — a large-scale, hand-labelled image dataset organised according to the WordNet semantic hierarchy.

The ImageNet project, begun in 2007 and substantially completed by 2009, was unusual at the time both for its scale (eventually exceeding 14 million labelled images) and for the labour model required to produce it: human annotators recruited via Amazon Mechanical Turk applied labels under careful quality-control protocols. The annual ImageNet Large Scale Visual Recognition Challenge (ILSVRC), held from 2010 to 2017, became the central evaluation benchmark for computer-vision research in the deep-learning era.

The 2012 ILSVRC, in which a deep convolutional neural network (AlexNet) substantially exceeded all previous results, has been widely credited as the catalysing event of the contemporary deep-learning wave. The dataset Li and her collaborators built was the substrate on which that demonstration was possible.

Google Cloud Chief Scientist Tenure and the Policy-Engagement Chapter

From 2017 to 2018, Li served as Vice President and Chief Scientist of AI/ML at Google Cloud. The role brought her substantial public visibility, including engagements with policymakers in the United States and Europe and contributions to the broader public discussion of AI ethics and governance. She returned to Stanford full-time in 2018, while remaining engaged in industry-policy discussions in subsequent years.

The same period saw her named to several US technology-policy advisory bodies. Her 2024 appointment as senior advisor on AI policy in the State of California — and her engagement with multiple US National AI Research Resource working groups — has reflected an ongoing commitment to the public-policy dimension of AI research.

Founding AI4ALL and the AI-Education Non-Profit

In 2017, Li co-founded the AI-education non-profit AI4ALL with Olga Russakovsky and Rick Sommer. The organisation’s stated mission is to expand the diversity of the AI-research talent pipeline, primarily through residential summer programmes for high-school students from underrepresented groups in technology.

AI4ALL has, since its founding, expanded to multiple US university campuses and has placed thousands of students through its programmes. Coverage in The New York Times and other US press has positioned the organisation as one of the most-cited examples of structured outreach in the AI-research field.

The Worlds I See (2023) and Its Critical Reception

In November 2023, Li published her memoir The Worlds I See: Curiosity, Exploration, and Discovery at the Dawn of AI. The book combines a personal account of her family’s emigration from China and her early years in the United States with a chronological narrative of her career through ImageNet, Stanford HAI, Google Cloud, and AI4ALL.

The memoir was widely reviewed in The New York Times, The Financial Times, and The Atlantic, with reviewers consistently identifying the personal-narrative chapters as among the strongest sections of the book. The book was named to several end-of-year notable-book lists for 2023.

Co-Founding World Labs (2024) and the Spatial-Intelligence Company

In April 2024, Li and several Stanford colleagues — including Justin Johnson, Christoph Lassner, and Ben Mildenhall — formally announced the founding of World Labs, a company described in its initial communications as focused on building “spatial intelligence” foundation models capable of reasoning about three-dimensional scenes from limited visual input.

The company’s announcement was widely covered in Bloomberg, Reuters, and The Information. Public reporting indicated substantial venture-capital backing from Andreessen Horowitz and Radical Ventures, among others. As of 2026, Li continues to lead the company alongside her continuing Stanford appointment.

Career Timeline

• 1976 — Born in Beijing
• 1992 — Family emigrates to Parsippany, New Jersey
• 1999 — Graduates from Princeton (BA Physics)
• 2005 — Completes PhD at Caltech
• 2007–2009 — Builds ImageNet at Princeton and Stanford
• 2009 — Joins Stanford faculty
• 2017 — Co-founds AI4ALL
• 2017–2018 — Vice President & Chief Scientist of AI/ML at Google Cloud
• 2019 — Co-founds Stanford HAI
• November 2023 — Publishes memoir The Worlds I See
• April 2024 — Co-founds World Labs

Sources & References

• Fei-Fei Li — Wikipedia
• Fei-Fei Li — Stanford profile
• Stanford Institute for Human-Centered AI (HAI)
• World Labs — Official Site
• AI4ALL — Official Site
• The New York Times — Profile coverage

Frequently Asked Questions