Fei-Fei Li's Former Students Now Lead World Labs, DAIR, and Princeton

Former Stanford Vision Lab PhD students co-founded or now lead four of AI's most closely watched organizations, including World Labs, which raised $1.23 billion since 2023. A breakdown of who went where and what they built.

6 min read
Fei-Fei Li, Stanford Vision Lab alumni network and World Labs, 2026
Fei-Fei Li speaking at the AI for Good Global Summit, Geneva, 2017.· Photo by ITU Pictures, via Wikimedia Commons (CC BY 2.0)

Former Stanford Vision Lab graduates founded or lead four of AI's most closely watched organizations, including World Labs, which raised $1.23 billion in spatial intelligence funding between 2023 and 2026. The alumni spread across commercial AI, academic computer science, and independent research institutes reflects both the breadth of Fei-Fei Li's research program and the field's rapid institutional diversification. (Bloomberg)

Justin Johnson and the $1.23 billion spatial intelligence bet

Justin Johnson completed his doctoral work under Li at Stanford and later served as an assistant professor at the University of Michigan, where he worked on visual reasoning, image synthesis, and 3D scene understanding. In 2023, he co-founded World Labs with Li, alongside Christoph Lassner and Ben Mildenhall, whose work on neural radiance fields gave the company a strong foundation in photorealistic 3D reconstruction.

World Labs emerged from stealth in September 2024 with a $230 million seed round that valued the company at $1 billion, backed by Andreessen Horowitz and Radical Ventures, among others. In January 2026, Bloomberg reported the company was in talks to raise additional capital at a $5 billion valuation, a fivefold increase in roughly fifteen months. The Series A, announced in February 2026 at $1 billion, brought in strategic investors including Nvidia, AMD, and Autodesk but closed without a publicly disclosed valuation.

World Labs launched Marble, its spatial intelligence product, in limited beta in November 2025, with the commercial release following in February 2026 alongside the funding announcement. Li described the company's mission at Bloomberg Tech 2026 in San Francisco: large world models that give AI systems a persistent, three-dimensional understanding of their environment rather than a flat, text-based one. StartupHub.ai data shows World Labs is the most heavily funded spatial intelligence startup in our database, having raised $1.23 billion since its founding.

Olga Russakovsky: from ImageNet benchmarks to Princeton and AI4ALL

Russakovsky completed her Stanford PhD in 2015 under Li. Her doctoral work was central to building and running the ImageNet Large Scale Visual Recognition Challenge (LSVRC), the annual benchmark competition from 2010 through 2017 that measured progress in object detection and classification. The final years of the challenge, in which AI models surpassed human-level accuracy, are now cited as a founding moment of the modern AI era.

After graduating, Russakovsky joined Princeton University's computer science faculty in 2017. She now directs the Princeton Visual AI Lab, where her group works on fairness in visual recognition systems, human-AI collaboration, and benchmark design. She received the NSF CAREER Award in 2022 and the Presidential Early Career Award for Scientists and Engineers, two of the most significant early-career honors in US scientific research.

In 2017, Russakovsky and Li co-founded AI4ALL, a nonprofit dedicated to increasing diversity in AI through programs for underrepresented high-school students at universities across the United States. Li chairs the organization; Russakovsky is a co-director. The pairing of a high-profile commercial startup (World Labs) and a nonprofit institution (AI4ALL), both directly involving Li's former students as co-founders, is an unusual double in AI's recent institutional history. The same Stanford Vision Lab lineage threads through a $1.23 billion startup and a nonprofit in the same year.

Timnit Gebru and the independent critical voice

Gebru completed her doctoral work at Stanford under Li before joining Microsoft Research as a postdoctoral researcher. She later became co-lead of Google Brain's Ethical AI team, a role focused on studying the risks and societal implications of large AI systems. Her departure from Google in December 2020 became one of the more widely discussed events in AI governance, centering on questions about corporate research independence, the publication of findings uncomfortable for technology companies, and who sets the agenda for AI safety research.

In 2021, Gebru founded the Distributed AI Research Institute, known as DAIR, as an independent, community-rooted research organization operating outside both big-tech corporate labs and traditional universities. Gebru serves as founder and executive director. DAIR's research focuses on the social, political, and community-level impacts of AI deployments, with a particular emphasis on populations underrepresented in mainstream AI development. It is one of the few AI research organizations structured to answer to neither a corporate parent nor a university administration.

Among the most prominent PhD graduates from Li's Stanford Vision Lab, two now lead for-profit AI ventures, one holds a named chair at a research university, and one directs an independent nonprofit research institute. Andrej Karpathy, another Stanford doctoral alumnus of Li's who co-founded Eureka Labs in 2023, extends the startup count. The three paths do not represent a single school of thought: Russakovsky's work centers on building better computer vision systems; Johnson's on commercializing spatial AI; Gebru's on scrutinizing both.

What it means

Li built ImageNet, helped establish the benchmark conditions that made deep learning commercially viable, and co-directed Stanford HAI during its founding phase. Her doctoral students are now, independently, building the companies, academic programs, and critical institutions that will shape what AI looks like through the 2030s. A mentor's legacy is typically measured in papers and citations; in Li's case, it is also measured in the organizations her graduates chose to build when they left her lab.

Sources

Editorial standards: every claim is sourced. Tips: [email protected]

© 2026 StartupHub.ai. All rights reserved. Do not enter, scrape, copy, reproduce, or republish this article in whole or in part. Use as input to AI training, fine-tuning, retrieval-augmented generation, or any machine-learning system is prohibited without written license. Substantially-similar derivative works will be pursued to the fullest extent of applicable copyright, database, and computer-misuse laws. See our terms.