Fei-Fei Li, revered as the ‘Godmother of AI,’ traces her improbable ascent to the steam presses and threadbare counters of her family’s New Jersey dry-cleaning shop. Now, as founder of World Labs—a startup valued north of $1 billion—Li’s immigrant grit is reshaping artificial intelligence’s frontier. Recent funding rounds and product launches underscore her pivot from academic pioneer to venture-backed mogul, blending personal resilience with cutting-edge spatial intelligence tech.

Li immigrated from China at age 16, landing in Parsippany, New Jersey, in 1992 with her mother. The duo toiled long hours at the family’s dry-cleaning business, a stark contrast to Li’s later Stanford professorship and stints at Google. ‘My immigrant experience helped shape my approach to AI research,’ Li told Business Insider in a November 2025 interview, highlighting how those years instilled a pragmatic ethos now evident in World Labs’ mission to make AI understand the physical world.

From ImageNet to World Models

Li’s breakthrough came with ImageNet, the 2009 dataset that ignited deep learning’s explosion by providing millions of labeled images for training vision models. Her work at Stanford and Google Cloud, where she led AutoML, democratized AI tools. Yet, Li saw limitations: machines excelled at static images but faltered in dynamic, 3D environments. Enter World Labs, launched in 2024 with $230 million from investors including Andreessen Horowitz and NEA, per Yahoo Finance.

World Labs focuses on ‘spatial intelligence’—AI that comprehends and predicts physical spaces, crucial for robotics, AR/VR, and autonomous systems. In a Bloomberg interview, Li explained, ‘Teaching machines to see as humans do,’ emphasizing world models that simulate real-world physics, as covered in Bloomberg.

Immigrant Hustle Meets AI Ambition

Li’s childhood wasn’t just hardship; it was a masterclass in perseverance. After her parents’ divorce, her mother sold jewelry to fund the U.S. move. Li, struggling with English, ironed shirts by day and studied nights, earning straight A’s to attend Marymount Manhattan College on scholarship, then Princeton for her Ph.D. This ‘dry-cleaning CEO’ phase, as IndexBox dubbed it, forged her human-centered AI philosophy.

At World Labs, that translates to accessible tools. The startup’s Marble platform generates interactive 3D worlds from text prompts, empowering creators from filmmakers to educators. Li posted on X about Marble enabling high school theater productions, calling it ‘lowering the barrier of creating theatre and films,’ via posts from @drfeifei analyzed in recent web searches.

Funding Surge and Valuation Leap

By late 2025, World Labs’ valuation soared past $1 billion, fueled by demand for spatial AI amid robotics booms at companies like Figure and 1X. Initial $230 million came from Sequoia, DST Global, and others, but follow-on investments reflect hype. ‘AI pioneer Fei-Fei Li raises $230 million for new startup World Labs,’ reported Yahoo Finance in September 2024, with updates in Business Insider confirming unicorn status.

Li’s vision counters text-heavy LLMs like GPT. ‘What is a really really hard problem in AI? Spatial Intelligence,’ she tweeted in 2024, a post with over 1 million views. World Labs’ tech powers applications in design, learning, and beyond, positioning it against OpenAI’s Sora and Google DeepMind.

Spatial Intelligence’s Technical Edge

Unlike 2D generative video, World Labs builds ‘world models’—simulations predicting object interactions, gravity, and motion. This draws from Li’s BEHAVIOR benchmark for robot tasks like cooking, introduced in 2021 via Stanford research shared on X. Recent demos show AI agents navigating virtual homes, folding laundry—a nod to her past.

Industry insiders praise the approach. In a Times of India piece, Li advised, ‘Prioritise passion over perfection,’ urging AI pros to chase impact. Her Stanford HAI institute forecasts AI’s societal ripple, from jobs to ethics.

US-China AI Race Looms Large

Li navigates geopolitics astutely. Born in Beijing, she’s vocal on US-China tensions. ‘AI policy must encourage innovation, set appropriate restrictions,’ she posted on X in 2024. Bloomberg noted her worries over an ‘automated future’ and arms race, yet she pushes democratization.

World Labs hires globally but prioritizes U.S. innovation. Competitors like China’s Baidu eye similar tech, but Li’s dataset legacy—ImageNet trained models worldwide—gives her edge. Recent X buzz from @drfeifei hypes demos, with users lauding ‘richness of the creator’s imagination.’

Human-Centered AI in Practice

Li’s mantra: AI for good, for all. AI4ALL, her nonprofit, trains underrepresented youth, echoing her path. At World Labs, ethics infuse products—no black-box models, transparent spatial reasoning. ‘Human-centered AI is AI of the future,’ she wrote in a 2018 New York Times op-ed referenced across sources.

Challenges persist: compute costs for 3D sims rival video gen, talent wars rage. Yet Li’s track record—from Google to Stanford—suggests staying power. As DNyuz echoed Business Insider, her story inspires: from dry cleaner to AI titan.

Future Horizons for World Labs

Next: robotics integration, partnerships with Tesla or Boston Dynamics. Li eyes ‘superintelligence’ tempered by safety, per recent interviews. With valuation milestones hit, World Labs eyes IPO trails blazed by Anthropic. Her dry-cleaning days? A reminder that AI’s next leap may hinge on overlooked origins.