Co-founder & CEO, World Labs · Stanford HAI · Expert Watch · monthly scan · our rolling analysis
Why we watch her. Fei-Fei Li sits directly on Dominic's frontier interests: world models, embodied AI and robotics simulation. She is among the clearest voices that after the LLM wave the next frontier is spatial intelligence, AI that perceives, generates and reasons about the 3D physical world, not just text. World Labs is the flagship case of taking foundational academic research (ImageNet) to a frontier-model company, and she pairs the thesis with a human-centered AI and democratization stance. A monthly tripwire on the world-models and spatial-intelligence debate.
Current headline view: spatial intelligence, machines that understand and generate 3D worlds, is AI's next frontier; "world models" are the missing piece beyond language.
Latest 1-page summary (as at 10 Jul 2026)
The eight core theses we track her against:
Spatial intelligence is the next frontier. LLMs cannot perceive or reason about physical space.
World models over LLMs. The coming wave is systems that perceive, generate and interact with the 3D world.
World models split into three functions. A functional taxonomy, not one monolith.
Spatial intelligence is the linchpin for creativity, simulation and embodied AI. 3D worlds become robot training and evaluation data.
Marble as proof-of-concept. Text, image, video and 3D-layout inputs to editable, persistent 3D worlds; research into shipping product.
Human-centered AI. AI should augment, not replace, humans.
Democratization of AI. The benefits should be broadly accessible, not concentrated.
Research-to-company frontier. Foundational academic work can seed frontier companies (World Labs, ~$1.23B raised, reported).
Method note. This v1 baseline is distilled from Fei-Fei Li's own Substack, her a16z and Bloomberg appearances, and the Lenny's Podcast interview (10 Jul 2026). It is not yet a full X post-by-post read. The monthly scan refines it from here, and flags any deviation.
Analysis backlog (newest first)
Deep source docs are held locally in the workspace at expert-watch/, not on this page.
10 Jul 2026 · v1 profile built: identity, surfaces, eight core theses and deviation baseline. Themes from her Substack, a16z and Bloomberg appearances, and the Lenny's Podcast interview.
Notable reactions (operators & researchers)
Up to three recent, high-signal reactions from operators or researchers (not commentators) to a specific new argument of hers. Challenges first. As at 10 Jul 2026.
No high-signal reactions logged this cycle (v1 baseline). The first weekly scan will populate this.
Recent media
Last 3 YouTube videos and last 3 podcast episodes, most recent first. As at 10 Jul 2026.