World model startups AMI Labs and World Labs stay quiet on products

World models — technology aimed at automating spatial intelligence, with potential applications from robotics to interactive video to self-driving systems — are one of the most closely guarded corners of AI. The field’s two biggest players, Yann LeCun’s AMI Labs and Fei-Fei Li’s World Labs, have both drawn significant funding and attention but rank low on the trying-to-make-money scale, according to a TechCrunch reporter who moderated a panel on world models at the All In conference this week.

Pressed on the panel about where the technology would actually be commercialized, the reporter found few clear answers. Michael Rabbat, an AMI Labs co-founder and the company’s VP of World Models, was the closest thing to an authority in the room, but stayed cagey when asked what AMI is working on: “We’ll talk about it when we’re ready to talk about it,” he said, later clarifying by email that “we’re still in a research and building phase, so we’re not talking publicly about any product plans or timeline.” AMI is less than a year old, so some caginess is reasonable — but the secrecy extends across the world-modeling space. World Labs’ Marble is probably the most fully developed product in the field, with demos ranging from media creation to building explorable game environments to CGI effects and some robotics use cases, though the platform still seems designed mainly to demonstrate capabilities rather than sell a defined product.

The secrecy even extends to suppliers: Alex de Vigan, CEO of data supplier Physicl, said he knows his company’s data has been useful to the world-model labs building with it, but is still in the dark about what exactly they’re building — “I wish they would tell us more. We could build more useful data if we knew what they were working on.” Part of the mystery comes from how versatile world models are: the same underlying approach that helps a self-driving car navigate traffic could also help a humanoid robot carry boxes or turn video footage into an explorable environment, and AMI has already touched manufacturing, biomedicine, robotics, and AI software for doctors through a partnership. As long as fundraising stays easy, labs have little pressure to commit publicly to one direction — announcing a breakthrough too early would draw competitors, including OpenAI and Anthropic, into the same race sooner.

Why it matters

The world-model field’s collective silence about commercialization means the technology’s first real business model — robotics, entertainment, self-driving, or something else — may only become visible once one lab decides the competitive cost of staying quiet outweighs the benefit.