Tech

World-model companies keep AI product plans under wraps

AMI Labs and World Labs are attracting funding and attention while disclosing little about their commercial strategies, products or timelines.

Editorial persona
Owen Mercer
Markets and Finance Editor
Published
Draft
Source: TechCrunch · View original source
Man in a suit sits onstage beneath a spotlight, framed by two shadowed audience figures.
Artificial intelligence

World-model companies are drawing significant funding and industry attention while keeping their product plans and commercial targets largely confidential, according to TechCrunch.

AMI Labs co-founder Michael Rabbat said the company remained in a research and building phase and was not discussing product plans or timelines. AMI has explored manufacturing, biomedicine, robotics and medical AI through its Nabia partnership, but has not indicated which areas it will ultimately pursue.

World Labs’ Marble is described as a relatively developed demonstration platform, with potential applications including media creation, explorable video-game environments and computer-generated imagery. Its demonstrations show capabilities rather than a confirmed commercial product strategy.

The secrecy also extends to suppliers. Physicl chief executive Alex de Vigan said the data company did not know precisely how its material was being used and could produce more useful data with greater visibility into its customers’ work.

World models are intended to automate spatial intelligence, with possible uses ranging from navigable maps and robotics to interactive video and autonomous driving. TechCrunch attributed the industry’s guarded approach partly to easy fundraising and concerns that detailed disclosures could attract rivals, including other laboratories, neolabs, OpenAI and Anthropic.

Continue reading

More from Tech

Read next: Essay proposes royalty model to fund open-source maintainers
Read next: No Dogs in Space returns with Punk 2.0 and a New York Dolls deep dive
Read next: Why smart TV standby mode can matter beyond faster start-up