Frenchman Yann LeCun raises over a billion dollars for his AI startup AMI Labs

After spending more than a decade at Meta, French researcher Yann LeCun is opening a new chapter in his career. His young company, AMI Labs, has just announced a spectacular funding round exceeding one billion dollars, an amount rarely reached in a first round of financing.

With this operation, the AI startup aims to accelerate the development of a new generation of systems capable of understanding the physical world. An ambition that clearly distinguishes itself from the language models currently dominant in the AI ecosystem…

A record fundraising round for AMI Labs

Recently founded by Yann LeCun and several co-founders, AMI Labs has successfully raised $1.03 billion, or approximately €890 million. This Series A funding round values the company at around $3.5 billion, a particularly high level for a still young company.

The funding round brought together some twenty investors from the global tech sector and the French ecosystem. Among them are Nvidia, Toyota, and Samsung, as well as influential figures such as Jeff Bezos and Eric Schmidt. French players such as the Muliez and Dassault families, the Artémis group, and Xavier Niel also participated in the operation.

The company, headquartered in Paris, already has offices in New York, Montreal, and Singapore. Operational management is handled by Alexandre Lebrun, former head of the French startup Nabla, while Yann LeCun serves as non-executive chairman.

With humor, the researcher commented on the company's valuation, referring to on France Inter as a unicorn "exceeding three billion", which he compared to a "triceratops".

The ambition to create "models of the world"

AMI Labs' objective is to develop "world models", AI systems capable of understanding and predicting how the real world works.

Unlike ChatGPT or Gemini, these systems must learn to analyze complex physical processes and anticipate their evolution. The idea is therefore to reproduce a form of reasoning similar to that of humans or animals.

In concrete terms, such AI could, for example, anticipate the fall of an object, virtually simulate scientific experiments, or even model the operation of a power plant, an aircraft engine, or a human organ.

Ultimately, this approach opens up possibilities in several industrial sectors, notably robotics, autonomous driving, and medical research…

A research-focused roadmap

Initially, the company plans to concentrate its efforts on research and development. Industrial applications could emerge as early as next year, with discussions underway with various partners. According to Yann LeCun, the ambition within three to five years is to create versatile AI systems capable of being used in numerous fields that require intelligent machines…