Advanced Machine Intelligence (AMI), a new Paris-based startup founded by Meta’s former chief AI scientist Yann LeCunannounced Monday that it has raised more than $1 billion to develop World examples of AI.
LeCun argues that most human thinking is based in the real world, not language, and that AI world models are necessary to develop true human-level intelligence. “The idea that you’re going to expand the capabilities of LLMs (large language models) to the point where they’re going to have human-level intelligence is pure nonsense,” he said in an interview with WIRED.
The funding, which values the startup at $3.5 billion, was led by investors such as Cathay Innovation, Greycroft, Hiro Capital, HV Capital, and Bezos Expeditions. Other notable contributors include Mark CubaFormer CEO of Google Eric Schmidtand French billionaire and communications executive Xavier Niel.
AMI (pronounced like the French word for friend) aims to build “a new type of AI systems that understand the world, have permanent memory, can think and plan, and are controllable and safe,” the company says in a press release. The startup says it will be global from day one, with offices in Paris, Montreal, Singapore, and New York, where LeCun will continue to work as a New York University professor in addition to leading the startup. AMI will be LeCun’s first commercial endeavor since leave from Meta in November 2025.
LeCun’s startup represents a bet against many of the world’s largest AI labs like OpenAI, Anthropic, and even his former workplace, Meta, which believe that adding LLMs will eventually produce AI systems with human-level intelligence or even super-intelligence. LLMs have used viral products such as ChatGPT and Claude Code, but LeCun has been one of the most prominent AI industry researchers speaking out about the limitations of these AI models. LeCun is well known for being to speak clearlybut as a pioneer of modern AI who won the Turing award in 2018, his skepticism carries weight.
LeCun says AMI aims to work with companies in industrial, biomedical, robotics, and other sectors that are data-rich. For example, he says AMI can create a real-world model of an aircraft engine and work with a manufacturer to help them improve efficiency, reduce emissions, or ensure reliability.
AMI was founded by LeCun and several leaders he worked with at Meta, including the company’s former director of research, Michael Rabbat; former vice president of Europe, Laurent Solly; and the former managing director of AI research, Pascale Fung. Other founders include Alexandre LeBrun, former CEO of AI healthcare organization Nabla, who will serve as AMI’s CEO, and Saining Xie, a former Google DeepMind researcher who will be the startup’s chief science officer.
The Case for World Models
LeCun does not exclude the general use of LLMs. Instead, in his opinion, these AI models are the latest promising trend of the technology industry, and their success has created “a kind of illusion” among the people who create it. “It’s true that (LLMs) are becoming very good at producing code, and it’s true that they’re probably going to be more useful in the larger application area where coding can help,” says LeCun. “That’s a lot of requests, but it won’t lead to human-level intelligence at all.”
LeCun has been working on world models for many years within Meta, where he founded the company’s AI Core Research lab, FAIR. But now he is convinced that his research is better done outside of the big social media company. He says it has become clear to him that the world’s top fashion brands will be sold to other companies, which do not fit well with Meta’s core consumer business.
As examples of AI worlds like Meta’s A Predictive Architecture for Co-Embedding (JEPA) became more sophisticated, “there was a reorientation of Meta’s strategy where it had to basically acquire the LLMs industry and do the same thing that other LLM companies are doing, which is not my interest,” says LeCun. “So sometime in November, I went to see Mark Zuckerberg and told him. He’s always been very supportive (of global model research), but I told him I can do this faster, cheaper and better outside of Meta. I can share the cost of development with other companies … His response was, OK, we can work together.”
