Lacan argues that most human reasoning is based on the physical world, not language, and that AI world models are necessary to develop true human-level intelligence. “The idea that you are going to expand the capabilities of the LLM [large language models] To the extent that they would have human-level intelligence is complete nonsense,” he said in an interview with WIRED.
The financing, which values the startup at $3.5 billion, was co-led by investors such as Cathay Innovation, Greycroft, Hero Capital, HV Capital and Bezos Expeditions. Other notable supporters include Mark Cuban, former Google CEO Eric Schmidt, and French billionaire and telecommunications executive Xavier Niel.
AMI (pronounced like the French word for friend) aims to create “a new breed of AI systems that understand the world, have persistent memory, can reason and plan, and are controllable and secure,” the company said in a press release. The startup says it will be global from day one with offices in Paris, Montreal, Singapore and New York, where Lekan will continue to serve as a New York University professor in addition to leading the startup. AMI will be the first commercial effort for Lacan following its departure from Meta in November 2025.
LeCun’s startup represents a bet against many of the world’s largest AI labs such as OpenAI, Anthropic, and even her former workplace, Meta, who believe that enhancing LLM will eventually provide AI systems with human-level intelligence or even superintelligence. LLM has powered viral products like ChatGPT and Cloud Code, but LeCun has been one of the AI industry’s most prominent researchers speaking out about the limitations of these AI models. LeCun is known for being outspoken, but 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 manufacturing, biomedical, robotics and other industries that have lots of data. For example, he says AMI can create a realistic world model of an aircraft engine and work with the manufacturer to help them optimize for efficiency, reduce emissions or ensure reliability.
AMI was founded by LeCun and several leaders who worked with him at Meta, including the company’s former director of research science, Michael Rabbat; former Vice-President of Europe, Laurent Soli; and Pascal Fung, former senior director of AI research. Other co-founders include Alexandre LeBrun, former CEO of AI health care startup Nabla, who will serve as AMI’s CEO, and Senning Xie, a former Google DeepMind researcher, who will be the startup’s chief science officer.
The case for the world model
LeCun does not dismiss the overall usefulness of the LLM. Rather, in his view, these AI models are simply the latest promising trend in the tech industry, and their success has created “a kind of confusion” among the people who create them. “This is true [LLMs] are becoming really good at generating code, and it’s true that they’re probably going to become even more useful in a broader range of applications where code generation can help,” LeCun says. “It has a lot of applications, but it’s not going to lead to human-level intelligence at all.”
LeCun has been working on world models for years inside Meta, where he founded the company’s fundamental AI research lab, FAIR. But now he is confident that his research is the best outside the social media giant. He says it became clear to him that the strongest application of the world models would be to sell them to other enterprises, which did not fit well into Meta’s core consumer business.
As AI world models like Meta’s Joint-Embedding Predictive Architecture (JEPA) became more sophisticated, “There was a realignment in Meta’s strategy, where it was basically to keep up with the industry on LLM and do the same thing that other LLM companies were doing, which is not my interest,” LeCun says. “So sometime in November, I went to meet Mark Zuckerberg and told him. He’s always been very supportive [world model research]But I told him I could do it faster, cheaper and better outside of Meta. I can share the cost of development with other companies… Their response was, OK, we can work together.
<a href