Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Advanced machine intelligence (AMI), a new Paris-based startup co-founded by Meta’s former chief AI scientist Yan to beIt announced on Monday that it had raised more than $1 billion for its development Models of the world of artificial intelligence.
LeCun argues that most human thinking is grounded in the physical world, not language, and that models of the AI world are necessary to develop true human-level intelligence. “The idea of expanding the capabilities of LLMs (large language models) to the point where they will have human-level intelligence is complete nonsense,” he said in an interview with WIRED.
The funding, which values the startup at $3.5 billion, was co-led by investors such as Cathay Innovation, Greycroft, Hiro Capital, HV Capital, and Bezos Expeditions. Other notable supporters include: Mark Cubanformer 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 breed of AI systems that understands the world, has a persistent memory, can think and plan, and is 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 professor at NYU in addition to leading the startup. AMI will be LeCun’s first business endeavor since then departure From dead in November 2025.
LeCun’s startup is a bet against many of the world’s largest AI labs such as OpenAI, Anthropic, and even his former workplace, Meta, which believe expanding the pool of LLM degree holders will eventually provide AI systems with human-level or even super-intelligence. LLMs have powered viral products like ChatGPT and Claude Code, but LeCun has been one of the most prominent researchers in the AI industry to speak out about the limitations of these AI models. LeCun is known for being sincereBut as a pioneer of modern artificial intelligence, who won the Turing Award in 2018, his doubts carry weight.
LeCun says AMI aims to work with companies in manufacturing, biomedical, robotics, and other industries that contain a lot of data. For example, he says AMI could build a realistic, global model of an aircraft engine and work with the manufacturer to help it improve efficiency, reduce emissions, or ensure reliability.
AMI was co-founded by LeCun and many of the leaders he worked with at Meta, including the company’s former director of research science, Michael Rabat; Former Vice President for Europe, Laurent Sully; and former Senior Director of Artificial Intelligence Research, Pascal Fong. Other co-founders include Alexandre Lebrun, former CEO of AI healthcare startup Nabla, who will serve as CEO of AMI, and Saining Xie, a former Google DeepMind researcher who will serve as the startup’s chief science officer.
The state of global models
LeCun does not rule out the public benefit of LLMs. Rather, he believes that these AI models are simply the latest promising trend in the technology industry, and their success has created “a kind of illusion” among the people who build them. “It’s true that LLMs are getting really good at generating code, and it’s true that they’ll probably become more useful in a wide range of applications where code generation can help,” LeCun says. “That’s a lot of applications, but it will never lead to human level intelligence.”
LeCun has been working on global models for years within Meta, where he founded the company’s core artificial intelligence research lab, FAIR. But he is now convinced that his research is better done outside the social media giant. He says it became clear to him that the most powerful global modeling applications would be sold to other organizations, which didn’t fit well with Meta’s core consumer business.
As AI world models like Meta’s Predictive architecture for co-embedding (JEPA) became more complex “There was a reorientation of Meta’s strategy where it had to essentially catch up with the industry with its MBA (LLM) and do the same thing that other LLM firms were doing, and that’s not my thing,” LeCun says. “So, sometime in November, I went to see Mark Zuckerberg and told him. He’s always been very supportive of[universal modeling research]but I told him I could do it faster, cheaper and better outside of Meta. I could share the development cost with other companies…and his answer was, well, we can work together.”