Meta’s former AI scientist, Yann LeCun, helped lay the groundwork for modern AI long before chatbots became industry buzzwords. Now, after leaving Meta in late 2025, he built a multibillion-dollar startup under the premise that the race toward AI superintelligence is starting from the wrong place. Speaking at the VivaTech conference on June 17, LeCun share more details about why he believes his startup, Advanced Machine Intelligence (AMI Labs), is on a better path toward human-level AI Earlier this year, the Paris-based company. $1.03 billion in revenue in the amount of $3.5 billion, giving LeCun and his team new support to pursue an alternative to today’s artificial intelligence models developed by tech giants.
“I think there is a need for new paradigms to go beyond the limitations of the current systems,” LeCun, 65, told Wired’s Steven Levy on stage.
Building better language models, as many AI companies are doing, could be one way, LeCun said, but “it’s a slow way.” According to LeCun, AMI Labs builds “global models” that learn from reality, understand outcomes, predict what happens next, and choose the best actions based on those predictions. “We’re very optimistic that this will be a complete revolution in smart programming,” LeCun said.
LeCun believes that much of Silicon Valley is already “live on the LLM” – even though it is overly confident that scaling large language models will eventually produce human-like intelligence. The result, he said, is a “monoculture” of AI systems that are powerful in some areas but still fundamentally limited.
While LLMs are already highly skilled in areas such as coding, they are built on language. Real-world data from cameras, robots, and other sensors is messy and difficult to predict, which is why he believes a different approach is needed. In AMI labs, the result can be “The most intelligent systems that can anticipate the outcome of their actions, that are able to plan, think and maybe up to the intelligence of people,” he said.
From LeCun, the criticism carries weight. A long-time professor at New York University, he is among three researchers who are often called the “Godfathers of AI” The others are Geoffrey Hinton, a professor at the University of Toronto and who worked at Google, and Yoshua Bengio, a professor at the University of Montreal who founded Quebec’s AI center, Mila. In 2018, see jointly won the prestigious Turing Award for achievements that have helped push deep learning into the mainstream.
From FAIR to AMI Labs
In 2013, LeCun he entered Facebook to launch what would become Facebook AI Research, or FAIR, and later transitioned to chief AI scientist at Meta. He said AMI has its roots as an internal FAIR project and has the support of corporate leaders, including Mark Zuckerberg and CTO Andrew Bosworth. But as the priorities of the Meta’s AI shifted to finding opponents, the environment became less attractive.
At that point, it made sense to “just leave the company, try to accelerate development, and build things that we can build with,” LeCun said. “And I realized that I could raise enough money to continue this.”
In March, AMI Labs raised $1.03 billion in seed funding—the largest in European startup history. Investors include Nvidia, Jeff Bezos and Eric Schmidt, among others. With LeCun serving as executive chairman, the startup is led by CEO Alex LeBrun, FAIR’s former head of engineering and founder of AI health company Nabla, an early partner of AMI Labs. The company has assembled a strong research team with experience in Meta, Google DeepMind and other leading AI labs.
AMI says it aims to develop applications in areas where this approach is important, including industrial process control, automation, wearable devices, robotics, and healthcare. “It won’t happen next year,” LeCun said. “We’re not going to have a world of geniuses in the data center next year, okay?”
At VivaTech, the conversation also expanded to a bigger question: who should control the next generation of AI? If the assistants of the future are created only by “a few companies” in Silicon Valley or China, LeCun warned, “the culture is in big trouble” and “democracy is in big trouble.” Banning open source models in the name of security, he said, misses the role AI can play in spreading information.
That position also informs LeCun’s work as the chief scientific advisor to Project Tapestry, an effort led by the AI Alliance to build a global AI model shared by a distributed network of contributors. The AI Alliance is a non-profit group, jointly launched in 2023 by Meta and IBM, to support the development of open source AI. With Tapestry, the goal is to allow countries, companies, universities and others to donate data or computing power without giving up control over that data. As AI begins to mediate what people read, watch, and search, LeCun says people need alternatives to what a few prominent American or Chinese companies decide to build.
LeCun’s comments also come as Meta’s open source situation has grown increasingly difficult. Llama 2, released in 2023 for free trade, makes the Meta a rare Big Tech counterweight to the more closed AI opponents. But Zuckerberg is since signed that Meta may not release all of its future “spy” models publicly.
“We need access to a wide variety of AI assistants for the same reason we need access to a wide variety of media to access multiple sources of information,” LeCun said. “The only way I see this happening is if there is an open source, free base model, upon which anyone can build their own personal assistant in their language or languages, their culture, their value system, their political biases, their interest centers. So open source has to exist.”
