In November, Yann LeCun left Meta after 12 years due to a disagreement with Mark Zuckerberg on the future of AI Frustrated by the limitations of major linguistic models (LLMs), the French computer scientist founded AMI Labs, a Paris-based startup focused on developing “world models.” The startup announced today (March 10) that it has raised $1 billion in what is Europe’s largest seed round.
The funding makes AMI $3.5 billion in funding and includes a number of high-profile backers such as Nvidia, Mark Cuban, Eric Schmidt, and Jeff Bezos, who co-led the venture along with venture capital firms Cathay Innovation, Greycroft, Hiro Capital, and HV Capital. LeCun will serve as executive chairman, guiding AMI’s long-term goal of building AI systems capable of understanding complex real-world data.
LeCun hired Alex LeBrun, a former partner at Meta, as CEO of AMI. LeBrun previously worked on Meta’s Fundamental AI Research (FAIR) group, which LeCun led for more than a decade. He also founded Nabla, a startup that builds AI tools for nurses.
In a LinkedIn posts today, LeBrun explained that “global models,” unlike current LLMs, can predict the results of actions and plan sequences to accomplish tasks. “AMI will advance AI research and develop applications where reliability, controllability, and safety are truly important,” he wrote, citing areas such as automation, robotics, healthcare and wearable devices.
LeCun, who moved from France to the US in the 1980s to work at AT&T Bell Laboratories, became a seminal figure in machine learning. He received the 2018 Turing Award for groundbreaking work in neural networks.
His departure from Meta last November followed his public criticism of LLMs as “dead end” to achieve truly intelligent systems. The move coincided with Meta’s internal reorganization around Alexandr Wang, the 29-year-old founder and former CEO of Scale AI, who now leads Meta Superintelligence Labs (MSL). LeCun has been critical of Wang’s leadership, calling him “the ignorant.” Despite their differences, LeCun said AMI could one day collaborate with Meta on projects like its Ray-Ban Meta smart glasses.
Other key members of AMI’s founding team include Saining Xie, a professor of computer science at New York University who will serve as chief scientific officer. Former Meta executives Laurent Solly and Michael Rabbat are stepping up as AMI’s chief operating officer and vice president of global models, respectively. Pascale Fung, a professor of engineering at the Hong Kong University of Science and Technology, is the chief research and innovation officer.
The company is also hiring across offices in Paris, New York, Montreal and Singapore.
According to PitchBook, AMI’s $1 billion raise is Europe’s largest seed round to date and marks another milestone in the developing world model sector. Other key players include Fei-Fei Li’s World Labs, which raised $1 billion last month, and General Intuition, which raised $134 million in October.
World models, or AI systems trained to simulate and reason about the real world rather than just generating text, are attracting increasing interest from researchers who doubt that LLMs can understand reality. AMI plans to focus first on developing the world model research before pursuing commercial applications, with possible early use cases in industry and hospitals.
One of AMI’s first external partners will be Nabla, where LeBrun will remain chairman and chief AI scientist. The partnership will give Nabla early access to AMI research, testing how AI based on a world model performs in a fast-paced, high-quality clinical care setting.
