From Meta to a New Frontier
Yann LeCun, the Turing Award-winning AI researcher and former Meta executive, has embarked on a major new venture. His startup, AMI, announced a monumental $1 billion (€890 million) funding round to develop artificial intelligence capable of understanding the physical world.
Investor Confidence and Corporate Structure
The funding round was led by five investment funds and attracted major corporate backers including Toyota, Nvidia, and Samsung. High-profile tech figures like former Google CEO Eric Schmidt and Amazon founder Jeff Bezos also participated. Prior to this raise, AMI was valued at €3 billion. The company is led by CEO Alexandre Lebrun, former head of French health-tech startup Nabla, with LeCun serving as non-executive president.
“When a company surpasses a billion valuation, it’s called a unicorn. We surpass three billion, so we are a triceratops,” LeCun remarked in an interview, highlighting the startup’s significant valuation.
Building “World Models” for Physical Understanding
AMI’s core mission is to develop “world models,” a different approach from the large language models (LLMs) behind chatbots like ChatGPT. These systems aim to comprehend and predict complex physical processes.
“The goal is to understand how the world works, in the manner of animals and humans,” LeCun explained. Initial applications could include analyzing jet engines, power plants, or patient physiology, with commercial deployments expected relatively soon.
Roadmap: From R&D to Universal Systems
The company, headquartered in Paris with offices in New York, Montreal, and Singapore, plans rapid expansion. LeCun stated the team will grow to 20-30 people imminently.
The development roadmap is clear:
- Year 1: Focus on core research and development.
- Year 2: Begin work on specific industrial applications with partners.
- 3-5 Years: Develop “somewhat universal intelligent systems” for applications like autonomous driving and robotics.
Ethical Considerations and Societal Role
LeCun addressed the critical debate surrounding AI ethics, particularly military applications—a point of contention among U.S. AI firms. He emphasized clear boundaries.
“There are things we must forbid ourselves from doing,” he stated. “In the end, the decision about the best use of AI for society should not be in the hands of someone like me or my colleagues. It’s up to society and its democratic institutions to decide.”
The funding marks a significant step for LeCun and his co-founders as they transition from foundational research at Meta to building applied AI systems designed to interact with and understand the real world.
