LeCun’s New Startup Raises $1.03 Billion for AI Beyond LLMs
Advanced Machine Intelligence aims to build AI that can reason about the physical world rather than just generate text.
News
- LeCun’s New Startup Raises $1.03 Billion for AI Beyond LLMs
- Anthropic Sues Pentagon Over “Supply Chain Risk” Blacklist; Tech Rivals Back Legal Challenge
- Claude AI Can Do Months of Work in Minutes, Says Ark Invest’s Cathie Wood
- US-Iran Escalation: The Rising Risks to Middle East Projects and Investments
- Amazon Faces Dual Disruptions: Drone Strikes Hit AWS Sites as Retail Platform Stumbles
- Enterprise AI Scales Fast, but Structural Change Lags, Study Finds
Advanced Machine Intelligence (AMI), a startup founded by former Meta chief AI scientist Yann LeCun, has raised $1.03 billion in funding at a $3.50 billion pre-money valuation. The company aims to commercialize a new class of artificial intelligence systems built around reasoning, planning, and so-called “world models”—an approach LeCun has long argued is necessary for building truly intelligent machines.
The funding round was co-led by Cathay Innovation, Greycroft, Hiro Capital, HV Capital, and Bezos Expeditions. The scale of the investment signals growing investor interest in alternative approaches to AI development, at a time when the field is largely dominated by large language models (LLMs).
LeCun has been one of the most prominent critics of the idea that today’s LLMs—systems trained primarily to predict the next word or token in a sequence—can eventually lead to human-level intelligence. Instead, he argues that such systems lack a deeper understanding of the physical and causal structure of the world.
AMI is positioned as a test of that thesis. According to LeCun, the company’s goal is to build AI systems capable of reasoning about complex environments, planning actions, and interacting with the physical world more effectively than current generative models.
In an interview with Reuters, LeCun said the company is focusing first on industrial applications where decision-making in complex systems is critical. Potential customers include manufacturers, automakers, aerospace companies, biomedical firms, and pharmaceutical organizations.
“We want to become the main provider of intelligent systems, regardless of what the application is,” he said.
LeCun joined Meta in 2013, where he founded Facebook AI Research—later renamed FAIR—and became one of the company’s most influential AI leaders. He left the company at the end of 2025.
Even as LeCun pursues a different technological direction, Meta has continued to expand its efforts in large language models. In June 2025, the company reorganized its AI initiatives into a new division, Meta Superintelligence Labs, led by former Scale AI CEO Alexandr Wang.
LeCun said AMI’s technology could eventually power consumer applications, particularly robotics. One example he pointed to is domestic robots that require a level of common sense to navigate everyday environments. He also said the company is discussing potential uses of the technology with Meta, including possible integration into Ray-Ban Meta smart glasses.
