Yann LeCun, one of the world's leading AI researchers, holds a pen upright on its tip. "We don't have robots that are nearly as good at understanding the physical world as a rat," he says. The pen falls – any toddler knows it will topple – but no human bothers to guess which direction. An LLM like ChatGPT, LeCun argues, would try to generate a single prediction based on statistical patterns, and would almost certainly be wrong. "They're not particularly smart. They don't have an underlying understanding."
LeCun worked as chief AI scientist at Facebook-owner Meta for a decade before leaving in 2025 to found Advanced Machine Intelligence Labs (AMI Labs) in Paris. His goal: build a new type of AI that can handle the messy, unpredictable real world – something he says ChatGPT, Claude and Gemini can never do. "They're not a path towards human level or human-like intelligence, or even animal-like intelligence, because they cannot deal with real world data," he told me on the sidelines of VivaTech, France's leading tech conference.
“Yann LeCun says AI lacks understanding as Bank of England warns of financial risk.”
AMI Labs is developing a system called Joint Embedding Predictive Architecture (JEPA), which creates abstractions of reality to assess possible outcomes – a flexible approach LeCun believes is essential for tasks like household chores. Investors are betting billions on the idea. In one of Europe's largest seed funding rounds, AMI Labs raised more than $1bn (£760m) from backers including US chip giant Nvidia and the fund that manages Jeff Bezos's private wealth.
But while LeCun pushes toward smarter AI, the Bank of England has warned that the UK financial system faces increased risk from artificial intelligence. The warning underscores a paradox: even as today's AI systems lack genuine understanding, their growing role in markets and lending already threatens stability.