Yann LeCun, chief AI scientist at Meta Platforms Inc., at the VivaTech conference in Paris, France, on Wednesday, May 22, 2024.
Bloomberg | Bloomberg | Getty Images
Meta’s former chief AI scientist, Yann Lecun, has called its 29-year-old AI boss “inexperienced,” and warned of a staff exodus from the company.
Alexander Wang, the billionaire co-founder of Scale AI, joined Meta as its chief AI officer in 2025 after the tech giant bought a 49% stake in his startup.
His hiring came amid an intensifying AI talent war, in which Meta reportedly offered $100 million signing bonuses to poach top talent from ChatGPT maker OpenAI, as it raced other hyperscalers to dominate the multibillion-dollar AI market with the most advanced models.
Lecun, who’s known as one of the “godfathers of AI” and quit Meta in November, called Wang “young” and “inexperienced” in an interview with the Financial Times.
“He learns fast, he knows what he doesn’t know,” Lecun said of Wang, who heads up Meta’s new AI research unit TBD Labs and is tasked with building new AI models. But Lecun added: “There’s no experience with research or how you practise research, how you do it. Or what would be attractive or repulsive to a researcher.”
Lecun, 65, added that Meta CEO Mark Zuckerberg “basically lost confidence in everyone who was involved” after the company was accused of gaming benchmarks to make its Llama 4 model look more impressive. Zuckerberg “basically sidelined the entire Gen AI organization,” Lecun said.
“A lot of people have left, a lot of people who haven’t yet left will leave,” Lecun added.
“We had a lot of new ideas and really cool stuff that they should implement. But they were just going for things that were essentially safe and proved,” he added. “When you do this, you fall behind.”
CNBC contacted Meta for a response, but had not received a response as this story went live.
AI superintelligence is a ‘dead end’ for LLMs
When asked about Meta’s AI hiring drive, Lecun said: “The future will say whether that was a good idea or not.”
But he added that “LLMs basically are a dead end when it comes to superintelligence.” “I’m sure there’s a lot of people at Meta, including perhaps Alex, who would like me to not tell the world that,” Lecun said.
Lecun’s new startup, Advanced Machine Intelligence Labs, is focusing on “world models:” AI systems that learn from videos and al data as well as language.
Nabla, a health tech AI startup that partnered with Lecun’s company in December, said in a press release that, unlike world models, LLMs “still face some structural constraints, including hallucinations, non-deterministic reasoning, and limited handling of continuous multimodal data, which make autonomous decision-making challenging.”
