Former Meta scientist and artificial intelligence pioneer Yann LeCun has raised sharp concerns over Meta Platforms’ recent overhaul of its AI leadership, describing the company’s new AI chief as too inexperienced for the role and warning that more senior researchers could leave the firm.
In an interview with the Financial Times, LeCun criticised the elevation of Alexandr Wang, the 28-year-old co-founder of Scale AI, following Meta’s aggressive push to regain ground in the global AI race under CEO Mark Zuckerberg
LeCun said Wang was “young” and “inexperienced” when it comes to running elite research teams, even though he acknowledged his intelligence and ability to learn quickly. According to LeCun, Wang does not yet fully understand what attracts or alienates top-tier AI researchers.
Also Read | Seven PlayStation games you can try free right now, no subscription needed
‘A lot of people will leave’
LeCun revealed that Wang briefly became his superior after Meta’s AI reorganisation but did not actively direct his work. He added that senior scientists typically do not respond well to being managed in ways that limit their independence.
More critically, LeCun warned that the leadership change could accelerate departures from Meta’s research ranks. “A lot of people have already left, and a lot of people who haven’t yet will leave,” he said, pointing to growing unease within the company.
Meta did not immediately respond to requests for comment.
Llama setbacks and a $14-billion bet
Wang’s rise at Meta followed the company’s $14-billion investment in Scale AI, a move widely seen as Zuckerberg’s attempt to inject urgency into Meta’s AI efforts. LeCun said the decision came after frustration inside Meta over slow progress on its open-source Llama models.
Internal confidence reportedly took a hit after Meta faced criticism for allegedly overstating benchmark results linked to Llama 4. LeCun said the episode angered Zuckerberg and led to much of Meta’s existing generative AI structure being sidelined.
Also Read | Planning new investments in 2026? Beware of fake trading courses and social media scams
Scepticism over large language models
LeCun, who announced in November that he is leaving Meta to launch a new venture called Advanced Machine Intelligence, also reiterated his long-standing doubts about large language models (LLMs).
He said he continues to view LLMs as a technological dead end and believes that true artificial intelligence will require entirely new architectures, even if that position clashes with Meta’s current strategy.
Meta has continued to double down on AI regardless, most recently acquiring Manus AI, a Chinese startup focused on autonomous agents, as part of a broader expansion of its AI portfolio.
