When an AI legend publicly questions your leadership choices, you've got a problem. Yann LeCun, the renowned AI pioneer who spent years at Meta Platforms Inc. (META), just delivered a scathing assessment of the company's recent AI shake-up in an interview with the Financial Times published Friday.

A 28-Year-Old Running Elite AI Research? LeCun Has Doubts

LeCun didn't mince words about Alexandr Wang, the 28-year-old Scale AI co-founder who recently became Meta's AI chief. The description: "young" and "inexperienced," particularly when it comes to understanding what makes elite AI research teams tick.

Sure, LeCun acknowledged Wang is smart and picks things up quickly. But knowing how to attract top researchers versus how to alienate them? That takes experience Wang simply doesn't have yet, according to LeCun.

Here's where it gets interesting: Wang briefly became LeCun's boss following Meta's AI reorganization. But LeCun made clear Wang wasn't actually directing his work, adding a pointed observation that senior scientists generally shouldn't be told how to do their jobs.

Meta declined to comment when contacted by MarketDash.

The $14 Billion Bet That Changed Everything

Wang's appointment didn't happen in a vacuum. It came on the heels of Meta's massive $14 billion investment in Scale AI, part of Mark Zuckerberg's aggressive campaign to catch up in the AI race.

According to LeCun, Zuckerberg grew increasingly frustrated with the slow progress on Llama, Meta's flagship open-source AI model. Things apparently came to a head after Meta faced criticism for allegedly overstating benchmark results related to Llama 4.

The controversy didn't just embarrass the company. It angered Zuckerberg enough that he effectively sidelined much of Meta's existing generative AI organization, LeCun said. The result? Internal confidence collapsed.

And the talent drain? It's real. "A lot of people" have already left Meta, LeCun warned, adding ominously that those who haven't departed yet "will leave."

LeCun's Controversial Take: LLMs Are a Dead End

LeCun, who announced in November he's leaving Meta to launch a new startup called Advanced Machine Intelligence, took the opportunity to double down on his controversial position about large language models.

People at Meta would prefer he stop calling LLMs "a dead end," he said. But he's not budging just because others think he's wrong. Real artificial intelligence, in his view, will require entirely new architectures beyond today's LLMs, even if that directly contradicts Meta's current strategic direction.

Meta Pushes Forward Despite the Drama

Meta isn't backing down from its AI ambitions. In December, the company acquired Manus AI, a fast-growing Chinese startup that specializes in autonomous agents.

The Manus deal adds to Meta's broader AI offensive this year, which includes the Scale AI investment that valued the data-labeling company at $29 billion and brought Wang into Meta's inner circle.

What Wall Street Thinks

Meta shares have gained 3.21% over the past 12 months, according to market data. That's not exactly setting the world on fire, but analysts see potential.

The consensus price target from 38 analysts sits at $831.64. The three most recent ratings from Rosenblatt, Baird and Wedbush are more bullish, averaging $937.33, which implies roughly 44.2% upside from current levels.

Interestingly, market analysis indicates Meta carries a bearish medium and long-term outlook, even while the stock shows positive momentum in the short term. It's the kind of mixed signal that reflects the uncertainty around Meta's massive AI investments and leadership turbulence.

The bottom line: Meta is doubling down on AI with billions of dollars and new leadership, but one of the field's most respected voices is publicly warning the strategy could backfire. Whether Zuckerberg's bet pays off or LeCun's predictions come true, the AI talent war is getting messy.