Former Meta AI chief Yann LeCun criticized the company’s new AI leadership, calling 29-year-old chief AI officer Alexander Wang “inexperienced” and warning of potential staff departures. Wang, co-founder of Scale AI, joined Meta in 2025 following the company’s acquisition of a 49% stake in his startup and now leads the new AI research unit, TBD Labs.
LeCun, who resigned from Meta in November, said Wang “learns fast” but lacks hands-on experience in AI research and the management of research teams. He also highlighted a loss of confidence in Meta’s Gen AI organization following allegations that the company manipulated benchmarks for its Llama 4 model, which resulted in a sidelining of the team by CEO Mark Zuckerberg.
Meta’s aggressive talent recruitment, including reported $100 million signing bonuses to lure engineers from OpenAI, reflects competition in the multibillion-dollar AI market. OpenAI has also raised the stakes by offering employees an average of $1.5 million in stock-based compensation, far exceeding peer startups, highlighting how top AI talent is commanding unprecedented pay across the sector. LeCun warned that pursuing “safe and proved” projects could cause Meta to fall behind in innovation.
LeCun added that large language models (LLMs) are “a dead end” for superintelligence, contrasting them with his new startup, Advanced Machine Intelligence Labs, which focuses on AI “world models” that learn from videos, data, and language. Health tech partner Nabla highlighted that LLMs still face structural limitations, including hallucinations and challenges in multimodal reasoning, making autonomous decision-making difficult.
