LeCun says Scale AI cofounder Alexandr Wang lacks the research background needed to lead Meta’s Superintelligence Labs.
Meta’s AI reshuffle followed internal frustration and controversy around Llama 4 benchmark results.
The episode highlights a deeper divide at Meta, with LeCun arguing that LLMs are unlikely to deliver true superintelligence.
Meta’s aggressive approach to make its world-class AI team has already put the company into the spotlight amid a wave of high-profile talent reshuffles across the industry. As the top researchers are switching firms in search of bigger bets, Yann LeCun, the pioneering AI researcher who recently stepped away from his role as Meta’s chief AI scientist, has openly questioned Mark Zuckerberg’s decision to place Alexandr Wang at the helm of the company’s new Superintelligence Labs.
SurveyDuring an interview with the Financial Times, LeCun mentioned some doubts about Wang’s readiness to lead cutting-edge AI research, arguing that the 28-year-old Scale AI cofounder is definitely quick to learn, but he lacks firsthand experience in running or practising academic-style research. According to LeCun, understanding what motivates top AI researchers and even what drives them away comes only with time spent inside research environments.
Wang’s appointment comes after Meta’s aggressive AI talent push, which included a $14 billion investment in Scale AI and Wang’s recruitment as a central figure in the company’s AI reorganisation. This marks a big shift for Meta after internal frustration about the progress of its Llama models, particularly following controversy around benchmark results tied to Llama 4.
LeCun also mentioned that those events broke Zuckerberg’s confidence in Meta’s existing generative AI leadership, leading to a broad sidelining of teams involved in the projects. Even after Wang briefly became his superior during the reshuffle, LeCun stated that the relationship was largely hands-off, underscoring that senior researchers are not typically “directed” in the traditional corporate sense.
The comments also exposed a deeper philosophical divide within Meta AI’s strategy. While Zuckerberg’s latest hires are heavily focused on large language models, LeCun has long argued that LLMs are fundamentally limited and unlikely to deliver true superintelligence. He reiterated that belief, saying he would not compromise his scientific views despite internal disagreement.
LeCun is now reportedly working on launching his own venture called Advanced Machine Intelligence. He will work as executive chair rather than CEO, acknowledging that his strengths lie in research vision rather than company management.
Follow Us
Ashish Singh
Ashish Singh is the Chief Copy Editor at Digit. He's been wrangling tech jargon since 2020 (Times Internet, Jagran English '22). When not policing commas, he's likely fueling his gadget habit with coffee, strategising his next virtual race, or plotting a road trip to test the latest in-car tech. He speaks fluent Geek. View Full Profile
