In a striking revelation that has drawn attention from across the tech industry, Yann LeCun, a prominent figure in artificial intelligence and former chief AI scientist at Meta Platforms Inc., has expressed concerns regarding the company's newly appointed 29-year-old head of AI, Alexander Wang. Describing him as "inexperienced," LeCun warns that this lack of experience could lead to significant staff departures from the organization.
Wang, who is also a billionaire co-founder of Scale AI, stepped into the role of chief AI officer at Meta in 2025 following the company’s acquisition of a 49% stake in his startup. His appointment comes at a time when the competition for AI talent has reached unprecedented levels, with reports indicating that Meta offered staggering signing bonuses of $100 million to attract top professionals from leading firms like OpenAI, known for developing ChatGPT. This aggressive strategy reflects Meta's ambition to secure a dominant position in the rapidly growing multibillion-dollar AI sector.
LeCun, often referred to as one of the "godfathers of AI," decided to leave Meta in November. In an interview with the Financial Times, he pointed out Wang's youth and relative inexperience, noting, "He learns fast, and he knows what he doesn't know." However, LeCun cautioned that Wang lacks a deep understanding of research practices and the nuances that can either attract or deter researchers in the field.
At age 65, LeCun pointed out that Meta's CEO, Mark Zuckerberg, has seemingly lost faith in the entire team involved with artificial intelligence after the company faced allegations of manipulating benchmarks to enhance the perception of their Llama 4 model. He stated that Zuckerberg had effectively marginalized the whole Generative AI division, leading to a mass exodus of talent. "Many people have left, and those who remain may soon follow suit," LeCun added, emphasizing a troubling trend within the company.
Moreover, LeCun elaborated on the creative potential within the organization, stating, "We had a lot of innovative ideas that should be acted upon. Unfortunately, they opted for safer, more established approaches, which ultimately causes you to lag behind in this fast-paced sector."
When approached for comment, CNBC reached out to Meta but had not received any feedback by the time this article was published.
As for the future of Meta's AI ambitions, LeCun shared a thought-provoking perspective during the discussion about the company's recruitment efforts, saying, "Time will reveal whether this decision was wise or not." He also made a controversial assertion regarding large language models (LLMs), declaring them a "dead end" in the quest for superintelligence. "I'm sure many at Meta, including possibly Alex, would prefer I keep that opinion to myself," LeCun remarked, hinting at potential internal tensions regarding the company’s direction.
Turning to his new venture, Advanced Machine Intelligence Labs, LeCun is shifting focus towards developing what are called "world models"—AI systems that integrate learning from videos, data, and language. A recent collaboration with Nabla, a health-tech AI startup, emphasized the limitations of traditional LLMs, stating that they encounter structural issues such as hallucinations, unpredictable reasoning, and difficulties handling continuous multimodal data. These challenges significantly hinder their ability to make autonomous decisions.
This ongoing discussion raises critical questions about the leadership and strategic direction of AI development at major tech firms. Do you agree with LeCun's assessment of Wang's capabilities? What do you think the implications of these staffing issues could mean for Meta's future in the competitive AI landscape? Share your thoughts in the comments!
