Yann LeCun, the Godfather of AI and a leading figure in artificial intelligence and head of a research lab at Meta Platforms Inc., has informed his team that he will leave the company at the end of this year to establish a new venture.
This announcement, made in a memo obtained by Bloomberg News, signals a significant transition for LeCun, who has been part of Meta since 2013.
In a memo shared with his colleagues, LeCun revealed that Meta intends to collaborate with him on his upcoming startup, although the specifics of this partnership are still being worked out.
He emphasized that Meta will have access to the innovations developed by his new company, which he is building to explore Advanced Machine Intelligence.
LeCun aims to create a company that focuses on AI systems capable of interpreting visual and sensory information.
He believes this technology will revolutionize AI by enabling machines to comprehend the physical world, maintain memory, reason effectively, and execute complex plans.
He explained that the startup’s mission is to drive a major transformation in AI, emphasizing the importance of developing systems that can understand and interact with their surroundings.
This venture will build on research he has conducted with colleagues at Meta and New York University, where he also teaches.
According to LeCun, the technology developed by his new company will have wide-ranging applications across various industries, some of which align with Meta’s interests, while others do not. He sees establishing this goal within an independent entity as a way to maximize its potential impact.
A spokesperson from Meta confirmed LeCun’s departure, although he has not yet publicly commented on the matter. Known as one of the pioneers of AI, LeCun’s work at Meta has primarily centered on long-term research, which may not yield immediate consumer benefits.
LeCun reportedly faced challenges in securing resources for his projects at Meta, as the company shifted its focus toward competing with rivals like OpenAI, Google, and Anthropic PBC. This shift in priorities created tension within the organization.
He has been vocal about his belief that Meta and other AI firms are making a mistake by prioritizing large language models (LLMs) as the key to achieving human-like intelligence, advocating instead for world models like JEPA as a more effective approach.
While Meta’s main AI product, Llama, relies on an LLM, LeCun’s criticisms of the company’s strategies led to conflicts with some colleagues.
This discord became more pronounced as Meta sought external talent to lead its new AI lab, investing heavily in recruiting top researchers.
Earlier this year, after the release of Meta’s Llama 4 model, which did not meet industry expectations, some employees began to distance themselves from LeCun, fearing that his views might not align with the company’s current AI strategy.
LeCun has also been a strong advocate for open-sourcing technology, a stance that contrasts with Meta’s recent shift away from this approach. Reports about his plans to start a new company have circulated in the media, highlighting his ongoing influence in the AI field.
Other Stories You May Like