Yann LeCun, Meta’s AI pioneer and Turing Award winner, is leaving the company after 12 years to start his own research venture. With this move, he aims to go beyond today’s AI trends and develop systems capable of understanding and reasoning about the real world — a vision he calls Advanced Machine Intelligence (AMI).

LeCun joined Meta (then Facebook) in 2013 and co-founded its FAIR (Facebook AI Research) lab, quickly becoming a central figure in AI development. Over the past 12 years, he has been instrumental in advancing deep learning, computer vision, and large-scale language models, shaping the technology behind Meta’s AI products.

Despite his success, LeCun has openly criticized the current focus on large language models (LLMs), calling them limited in achieving true human-like intelligence. His new startup will pursue Advanced Machine Intelligence, AI capable of perceiving, reasoning, and planning complex tasks in the physical world.

While he leaves the day-to-day operations at Meta, the company will remain a partner in his venture, collaborating on certain projects while he explores independent foundational research. This reflects a strategic divergence: Meta is increasingly prioritizing commercial AI applications, whereas LeCun seeks long-term breakthroughs in intelligence.

LeCun’s departure has drawn attention across the tech community. As one of AI’s most influential figures, his exit underscores debates about the future of artificial intelligence — balancing immediate products with deeper, world-model-based research.

Through his startup, LeCun aims to push AI boundaries, creating systems that are not only smarter but also better at interacting with and understanding the real world. His vision could define the next chapter in AI innovation, influencing both research priorities and the technologies that shape everyday life.