Yann LeCun, Meta’s chief artificial intelligence scientist and a Turing Award-winning pioneer of modern AI, is set to leave the company to launch his own AI start-up, marking a significant shift in Meta’s internal AI strategy. According to sources familiar with the matter, LeCun has already begun early fundraising for his new venture, which will focus on developing "world models"—next-generation AI systems designed to understand the physical world through visual and spatial data, rather than relying solely on text. These models aim to replicate human-like reasoning, perception, and planning, a long-term goal LeCun believes could take a decade to achieve. LeCun’s departure comes amid a major restructuring of Meta’s AI division under CEO Mark Zuckerberg, who has prioritized rapid commercialization of AI products over long-term research. This shift has strained relations with LeCun, who has publicly criticized the industry’s heavy reliance on large language models (LLMs), calling them "useful but fundamentally limited" in their ability to reason, plan, and understand the real world. He has long championed a more fundamental approach to AI, centered on building systems that learn from experience and environment—core tenets of the research he led at Meta’s Fundamental AI Research (FAIR) lab, which he founded in 2013. Zuckerberg’s strategy has moved FAIR away from basic research toward product development, particularly in LLMs. This pivot followed the underwhelming launch of Meta’s Llama 4 model, which failed to match the performance of rival models from OpenAI, Google, and Anthropic. In response, Zuckerberg has aggressively reshaped Meta’s AI leadership. He hired Alexandr Wang, founder of Scale AI, paying $14.3 billion for a 49% stake in his company and appointing him to lead Meta’s new Superintelligence division—now the top AI unit, to which LeCun reports. Zuckerberg also launched the elite TBD Lab, offering $100 million compensation packages to attract top AI talent from competitors. LeCun’s exit adds to a wave of leadership changes at Meta. Joelle Pineau, Meta’s vice-president of AI research, recently left for Cohere, while the company laid off 600 employees from its AI division. Meanwhile, Shengjia Zhao, co-creator of ChatGPT, joined Meta as chief scientist of the Superintelligence Lab, signaling a continued push toward competitive AI product development. The turmoil follows a sharp drop in Meta’s stock, which fell 12.6% in late October after Zuckerberg revealed that AI spending could exceed $100 billion in the coming year. The move sparked investor concern, contributing to a $240 billion market value loss. LeCun’s departure reflects a growing philosophical divide within Meta: between long-term, foundational AI research and the race to deliver commercially viable AI products. His new start-up could emerge as a direct competitor in the pursuit of artificial general intelligence (AGI), challenging Meta and other tech giants in the race to build machines that think and understand like humans. Meta closed Tuesday at $627.08, down 0.74%, and was trading slightly lower after hours at $627.00 on the NasdaqGS. The views expressed are those of the author and do not necessarily reflect those of Nasdaq, Inc.