According to CNBC, Yann LeCun—Meta’s chief AI scientist and one of the original “godfathers” of modern AI—is leaving the company to start his own AI startup. The 65-year-old researcher announced his departure in a LinkedIn post on Wednesday, revealing plans to create a company focused on developing “world models” that analyze information beyond web data to better understand the physical world. LeCun joined Facebook back in 2013 to lead their FAIR AI research division while maintaining his position at NYU, and his departure comes during major upheaval in Meta’s AI unit following disappointing response to their Llama 4 model. Meta CEO Mark Zuckerberg has been spending billions recruiting new AI talent, including a $14.5 billion investment in Scale AI to bring aboard 28-year-old CEO Alexandr Wang as Meta’s new chief AI officer. The company will partner with LeCun’s startup, which aims to create systems with persistent memory, reasoning capabilities, and complex planning abilities.
The great Meta AI shakeup
Here’s the thing—this isn’t just another executive departure. LeCun has been the philosophical heart of Meta’s AI research for over a decade. His exit signals a fundamental shift in how Meta approaches artificial intelligence. Remember when Zuckerberg went on that massive hiring spree this summer? Bringing in the Scale AI CEO, former GitHub CEO Nat Friedman, ChatGPT co-creator Shengjia Zhao? That was basically a complete overhaul of the AI leadership team.
And it created some serious tension. Sources say LeCun rarely interacted with Wang or the new “TBD Labs” unit that’s now running Meta’s Llama development. We’re talking about a clash of cultures here—LeCun has always been an open-source champion, while Wang’s team apparently favors a more closed approach amid competition from OpenAI and Google. When you combine that with the October layoffs that hit 600 people from the Superintelligence Labs division (including some from LeCun’s original FAIR team), it’s not hard to see why he’d want out.
LeCun’s new vision versus current AI
So what exactly is LeCun building that’s so different? He’s been pretty vocal about his skepticism toward the current AI gold rush. While everyone’s pouring billions into large language models, LeCun thinks they’re fundamentally limited. He told the Financial Times last year that today’s AI “just doesn’t work” for true reasoning and planning. His new startup will focus on what researchers call “world models”—AI that actually understands physical reality, not just patterns in text data.
Basically, he’s chasing artificial general intelligence through a completely different path. Current LLMs are brilliant pattern matchers, but they don’t really get how the world works. LeCun wants systems that can reason about cause and effect, remember things over time, and plan complex actions. It’s the difference between a chatbot that can write a sonnet and an AI that could actually navigate a kitchen to make coffee.
What this means for AI development
This split represents a fascinating philosophical divide in AI research. On one side, you’ve got Meta, OpenAI, Google—all doubling down on scaling up existing transformer architectures. On the other, you’ve got pioneers like LeCun saying “Wait, we need to fundamentally rethink this whole approach.”
The partnership angle is interesting though. Meta isn’t cutting ties completely—they’ll still work with LeCun’s startup. That suggests they see value in his alternative approach, even as they pursue the mainstream LLM path. It’s like having a moonshot research division that just happens to be external now. For the broader AI ecosystem, having someone of LeCun’s stature pursuing a completely different architecture could lead to breakthroughs that the current scaling approach might never achieve.
Look, the timing is everything here. Meta’s AI division is in full rebuild mode, and one of its founding architects is walking out the door to pursue his vision independently. That tells you something about how messy and competitive this AI race has become. When even the godfathers are starting their own shops because they disagree with the direction of the field, you know we’re in for some interesting years ahead.
