AMI Labs, Yann LeCun’s newly launched AI startup, is emerging as a pivotal player in the next wave of artificial intelligence research, centered on developing “world models” — AI systems that understand and interact with the physical world in a persistent, reasoning-driven manner. LeCun, Meta’s former chief AI scientist and a Turing Award laureate, stepped down from Meta to co-found AMI Labs, positioning it as a direct counterpoint to the dominant large language model (LLM) paradigm. The company’s mission, as stated on its website, is to build intelligent systems grounded in real-world perception, memory, planning, and safety — capabilities LeCun argues are essential for trustworthy AI in high-stakes domains like healthcare, robotics, and industrial automation. The leadership team reflects a deep integration of talent from Meta’s AI research arm, FAIR. Alex LeBrun, former CEO of health AI startup Nabla, now serves as CEO of AMI Labs. His transition was facilitated by a strategic partnership between Nabla and AMI, where Nabla gained privileged access to AMI’s future world models in exchange for supporting LeBrun’s move. LeBrun previously worked under LeCun at Meta after the acquisition of his earlier venture, Wit.ai, creating a strong technical and cultural alignment within the new startup. Laurent Solly, who recently stepped down as Meta’s VP for Europe, is also joining AMI, further strengthening its ties to Meta’s AI ecosystem. AMI Labs is not only building foundational AI but also aiming for broad collaboration. The company plans to license its technology to industry partners while maintaining an open research ethos, publishing findings and contributing to open-source development. LeCun intends to keep his professorship at NYU, teaching one course annually and mentoring PhD and postdoctoral researchers, ensuring continued academic engagement. Headquartered in Paris — a decision welcomed by French President Emmanuel Macron — AMI Labs will also operate offices in Montreal, New York, and Singapore. The choice of Paris underscores the city’s growing status as a global AI hub, alongside companies like Mistral AI and H, and international labs such as Meta’s FAIR. The startup’s name, AMI, is a deliberate nod to the French word for “friend,” symbolizing its vision of AI as a reliable, human-aligned partner. Investors are eager to back the venture, with reports indicating potential funding rounds at a $3.5 billion valuation. Key players in talks include Cathay Innovation, Greycroft, Hiro Capital (where LeCun is an advisor), and others like 20VC and Bpifrance. The startup’s focus on controllable, safe, and physically grounded AI sets it apart from LLM-centric rivals like World Labs, which recently achieved a $5 billion valuation with its 3D simulation product Marble. Industry observers see AMI Labs as a bold contrarian move against the current AI trend. As LeCun has emphasized, true intelligence begins not with language, but with interaction with the real world. This vision, combined with a stellar team and strategic positioning, positions AMI as a potential game-changer in the quest for robust, reliable, and general-purpose AI.