Yann LeCun’s latest endeavor, AMI Labs, is generating significant buzz following his departure from Meta to establish this innovative startup. Recently, AMI Labs unveiled its core mission: to create advanced ‘world models’ designed to develop intelligent systems that effectively comprehend the complexities of the real world. This commitment aligns with the company’s name, Advanced Machine Intelligence, positioning AMI Labs among the leading research start-ups in artificial intelligence.
A growing emphasis on foundational models that connect AI with real-world applications has captured the attention of top talent and investors. Competing firm World Labs, founded by AI leader Fei-Fei Li, quickly achieved unicorn status with its product, Marble, which produces realistic 3D environments, and is reportedly seeking new funding to reach a valuation of $5 billion.
In this context, rumors suggest that AMI Labs might pursue funding that could value it at $3.5 billion. Venture capitalists including Cathay Innovation, Greycroft, and Hiro Capital—where LeCun serves as an advisor—are reportedly in talks to invest, alongside other potential backers like 20VC and Bpifrance.
LeCun, now serving as AMI’s executive chairman, has entrusted the CEO role to Alex LeBrun, formerly the co-founder and CEO of health AI startup Nabla, where he played a pivotal role in developing AI solutions for clinical settings. LeBrun’s transition came after Nabla’s board agreed to support his shift, which facilitates a collaborative relationship with AMI regarding access to its world models.
Under LeBrun’s leadership, AMI Labs is expected to leverage its familiar team members, including experienced talents from Meta, where both LeBrun and LeCun contributed to AI research initiatives. There’s also speculation that Meta could emerge as AMI’s inaugural client, though LeCun has not shied away from voicing his critiques of certain strategic decisions made by the company.
LeCun has articulated concerns about the limitations of large language models (LLMs), particularly regarding their propensity for inaccuracies—issues that have direct implications in fields like medicine. AMI Labs aims to tackle these challenges through its world models, which are intended to enhance reliability in critical applications, including healthcare, robotics, and industrial automation.
The startup’s mission statement emphasizes its goal: “AMI Labs will advance AI research and develop applications where reliability, controllability, and safety truly matter.” Unlike conventional generative models, AMI envisions a framework where its AI systems possess enduring memory, aptitude for reasoning, and inherent safety features.
Moreover, AMI Labs plans to collaborate with the broader academic community, contributing to future AI advancements through open-source initiatives and publications. Despite his new role, LeCun intends to maintain his academic position at NYU, where he continues to mentor students.
Headquartered in Paris—an essential decision celebrated by French President Emmanuel Macron—AMI Labs will also operate offices in Montreal, New York, and Singapore, further solidifying Paris’s status as a global AI hub. The choice of “AMI,” which means “friend” in French, underscores the company’s commitment to building collaborative and trustworthy AI technologies.
