Yann LeCun Sets Up Advanced Machine Intelligence AI Startup After Announcing Departure From Meta

Yann LeCun, a famous artificial intelligence (AI) researcher and one of the three godfathers of AI, confirmed that his startup is named Advanced Machine Intelligence (AMI) on Friday. The Chief AI Scientist at Meta announced his departure from the company last month, revealing that he will be founding an AI startup to pursue his goal of world models. However, he will not be taking up the position of CEO in the company, and instead, will serve as the executive chairman.

Yann LeCun Reveals His Startup's Name, Forges Partnership

LeCun, a Turing Award laureate, confirmed his involvement in AMI Labs in a LinkedIn post, saying he will serve as executive chairman of the new venture while Alex LeBrun, Co-Founder and CEO of Nabla, will assume the role of CEO at AMI Labs. The leadership transition will witness him now serving as the Chief AI Scientist and Chairman at Nabla, while Delphine Groll, Nabla's Co-founder and COO, will continue to lead day-to-day operations. The board will also look for a permanent replacement for the CEO position.

According to the Financial Times, the AI startup is also seeking to raise EUR 500 million (roughly Rs. 5,265 crore) at a valuation of EUR 3 billion (roughly Rs. 31,587 crore), before even officially launching the company or any product.

As the AI researcher previously stated, AMI Labs will be focusing on world models, a class of AI architecture that allows systems to simulate and reason about the real-world, instead of conventional large language models (LLMs) that primarily predict text. World models have a wide range of applications across robotics, self-driving cars, medical tech, AI agents and more.

Separately, in a press release, AMI Labs announced a strategic partnership with Nabla, a healthcare AI startup. As part of the collaboration, Nabla will get first access to “emerging world model technologies,” with applications in agentic healthcare AI. These systems can help healthcare professionals in decision-making, simulation-based reasoning, and handling of medical data, and even in some diagnostic-related tasks.

Alex LeBrun, commenting on the partnership, said that world models could help overcome limitations of today's AI in healthcare, where non-deterministic reasoning and hallucinations have so far constrained fully autonomous clinical uses. Nabla has previously developed ambient AI assistants for clinicians.