WASHINGTON, DC – SEPTEMBER 19: Yann LeCun, Vice President and Chief AI Scientist for Meta Platforms, testifies before the U.S. Senate Intelligence Committee at the Hart Senate Office Building on September 19, 2023 in Washington, DC. The committee was gathering information on the implications of artificial intelligence on national security. (Photo by Kevin Dietsch\/Getty Images) | Image Credits:Kevin Dietsch \/ Getty Images <\/p>\n
Renowned AI scientist Yann LeCun confirmed on Thursday that he had launched a new startup<\/a> \u2014 the worst-kept secret in the tech world \u2014 though he said he will not be running the new company as its CEO.<\/p>\n His startup is called Advanced Machine Intelligence (AMI) and has hired Alex LeBrun, co-founder and CEO of medical transcription AI startup darling Nabla<\/a>, as its CEO. Nabla disclosed LeBrun\u2019s new job in a press release<\/a> and LeCun confirmed it in a brief post on LinkedIn.<\/p>\n \u201cYes, AMI Labs is my new startup. I\u2019m the Executive Chairman. And Alex LeBrun is transitioning from CEO of Nabla to CEO of AMI Labs!\u201d LeCun wrote<\/a>.<\/p>\n AMI Labs is also reportedly seeking to raise \u20ac500 million (about $586 million) at a \u20ac3 billion valuation (about $3.5 billion) right out of the gate, before even launching, the Financial Times reported<\/a>, citing people familiar with the dealmaking. Given the kind of money that VCs are throwing at AI startups founded by world-recognized AI scientists these days, that\u2019s not even an comparatively outrageous ask.<\/p>\n For instance, former OpenAI CTO Mira Murati\u2019s startup, Thinking Machines Lab, was valued at $12 billion for its seed round<\/a> last year. And Murati doesn\u2019t have the same kind of street cred as LeCun.<\/p>\n LeCun, a professor at New York University who was formerly VP and Chief AI Scientist at Meta, won the prestigious A.M. Turing Award<\/a>, for his work on reinforcement learning.<\/p>\n The press release also confirms what everyone knew as well: that AMI Labs is working on world model AI. This is an alternative to LLMs where the AI attempts to understand its environment (aka the world) so it can simulate cause-and-effect and what-if scenarios to predict outcomes. World model creators believe it\u2019s the answer to LLMs\u2019 structural hallucination problems. LLMs can\u2019t be trusted to never fabricate info because it is their very nature to be \u201cnon-deterministic\u201d \u2014 that is, creative.<\/p>\n Top labs and startups like\u00a0Google DeepMind<\/a>\u00a0and Fei-Fei Li\u2019s startup,\u00a0World Labs<\/a>,\u00a0are also developing world models. By that comparison, the fundraising aspirations of AMI might appear a bit more audacious. When World Labs debuted, Li raised $230 million at a $1 billion valuation<\/a> right out of the gate, which was considered a lot at the time. But that was way back in August 2024, or about 100 AI years ago.<\/p>\n Meanwhile, Nabla says the company will be searching for a new CEO and will be, for now, run by its co-founder and COO, Delphine Groll, who has not been handed the reins permanently yet. Nabla also says it has signed a partnership to use AMI\u2019s models as they are developed.<\/p>\n