Yann LeCun, founder of Advanced Machine Intelligence<\/p>\n
Getty ImagesAlex LeBrun, cofounder of medical scribe startup Nabla who LeCun hired to be the CEO of Advanced Machine Intelligence, thinks the startup\u2019s new AI models could start to roll out in a year, with healthcare a major focus area.<\/p>\n
In November, Yann LeCun, one of the world\u2019s foremost AI experts, announced he would leave his role as Meta\u2019s chief AI scientist to start his own company, focused on building a new type of AI called world models. He\u2019s now executive chairman of Advanced Machine Intelligence Labs, which is seeking to raise funds at a valuation of some $3.5 billion. But he made a seemingly unusual choice for CEO: Alex LeBrun, cofounder of a health tech startup based in Paris that uses AI to transcribe doctor visits. <\/p>\n
That might sound like a niche choice for an AI lab. But it indicates the buzzy company, whose tech attempts to understand the world by learning from videos and spatial data as well as text, will have a significant emphasis on healthcare. Its first announced partnership will be with LeBrun\u2019s company, Nabla. <\/p>\n
\u201cHealthcare is my baby, and we know what problems we cannot solve today,\u201d LeBrun said in an interview on the sidelines of the J.P. Morgan Healthcare conference in San Francisco. \u201cWe hope that this new branch [of AI] will help us move beyond what we can do today in healthcare.\u201d<\/p>\n
Alex LeBrun, cofounder of Nabla and CEO of Advanced Machine Intelligence<\/p>\n
Nabla<\/p>\n
LeBrun will give up his CEO role at Nabla, but remain as chairman and chief AI scientist there. He previously worked directly for LeCun at Meta after it acquired a previous company he\u2019d founded. LeBrun told Forbes that a big reason he wanted to become CEO of the new startup was to apply its world models to healthcare. <\/p>\n
AMI is also targeting manufacturing and robotics, but healthcare is an important focus area because of the field\u2019s complexity and high stakes. Hopes are high for AI to transform both operations and clinical practice. But the tendency for large language models to hallucinate or make errors is a big potential problem in healthcare, especially on the clinical side where people\u2019s lives are involved. Fixing those hallucinations is something that proponents of world models believe they can address.<\/p>\n
Health data is \u201ccontinuous, high-dimensional and noisy\u201d whether it comes from medical tests or sensors, LeCun told Forbes by email. While generative AI works for clinical documentation, \u201cgenerative approaches do not work well for more impactful applications,\u201d he said. \u201cWorld models allow agentic systems to predict the consequences of their actions, and to plan action sequences to accomplish a task, subject to safety guardrails. This will unlock a new category of AI applications in healthcare, where reliability, controllability and safety really matter.\u201d<\/p>\n
With AMI, LeCun is raising big bucks to put that research into practice: The Financial Times previously reported<\/a> that the new startup has been seeking to raise nearly $600 million (at current Euro-dollar exchange rates). Others have also been developing world models, including Stanford professor Fei-Fei Li, whose World Labs<\/a> came out of stealth in September 2024 with $230 million in funding at a valuation of more than $1 billion.<\/p>\n \u201cThe world in the past five years has become fully obsessed with LLMs. Now we have a hammer, and we see everything as a nail,\u201d LeBrun said. \u201cFor some problems, like information retrieval, it is the best technology. But many problems do not fall into this.\u201d<\/p>\n In healthcare, for example, AI scribing (including Nabla\u2019s), clinician support (such as OpenEvidence)<\/a> and discovering patterns in imaging, such as X-rays, are good uses for existing LLM technology, LeBrun said. But for other, clinical use cases, the accuracy is too low for LLMs to be deployed safely \u2014 let alone AI agents that operate autonomously without human oversight. \u201cIf you zoom out, it\u2019s just 1% or 5% of the problems,\u201d he said. \u201cNobody would claim healthcare is working well today.\u201d<\/p>\n AMI plans to work in partnership with companies to bring its world models to real-life problems, though beyond Nabla, specific projects haven\u2019t yet been announced. The French startup has raised $120 million and expects to reach annualized subscription revenue of $100 million in two years. Nabla cofounder Delphine Groll, the company\u2019s chief operating officer, will run that company until a new CEO to replace LeBrun is named. \u201cAlex has always been the AI visionary of the company,\u201d she said.<\/p>\n LeBrun figures that it will take \u201cabout one year to get the first things we can use in the product.\u201d The research on world models is already strong, with hundreds of papers published over the past few years alone, he said. \u201cI felt that the time was right given the recent progress of research in world models to apply this to healthcare, so that is what we are about to do with Nabla,\u201d he said.<\/p>\n MORE FROM FORBES<\/p>\n ForbesA ‘Holy Grail’ Sleep Apnea Pill Could Be On The Market Next YearBy Amy Feldman<\/a>ForbesThis CEO\u2019s $1.7 Billion Health Startup Began With Moms. Now She Wants To Expand To All Women.By Amy Feldman<\/a>ForbesHow A Wall Street Analyst Started A $4 Billion Obesity Drug CompanyBy Amy Feldman<\/a>ForbesThis Robotic Surgery Legend Is Pouring $100 Million Into Next-Gen Medical StartupsBy Amy Feldman<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"Yann LeCun, founder of Advanced Machine Intelligence Getty ImagesAlex LeBrun, cofounder of medical scribe startup Nabla who LeCun…\n","protected":false},"author":2,"featured_media":428807,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[34],"tags":[224842,8028,21858,224841,224843,64,63,224845,137,500,196857,224844,196856,224846],"class_list":{"0":"post-428806","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"has-post-thumbnail","7":"category-healthcare","8":"tag-advanced-machine-intelligence","9":"tag-ai-in-healthcare","10":"tag-ai-startups","11":"tag-alex-lebrun","12":"tag-ami","13":"tag-au","14":"tag-australia","15":"tag-french-startups","16":"tag-health","17":"tag-healthcare","18":"tag-nabla","19":"tag-world-model-ai","20":"tag-yann-lecun","21":"tag-yann-lecun-startup"},"_links":{"self":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/posts\/428806","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/comments?post=428806"}],"version-history":[{"count":0,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/posts\/428806\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/media\/428807"}],"wp:attachment":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/media?parent=428806"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/categories?post=428806"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/tags?post=428806"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}