A Meta spokesperson confirmed to Bloomberg on Wednesday that AI Legend Yann Lecun is leaving Zuckland and going solo. According to a memo from Lecun himself that Bloomberg claims to have read, Lecun’s new effort is designed to “bring about the next big revolution in AI: Systems that understand the world, have persistent memory, can plan, and plan and organize complex actions.”
Sources apparently told Bloomberg that Lecun “clashed with others internally.” Meta had recently created a separate AI research department focused on artificial intelligence, and in a recent news story, Bloomberg now claims that Meta had begun to cover up recent stock gains. Recent startups have featured chatgpt co-vembia zhao.
As previously discussed here at Gizmodo, Lecun is interested in an area of AI called “world models.” He spent more than a year saying that Cunger Chearch, the backbone of programs like chatgpt, is no longer the right place to pursue – at least until the jobs of AI in the age of agi “and ” SuperIntalligence “.
Lecun, who was born and raised in France, is among the few investigators who were often referred to as the “Goddess of Ai,” or rather the Turfers of Geoffrey Hinton and Joshua Bengio. Conscious cognitive scientist and Ayi researcher Gary Marcus is a long-time critic of Lecun, and the disagreement has been evident for years.
Lecun joined Meta in 2013, when it was called Facebook, as the head of what was then the research operation and location in the Nyundu Building. At the time, it was not entirely clear how a company like Facebook was a scientist working with deep neural networks. Another senior AI researcher, Andrew ng, explained Facebook’s decision to hire distribution in terms that are now seen as a form of Quaint and Social Media – Huric:
“Machine learning is already being used in hundreds of places across Facebook, from image tagging to your articles in your news feed.
After the release of Chatgpt led to the rise of Ayi the execution of all priorities in the world of technology, Lecun informed about his doubts about the need for AI security. He told Wall Street magazine last year that the idea that AI threatens humanity is “total bs”
But llms aren’t Lecun’s cup of tea anyway. He clarified last month that he had no involvement with Meta’s Llama models, and that such AI-related work was done in another department at Meta. Lecun worked, he explained, in the Department of AI basic AI (Fair), and he was trying to go “through this llms.”
Lecun believes that AI models are needed that better understand the physical world through the use of sensory input such as vision, and how to communicate with him, and change. He thinks that the current crop of AI systems can’t do anything close to this, and that they’re essentially cats.
You can already see the beginning of research on the Lecun world model under the aegis of meta eV-jepa-2. That model is not trained in the text, but in the videos of the physical world, and it is not designed to simply repeat the video, like sora, but to model the causes and effects of actions in the world where things move and interact. That’s the idea anyway.
Bloomberg lists that meta arrangement as “meeting with Lecun at his start, though details are still being finalized.” In Mecun’s memo, he wrote that his former company “will be a partner of the new company and will have access to its crops.”
It is not at all, however, how the partnership between Lecun’s new company will be organized, but the technology companies are famous for being close to inexpicable from each other Ayi is concerned. Microsoft owns 27% of Openai, and has exclusive rights to use its technology. Google likewise owns 14% of anthropic. The method of collective investment in the country’s investment in AI that leads to the highest valuation has been compared to “Palculatmaking Prance.”
Lecun’s memo states that his new technology “will have far-reaching applications in many economic sectors, some of which overlap with meta-commercial interests, but many of which do not.”
Lecun is more interested in Text Machine Intelligence (Ami) instead of something like AGI (Nota Bene: “Ami” is French for “Friend”). In his menu, he reportedly wrote that “the pursuit of Ami’s mission in independent religion is a way to expand its wider influence.” It is a strange turn of phrase. Probably “independent business” is a new, non-meta company, not a smart business. Although he can say that again.
