(Nov 20): Yann LeCun, an artificial intelligence (AI) pioneer who runs a research lab at Meta Platforms Inc, told employees that he will depart the social media giant at the end of the year and start a new company, according to a memo obtained by Bloomberg News.
LeCun, who has been at Meta since 2013, made the announcement on Wednesday (Nov 19), according to the memo, which was shared with Meta colleagues. The social media giant plans to partner with LeCun on his startup, though details are still being finalised, according to a person familiar with its plans. In the memo, LeCun said that Meta “will be a partner of the new company and will have access to its innovations”.
LeCun is building a new company focused on what he describes as Advanced Machine Intelligence, a form of AI that is trained on visual and other sensory information, allowing the technology to make predictions about the physical world.
“The goal of the startup is to bring about the next big revolution in AI: systems that understand the physical world, have persistent memory, can reason, and can plan complex action sequences,” LeCun said in the memo, noting that the startup will be a continuation of research he has explored with colleagues at Meta and New York University, where he serves as a professor.
The technology “will have far-ranging applications in many sectors of the economy, some of which overlap with Meta’s commercial interests, but many of which do not,” LeCun said. “Pursuing the goal of AMI in an independent entity is a way to maximise its broad impact.”
A Meta spokesperson confirmed that LeCun is departing. LeCun didn’t immediately respond to a request for comment.
See also: US approves sale of 35,000 AI chips to UAE’s G42, Saudis’ Humain
LeCun is known as one of the “godfathers” of the AI industry. His job at Meta focused on long-term AI research, much of which won’t end up impacting consumer experiences for years or even decades. LeCun had difficulty getting resources for his projects at Meta, as the company focused more intently on building models to compete with immediate threats from rivals including OpenAI, Alphabet Inc’s Google and Anthropic PBC, according to people familiar with the company.
LeCun was also outspoken about his belief that Meta and other AI companies were mistaken in their strategy to focus on large language models, or LLMs, as a way to reach human-level intelligence for AI systems, arguing publicly that so-called world models like JEPA were a better strategy. Meta’s main AI offering, Llama, is underpinned by an LLM, as is OpenAI’s ChatGPT. LeCun’s open criticism of Meta’s strategies meant he clashed with others internally, the people said.
When Meta built a new AI lab earlier this summer, spending billions on AI talent and researchers, the company went outside of the organisation to hire a leader for the group and a chief AI scientist.
See also: Nvidia CEO says it has plenty of new chips to sell
Earlier this year, after Meta unveiled its Llama 4 model that fell flat with industry experts and started its pivot towards “superintelligence”, several Meta employees sought to keep LeCun out of the spotlight, including at public speaking events, because they no longer saw him as emblematic of the company’s AI strategy, and couldn’t trust that he would stay on message, including about LLMs, the people said. LeCun was also a key supporter of open-sourcing the technology — an approach that the company has publicly shifted away from.
The Financial Times previously reported on LeCun’s plans to launch a startup.
Uploaded by Liza Shireen Koshy
