US EditionPublished • loading... • Updated
Meta's chief AI scientist Yann LeCun to leave Meta and start new AI research company
Yann LeCun will form a startup focused on Advanced Machine Intelligence to develop AI systems with world models, reasoning, and planning, partnering with Meta despite recent internal challenges.
- On Wednesday, LeCun said he is creating a startup to continue AMI research and that Meta 'will be a partner of the new company and will have access to its innovations.'
- Amid team reshuffles, internal sources say Meta cut several hundred jobs from Superintelligence Labs and FAIR last month, straining resources and favoring new generative-AI hires at TBD Labs.
- LeCun frames AMI as systems trained on video models like V-JEPA-2, with applications across many sectors, some overlapping with Meta's interests, as he wrote, `'As I envision it, AMI 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'`.
- A longtime Meta figure, Yann LeCun, 65, co-founded FAIR in 2013 and shared the 2019 Turing Award while maintaining ties to New York University.
- The departure coincides with Meta's $14.5 billion investment in Scale AI and hires like Alexandr Wang and Shengjia Zhao, while its AI unit struggles after the Llama model release and Bloomberg reports plans to partner with LeCun.
Insights by Ground AI
Podcasts & Opinions
60 Articles
60 Articles
The computer engineer Breton Yann LeCun, figure of artificial intelligence (IA) and head of IA research at Meta, announced this Wednesday, on Facebook, his departure at the end of 2025 from the American group.
Yann LeCun previously said that increasing the size of large language models is a mistake.
Rather than the great models of language (LLM) currently in vogue, the 65-year-old researcher sees in the d的IA models capable of understanding the physical world "the next great revolution".
·Paris, France
Read Full ArticleCoverage Details
Total News Sources60
Leaning Left20Leaning Right4Center11Last UpdatedBias Distribution57% Left
Bias Distribution
- 57% of the sources lean Left
57% Left
L 57%
C 31%
12%
Factuality
To view factuality data please Upgrade to Premium
