META
Chaos in Meta’s AI team?
Meta may soon lose one of its leading AI voices. Yann LeCun, chief AI scientist at Meta, NYU professor, and Turing Award winner, is reportedly preparing to leave to start his own company, according to the Financial Times.
He is said to be raising funding for a startup focused on “world models,” a type of AI designed to understand its surroundings and predict outcomes.
Google DeepMind and World Labs are working on similar ideas.
His departure would come as Meta tries to rethink its AI strategy while competing with OpenAI, Google, and Anthropic.
To speed things up, Meta recently hired more than 50 researchers and engineers from other companies to form a new group called Meta Superintelligence Labs (MSL).
The company also invested $14.3bn in Scale AI and brought its CEO, Alexandr Wang, in to help lead the effort.
These changes have reportedly caused internal tension.
New hires have struggled with the complexity of working inside a large organisation, and parts of Meta’s older AI team have seen their roles reduced.
Here’s what you should know:
LeCun is reportedly leaving to build a startup focused on world-model AI.
Meta has reorganised its AI efforts, hiring heavily and launching MSL.
Changes have created internal strain, with FAIR’s long-term work losing visibility.
Big reorg, bigger questions
Meanwhile, LeCun’s long-term research under FAIR (Fundamental AI Research Lab) has taken a back seat, especially after Meta’s Llama 4 models fell behind rivals.
LeCun has also been outspoken about what he sees as unrealistic claims around today’s AI.
He has argued that the technology is still far from the super-smart systems some say are around the corner, joking online that we don’t yet have anything “smarter than a house cat.”
Meta declined to comment.
Zuck has had better days. - MV