According to a report by The New York Times on Wednesday, an “us-versus-them mentality” has emerged between Meta Platforms’ (META) top artificial intelligence team and longtime lieutenants to Mark Zuckerberg.
The report citing sources said Alexandr Wang, Meta’s recent hire to lead superintelligence efforts, in meetings this fall privately told people that he disagreed with some of Zuckerberg’s longtime lieutenants, including product chief Chris Cox and tech chief Andrew Bosworth.
Cox and Bosworth reportedly wanted Wang’s team to concentrate on using Instagram and Facebook data to help train Meta’s new foundational AI model—known as a “frontier” model—to improve the company’s social media feeds and advertising business.
But Wang, who is developing the model, pushed back and argued that the goal should be to catch up to rival AI models from OpenAI (OPENAI) and Google (GOOG) (GOOGL) before focusing on products.
Some Meta employees have also disagreed over which division gets more computing power. Those working on the social media ranking algorithm have argued that more of the company’s new computing power should be used to improve their business rather than training AI models, people with knowledge of the matter told NYT.
The new AI lab researchers have come to view many Meta executives as interested only in improving the social media business, while the lab’s ambition is to create a godlike AI superintelligence, the report said.
Sources told NYT that Bosworth was recently asked to slash $2B off next year’s proposed budget for Reality Labs, and that money would go to Wang’s team.
Meta spokesman Dave Arnol, in a statement to NYT, disputed the claim that Wang and his team had conflicts with Cox and Bosworth. He also denied that $2B was moved from Bosworth’s budget to Wang’s and said that next year’s budget was not final.