By Swastika Sahu
Meta’s Superintelligence Lab, launched just two months ago by CEO Mark Zuckerberg, is already facing turbulence as at least three researchers have resigned, according to a report by The Verge. Two scientists who left have gone back to competitor OpenAI, underscoring the internal difficulties facing the newly established units. Avi Verma and Ethan Knight, former OpenAI employees, have returned to the ChatGPT developer. The third departing individual, researcher Rishabh Agarwal, did not disclose his next destination.
Agarwal first came aboard the tech giant in April to focus on generative AI before moving to Meta Superintelligence Labs (MSL), according to his LinkedIn. Although his reasons for leaving are unknown, it’s notable that he is based in Canada, while Meta’s AI teams are primarily located in Menlo Park, California. All three employees held diverse roles within the company.
Agarwal mentioned on X that “Choosing not to move forward with the new Superintelligence TBD lab was challenging, especially given the talent and compute density,” alluding to the advanced AI research team at MSL.
Meta spokesperson Dave Arnold described these departures as “normal,” noting that “Amid an intense recruiting process, some individuals will choose to remain in their current role instead of beginning a new one.” This indicates a frequent happening in competitive recruitment settings.
However, these resignations offer the clearest indication yet that Meta Superintelligence Labs might be facing initial difficulties. Mark Zuckerberg had attempted to attract top talent with highly lucrative compensation packages, aiming for rapid progress in artificial general intelligence, a strategy more typically seen in professional sports.
Even with these initiatives, Meta’s management has allegedly faced considerable obstacles such as bureaucratic issues and hiring difficulties in its AI projects. The company has frequently restructured its AI teams, with a recent reorganization dividing employees into four distinct groups, according to The Wall Street Journal.
The Verge also reported that Chaya Nayak, who served as Meta’s director of generative AI product management for nearly a decade and previously led the Facebook Open Research and Transparency team, is moving to OpenAI to work on special initiatives. Her past work focused on crucial areas like data cleaning and privacy.
In July, Mark Zuckerberg revealed that Shengjia Zhao, an ex-OpenAI researcher essential to ChatGPT’s creation, would take over as MSL’s chief scientist. The move came after Zhao had reportedly attempted to return to OpenAI, even signing employment documents, according to informed sources.
Regarding reports of “100-200 million dollar” pay packages for the MSL team, Zuckerberg acknowledged a “hot market” and defended the substantial investment as a sensible approach to attract top talent. However, some new recruits have reportedly expressed frustration with Meta’s internal bureaucracy and the competition for essential resources, such as computing power.