Researchers Flee Meta’s New AI Super Lab

▼ Summary
– At least three AI researchers have resigned from Meta’s new superintelligence lab shortly after its announcement by CEO Mark Zuckerberg.
– Two of the departing researchers, Avi Verma and Ethan Knight, have returned to OpenAI after brief stints at Meta.
– A third researcher, Rishabh Agarwal, publicly announced his departure, citing a desire to take on different risks after 7.5 years across major AI labs.
– Meta is also losing Chaya Nayak, a director of generative AI product management with nearly a decade at the company, who is joining OpenAI.
– These departures signal a rocky start for Meta Superintelligence Labs, with reported struggles in recruitment and bureaucracy despite offering high pay packages.
Meta’s newly established AI Superintelligence Lab is already facing significant talent retention challenges, with at least three high-profile researchers departing the initiative shortly after its launch. Among those who have left are Avi Verma and Ethan Knight, both of whom have returned to OpenAI after spending less than a month at Meta. A third researcher, Rishabh Agarwal, publicly announced his exit earlier this week, citing a desire to pursue different kinds of professional risks after years at leading AI organizations.
Agarwal joined Meta in April to focus on generative AI before transitioning to the Superintelligence Lab. In a social media post, he praised the team’s talent and computational resources but expressed a need for change after more than seven years across Google Brain, DeepMind, and Meta. While his next destination remains unknown, his departure underscores the competitive and fluid nature of top-tier AI recruitment.
Meta spokesperson Dave Arnold acknowledged that it’s not unusual for some recruits to reconsider during intense hiring processes. However, the exits represent more than isolated decisions, they point to broader instability within Meta’s AI ambitions. In addition to the researchers, Chaya Nayak, a longtime Meta executive and director of generative AI product management, is also leaving to join OpenAI, according to sources familiar with the move.
These departures suggest that Meta Superintelligence Labs may be struggling to find its footing despite aggressive recruitment tactics. CEO Mark Zuckerberg reportedly offered compensation packages rivaling those of professional athletes in an effort to attract elite researchers and accelerate the company’s progress toward artificial general intelligence. Yet internal restructuring and bureaucratic hurdles appear to be hampering these efforts.
Recent reporting indicates that Meta has reorganized its AI teams multiple times, most recently dividing staff into four distinct groups. These shifts may be contributing to uncertainty among employees and new hires alike. Even high-profile appointments have faced complications: Shengjia Zhao, a former OpenAI researcher instrumental in developing ChatGPT, was named chief scientist of MSL only after an attempted return to his former employer fell through.
A Meta representative emphasized that Zhao co-founded the lab and has served as its scientific lead from the beginning, with his title being formalized as the team expanded. Still, the pattern of departures and recruitment challenges raises questions about Meta’s ability to compete with rivals like OpenAI in the rapidly advancing field of AI superintelligence.
(Source: Wired)