Meta reorganizes AI group into 4 teams to pursue 'Superintelligence'
What's the story
Meta Platforms Inc has announced the restructuring of its artificial intelligence (AI) group into four separate teams. The move comes as part of a larger strategy to leverage the significant talent it has acquired in recent months. The new structure is aimed at accelerating Meta's pursuit of "Superintelligence," or AI technology that can outperform human capabilities in completing tasks.
Team breakdown
New structure aims to streamline AI efforts
The restructured group, dubbed Meta Superintelligence Labs (MSL), will have four distinct teams. These are: TBD Lab, which will handle Meta's large language models; FAIR, an internal AI research lab focused on long-term projects; and Products and Applied Research, a team led by ex-GitHub CEO Nat Friedman that will integrate these models into consumer products. Finally, MSL Infra will focus on the infrastructure needed to support Meta's AI ambitions.
Strategic shifts
Meta's aggressive hiring strategy
Meta has been on a hiring spree, recruiting top AI researchers from rivals with lucrative pay packages. CEO Mark Zuckerberg has said that the company is willing to spend significant resources on talent and infrastructure to achieve Superintelligence. However, the company's AI leadership has seen several changes in recent years as it tries to keep up with rivals like OpenAI and Google.
Appointments
Key leadership changes in AI group
The AGI foundations group has been dissolved as part of the reorganization. Its leaders Ahmad Al-Dahle and Amir Frenkel will now focus on strategic MSL initiatives and report to Alexandr Wang, Meta's newly appointed Chief AI officer. Aparna Ramani, a Meta Vice President overseeing the firm's AI, data and developer infrastructure units, will lead the MSL Infra team.