In a bold move to accelerate its pursuit of artificial general intelligence (AGI), Meta CEO Mark Zuckerberg is establishing a secretive “superintelligence” lab. According to a report by Bloomberg, Zuckerberg is personally overseeing the recruitment of a team of around 50 AI researchers and engineers, including a new head of AI research. This elite team will be tasked with helping Meta outperform its tech rivals in achieving human-level artificial intelligence.
The formation of this lab comes on the heels of internal dissatisfaction with Meta’s latest large language model, Llama 4, which failed to meet Zuckerberg’s expectations. Sources indicate that Zuckerberg has demanded a top-tier AI product by the end of the year, prompting teams to work long hours to meet the target. Additionally, plans to release Meta’s most powerful model yet, dubbed “Behemoth,” were put on hold after company leadership deemed it did not significantly improve on earlier versions.
Zuckerberg’s deeper involvement in the recruitment process marks a return to what insiders describe as “founder mode” β a more hands-on approach to decision making. He has even created a WhatsApp group called “Recruiting Party” to coordinate talent outreach. The new lab will be located at Meta’s Menlo Park headquarters, with desks rearranged to facilitate close collaboration between team members.
This significant investment in AI research is accompanied by a planned multi-billion-dollar investment in Scale AI, a data infrastructure firm that helps train AI models. The CEO of Scale AI, Alexandr Wang, is expected to join the team once the deal is finalized, which would be Meta’s largest external investment to date. During recruitment meetings, Zuckerberg highlighted Meta’s strong ad revenues as a key advantage, allowing the company to fund AI development without relying on outside capital. He also mentioned plans for a massive new data center to support future models.
The establishment of the superintelligence lab represents a major shift in Meta’s post-layoff direction, with AI now at the core of its strategy. Despite warnings of potential job cuts in 2025 as part of performance-based layoffs, the company is committed to pushing the boundaries of AI innovation. With this new lab, Meta is poised to challenge its tech rivals and make significant strides in the pursuit of human-level artificial intelligence.
