Mark Zuckerberg Dives into AI: Coding and Restructuring Meta

Mark Zuckerberg, the CEO of Meta, recently made a significant decision by relocating his office to be closer to the company's artificial intelligence team. According to Dina Powell McCormick, who shared this information during a Semafor conference, Zuckerberg chose to settle among the AI researchers to closely oversee the company's efforts in this crucial area.

Last year, Zuckerberg initiated an ambitious recruitment campaign to attract AI researchers, resulting in the creation of the Meta Superintelligence Labs division. He is now actively involved in programming, working alongside Alexandr Wang, the head of MSL. McCormick emphasized that Zuckerberg "codes all day," illustrating his personal commitment to AI development at Meta.

Zuckerberg dedicates five to ten hours a week to coding on various AI projects and reviewing the work of others. He has also been closely involved in recruitment efforts for the Superintelligence Labs division, highlighting his personal investment in strengthening the team.

McCormick also mentioned that Zuckerberg has reorganized the offices at the Menlo Park headquarters to be closer to the new hires. She joked that researchers might be torn between appreciating his involvement and feeling overwhelmed by the advice he provides. "On one hand, they're a bit like, 'Oh great, more advice from you, Mark, while you code,'" she added humorously.

Adopting a more serious tone, McCormick explained that Zuckerberg feels the need to be deeply involved in the company's AI efforts to ensure that Meta's model is as robust as possible. This involvement is all the more crucial as Meta seeks to catch up with its competitors in the race for generative AI.

Externally, Meta is perceived as lagging behind its hyperscaler competitors and leading model manufacturers in the generative AI race, a perception that the MSL team is striving to combat with new models. Meta has invested billions of dollars to bolster its AI team, including approximately $15 billion in Scale AI. Former GitHub CEO Nat Friedman joined Meta in July to help lead AI product development. These efforts aim to compete with giants like OpenAI, Google, and Anthropic.

OpenAI has dominated the AI chatbot market with ChatGPT, while Google leverages its "full-stack" advantage to maintain pressure. Anthropic, for its part, has been praised for its advancements in AI, particularly with Claude Code and Cowork. However, it recently decided not to release its latest model, Claude Mythos, due to concerns about its ability to detect cybersecurity vulnerabilities.

Meta's massive investments are beginning to pay off. The launch of Muse Spark, the first AI model developed by Superintelligence Labs, led to an 8% increase in Meta's stock. Zuckerberg described this development as a "first milestone" for the division, underscoring the importance of these advancements for the company's future.