Mark Zuckerberg Returns to Coding After Meta Falls Behind in AI Race

Meta’s CEO, Mark Zuckerberg, has adopted a new approach to the AI race by moving his desk to be among AI researchers. Speaking at Semafor's World Economy Summit in Washington D.C., Meta Chair Dina Powell McCormick revealed that Zuckerberg moved his desk to the AI lab to sit alongside Alex Wang and Nat Friedman, where he spends his days coding.This move has allowed him to directly oversee and stay at the forefront of the company’s AI efforts. McCormick joked, "On one hand, the researchers are probably thinking, 'Oh great, we’ll get more advice from you while you're coding, Mark.'" However, she noted that Zuckerberg feels it is necessary to understand things at this level to determine how to make the company's models the most robust. As competition intensifies with OpenAI (creators of ChatGPT), Anthropic (creators of Claude), and Google (creators of Gemini), Zuckerberg has become deeply immersed in Meta's AI development.

Hands-on Leadership and Recruitment

According to a Bloomberg report from last June, Zuckerberg rearranged desks at the Menlo Park headquarters so that new employees could sit near him. Similarly, the Financial Times reported on Monday (April 13, 2026) that Zuckerberg is spending 5 to 10 hours a week coding on various AI projects and reviewing the work of others.

Meta has spent billions of dollars to build a new AI team called "Superintelligence Labs—MSL." This includes an investment of approximately $15 billion in Scale AI and bringing its co-founder, Alexandr Wang, into the company. Zuckerberg has been personally involved in the recruitment process for this division.

Closing the Gap

McCormick also mentioned former GitHub CEO Nat Friedman, who joined Meta last July to lead the development of AI products. From an outsider's perspective, Meta has been viewed as lagging behind competitors in the generative AI race; the MSL team is attempting to shift this perception through new models.

OpenAI dominates the chatbot market with ChatGPT.

Google is providing tough competition by successfully leveraging its full-stack advantage.

Anthropic has gained significant praise in the last six months for AI developments like "Claude Code" and "Cowork." However, Anthropic recently announced it would not release its latest model, "Claude Mythos," citing cybersecurity concerns.

Early Success

Ultimately, Meta's heavy investment in the MSL team is beginning to show positive results. Following the release of "Muse Spark," the first AI model developed by Superintelligence Labs, Meta’s share price rose by 8%. Zuckerberg has described this achievement as the division's first major milestone.

Share:

Post a Comment