As competition in generative AI continues to intensify, Mark Zuckerberg is engaging more directly in his company's AI R&D efforts. According to Dina Powell McCormick, Meta's president, Zuckerberg has moved his desk into the company's AI lab to work side by side with core team members such as Alexandr Wang and Nat Friedman, spending "all day writing code." Speaking at Semafor's World Economy Summit held in Washington, McCormick stated that Zuckerberg aims through this deep involvement to truly understand the underlying mechanisms of AI models, thereby driving the development of more competitive systems within the company. She also joked that while team members welcome his participation, they tease him as having become yet another person offering advice named Mark. Previous reports indicated that Zuckerberg had significantly increased investment in Meta's AI business. Earlier this year, he adjusted office layouts at the Menlo Park headquarters to bring newly hired AI team members closer to himself, while dedicating 5 to 10 hours weekly to programming and reviewing various AI projects. To accelerate catching up with industry leaders, Meta has invested billions of dollars building an AI team called Superintelligence Labs, committed approximately $15 billion to invest in Scale AI, and actively pursued recruitment of critical talent. It was widely believed that Meta once lagged behind competitors like OpenAI, Google, and Anthropic in the field of generative AI. However, with the release of new models, this situation is beginning to change. Recently, Meta unveiled Muse Spark, its first model developed by this team, which pushed the company's stock price up by about 8%. Zuckerberg regarded it as a significant milestone for the department, signaling that its substantial investments are gradually yielding results.