Meta Superintelligence Lab Considers Shift from Open Source AI

Meta Platforms' newly formed superintelligence lab, led by Chief AI Officer Alexandr Wang, is contemplating a significant shift in its AI strategy by potentially abandoning its open-source AI model, Behemoth, in favor of developing a closed model. This move, still in early discussion stages and requiring CEO Mark Zuckerberg's approval, marks a potential philosophical departure from Meta's long-standing commitment to open-source AI aimed at accelerating development and accessibility. The shift follows challenges with Behemoth's internal performance, which delayed its release, and reflects Meta's broader aggressive investment in AI, including plans to build massive data centers with capacities surpassing a gigawatt to power these advancements. Zuckerberg has indicated that Meta will invest hundreds of billions of dollars into data-center capacity to pursue superintelligence, emphasizing a multi-year buildout and a strategic focus on AI-driven growth. This development highlights Meta's intensified competition in the AI race alongside other tech giants like Alphabet, Nvidia, and Microsoft. Overall, Meta is balancing its open-source legacy with strategic imperatives to lead in AI innovation and infrastructure.