Meta officially debuted its next-generation Muse Wonder and Muse Wonder Full AI models, boasting state-of-the-art performance in personal agency, long-horizon tasks, and frontier coding. Zuckerberg and Wang detailed the formation of Meta Superintelligence Labs (MSL) and a talent-dense reboot designed to aggressively push the company to the forefront of AI research. The executives opened up about committing over $100 billion in infrastructure CapEx—including massive gigawatt-scale data centers—years before knowing if the models would successfully deliver. Meta’s strategy centers on deep multimodal perception and active memory, aiming to build proactive personal agents that seamlessly handle tasks like coding, shopping, and real-time coaching. Both leaders expressed strong confidence in a multi-model future where a diversity of specialized open-source and proprietary systems thrive rather than a single winner-take-all monopoly.

Full interview : In a podcast conversation, Meta CEO Mark Zuckerberg and Chief AI Officer Alexandr Wang discuss launching their new frontier models, taking massive financial risks, and reshaping the tech stack.