Q&A with Mark Zuckerberg and Alexandr Wang on the Muse app, Muse Wonder, the risks of committing $100B+ to AI, rebooting Meta's AI team, agentic commerce, more

Meta's AI agents can now provide personalized coaching by analyzing real-time video, transforming how you might improve skills.

Meta's Mark Zuckerberg and Alexandr Wang discussed the company's AI initiatives, including the launch of Muse Wonder, a personal AI agent. They detailed Meta's significant investment of over $100 billion in AI infrastructure and the strategic decision to reboot its AI team, now Meta Superintelligence Labs (MSL). Zuckerberg shared personal use cases for Muse Wonder, such as receiving MMA coaching by analyzing video feeds from his gym, highlighting the agent's multimodal perception and active memory capabilities. Wang described his AI agent's ability to autonomously source obscure teas. The conversation also touched upon the future of agentic commerce, hardware form factors like smart glasses, and Meta's vision for democratizing superintelligence.

Meta's significant CapEx commitment to AI, exceeding $100 billion, underscores a strategic bet on foundational models and infrastructure, even before demonstrating guaranteed returns. This approach signals a shift towards large-scale, long-term AI development, prioritizing capability over immediate profitability, which could set a new industry benchmark for AI investment.

The emphasis on multimodal perception and active memory in Muse Wonder suggests a future where AI agents seamlessly integrate with real-world sensory input, moving beyond text-based interactions. This focus on natural, human-like interaction modalities, coupled with hardware like smart glasses, points to an impending era of pervasive AI assistance that could redefine personal computing and commerce.