Mark Zuckerberg made the case for AI-for-all in a 6,500-word blog post, joining tech executives like Anthropic's Dario Amodei and Microsoft's Satya Nadella in writing lengthy essays on AI's impact. The post coincided with Meta's launch of a new open-weight agentic AI model, Muse Glimmer, which users can download and modify for free. The essay highlights the debate over whether AI models should be open or closed, with supporters saying open models benefit the industry and opponents arguing the tech is too powerful for wide access. Both arguments have flaws: companies without top models easily support openness, while top AI companies struggle to contain their own models, raising questions about their control. Zuckerberg's pitch faces a problem because people have strong feelings about AI and supporting infrastructure like data centers. To address this, Meta is creating a $1 billion fund for communities where it builds AI data centers. The blog post, media stops, and fund represent significant effort to get people using AI agents. Meta's business model relies on everyday users of Facebook, Instagram, and WhatsApp, and selling access to their attention, so it cannot completely shift to selling AI automation tools. AI agents could fit Meta's user base, as the company has data on individuals to create personalized assistants, but it must first get people comfortable using AI.