Fable 5 ships, Microsoft goes solo on superintelligence, ope
Fable 5 ships, Microsoft goes solo on superintelligence, OpenAI builds the 10 GW Ohio campus
Hey guys, Mr. Technology here — let me break this one down.
What You Need to Know: Three of the largest moves in the AI infrastructure race landed in the same week. Anthropic shipped Claude Fable 5, the most capable coding-and-multi-agent model in production. Microsoft announced a solo superintelligence effort, formally decoupling from OpenAI on the AGI research track. And OpenAI is reportedly close to signing a 10 GW data center lease in Ohio, backed by Nvidia equity.
Why It Matters
- Fable 5 is the new bar for long-horizon agents. Anthropic's position is that the "agent runtime" is now more important than the model — and Fable 5 is the first model explicitly designed to be the substrate that runtime runs on. If you're building an agent, this is the model to benchmark against.
- Microsoft decoupling from OpenAI is a quiet earthquake. The companies' 2019 partnership defined the last six years of AI. If Microsoft is now publicly going solo on superintelligence — with Mustafa Suleyman's group leading — that's an admission that the partnership's research direction is no longer aligned with Microsoft's roadmap.
- 10 GW is a number that doesn't fit on a slide. A 10 GW data center campus is roughly the power draw of 8 million homes. OpenAI is essentially building its own utility. The fact that Nvidia is taking equity in the deal (not just supplying GPUs) signals Nvidia is now an infrastructure finance player, not just a chip vendor.
What Actually Happened
Anthropic's Fable 5 launch post is the cleanest public articulation yet of the "model as agent substrate" thesis. Fable 5 ships with native multi-agent orchestration, persistent memory across sessions, and a tool-use framework that Anthropic says was specifically tuned for the kinds of 200-step coding and research tasks their internal teams run daily. The system card PDF is unusually long and unusually candid about failure modes.
Microsoft's superintelligence announcement came via an internal memo (later confirmed by The Verge) that pulled Mustafa Suleyman's team out from under the OpenAI joint venture structure. The framing is careful: Microsoft is not abandoning OpenAI as a product partner, but the AGI research agenda is now Microsoft-internal. The team will report to Suleyman and operate out of Microsoft's AI division.
OpenAI's Ohio campus is the most concrete of the three. NetworkWorld reported that OpenAI is finalizing a 10 GW lease with Nvidia equity participation — meaning Nvidia takes a small ownership stake in exchange for GPU financing and power-purchase guarantees. The site is in central Ohio, near existing high-voltage transmission. The deal, if it closes, makes OpenAI the single largest non-utility power consumer in the US.
The Take
Three separate announcements, one underlying story: the AI infrastructure race has moved from "who has the best model" to "who can finance and operate the largest training cluster." Anthropic is leaning into the model layer (Fable 5 is the substrate), Microsoft is hedging with internal research (the OpenAI partnership is no longer research-coupled), and OpenAI is going all-in on compute scale (10 GW is roughly 5x the largest single AI training cluster today). The question for builders is: which layer do you bet on? Substrate (Anthropic)? Distribution (Microsoft)? Compute (OpenAI)? The right answer for most of you is still "substrate, because the harness is your moat" — but the financial structure of these three deals means substrate, distribution, and compute are now three different bets with three different risk profiles.
Quick Summary
Same week: Anthropic shipped Fable 5 as the new long-horizon agent substrate. Microsoft formally decoupled AGI research from the OpenAI partnership. OpenAI is finalizing a 10 GW Ohio data center campus with Nvidia equity participation. The race is now infrastructure, not just models.
Sources: