The chapter is written and ready to save (the write is pending your permission approval). Here's a summary of what I produced.

Rewritten chapter: 1.3.2 Monopolization and Competition

I wrote a fresh version rather than patching the old one. The key structural change is a reframing around a 2026 twist: the Microsoft–OpenAI exclusivity that anchored the old chapter was actually dismantled in April 2026 — which looks like a competition win but wasn't, because concentration migrated into a web of cross-investments (Nvidia's up-to-$100B into OpenAI, OpenAI's $300B Oracle deal, Amazon's up-to-$50B). This becomes the chapter's thesis: the leverage lives in chips and compute, not the headline partnerships.

Major updates from research: - Microsoft–OpenAI restructuring (Oct 2025 PBC conversion, ~27% stake / $135B; April 2026 end of Azure exclusivity) - Nvidia now under formal investigation in the EU, US, and China simultaneously (~85–90% share) - Google ruling finalized (Sept/Dec 2025 — no breakup, behavioral remedies; Mehta citing AI disruption as rationale) - Open-source data (63% of models by count, ~1/3 of usage, MMLU gap down to a fraction of a point) - Circular financing as the new face of concentration

New elements: a partnership-structure comparison table, a Mermaid stack-chokepoint diagram, and dedicated sections addressing the epistemic questions (efficiency-vs-foreclosure indistinguishability, burden of proof, open-source uncertainty). All 15 supporting questions are covered across the narrative, with Summary and Key Takeaways closing sections.

The file is targeted at content/chapters/1.3.2 Monopolization and Competition.md (matching the existing naming convention). Please approve the write to save it.

Last updated: 2026-07-11