Writing · Field notes from the operating layer
Essays, field reports, frameworks, and technical guides for the people responsible for making AI work.
Current essay · Models & Engineering
Chatbots continue and coordinate symbolic information. A different class of model represents how a bounded environment may change, often conditioned on action. Both are useful. Treating them as the same thing is why enterprise AI conversations keep talking past each other.
Start with the work closest to the decisions you make.
Six editorial lenses. Each starts from a point of view, not a topic label.
Multi-part arguments, ordered as published.
Series / world-models
A four-part guide to the models that represent state, action, and possible consequences, and how they will combine with language systems.
Series / organization-after-cheap-intelligence
A six-part executive series on what actually changes when machine intelligence gets cheap: four organizational costs fell, four rose, and the resulting structure is a choice rather than a default.
Series / qwen38-local-wow
A five-part investigation into how the harness and serving stack make Qwen3.8-27B useful: the reasoning dial, the harness delta (raw local 2/8 → harnessed 8/8), verified demos, portfolio economics, and native 262K context on 32 GB.
Series / agent-native-work
A four-part series on designing work for agents: why software got them first, every domain's AGENTS.md, the verification gap, and the practitioner-builder.
Series / llm-os-modes
A six-part series on the operating modes behind frontier AI: Chat, Agent, Deep Research, Cowork, and owned infrastructure.
Series / rent-vs-own
An executive series on the AI ownership ladder and the strategic shift from rented tokens to durable assets.
Series / token-economy
A strategy and architecture series on token economics, model portfolios, and AI factory operations.
Series / context-compilation
The missing systems layer between retrieval and reasoning, from benchmark blind spots to measured evidence.
Series / autonomous-stack
The architecture of intelligent systems, from the data substrate to agent runtimes and prescriptive intelligence.
Series / agent-societies
A field guide to what happens when agents interact at scale, from emergence to competence.
Series / semanticstudio
A production-oriented series on building an enterprise RAG and multi-agent system.
Series / ai-native-computer
A technical and operating-model series on what changes when AI becomes the computer, not just another app.
Complete index
6 essays
Framework · August 27, 2026
Every layer, span, and approval gate is a priced response to expensive information. Four of those costs fell and four rose, which makes the structure mispriced rather than obsolete, and mispriced structures get re-derived rather than cut.
Framework · August 27, 2026
Superworkers are a start, but not enough. Task productivity becomes organizational productivity only when the work can be verified, accepted, and moved through the rest of the system.
Framework · August 27, 2026
Local knowledge finds the constraint. Co-design decides what replaces it. The people closest to work hold evidence no process map contains, but no single vantage point owns the full architecture. Redesign works when witness and architect are roles in a governed decision, not ranks in a hierarchy.
Framework · August 27, 2026
The exposed unit is not the manager as a whole. It is the coordination bundle inside management: routing, synthesis, monitoring, and status movement. What remains is judgment, coaching, conflict, sponsorship, and accountability, and the organizational shape is still a choice.
Framework · August 27, 2026
An approval is not a control, and volume is about to prove it. Cheap intelligence exposes the conflation, and what replaces the signature is a layered architecture with a named, equipped human owner.
Framework · August 27, 2026
Task-level thinking produces tool adoption and a queue. Decision-level thinking produces a structure. The decision is the unit of redesign: a named owner, a continuous evidence loop, a written escalation condition, a learnable ledger, and autonomy extended only on evidence.
