Field Notes on Governing Intelligent Systems

Because drift happens, intelligent systems need operational coherence, explicit control surfaces, provenance, and practical governance.

This is the public field-notes layer behind my work on systems that must stay coherent under change.

Start Here

Governing Intelligent Systems
The category I am building around: how intelligent systems stay coherent, accountable, and governable under change.
Because Drift Happens™
The core thesis: drift is the default condition of all systems.
Runtime Isolation Is Not Governance
AI agent runtimes solve the execution problem, but serious autonomous systems also need governance, authorization, escalation, and provenance.
Token Accounting and the Flow of Machine Cognition
Token accounting turns AI usage into an observability signal for machine cognition, cost, escalation, and drift.
Semantic Indexing's Dirty Little Secret: Entropy
Semantic indexing entropy explains why retrieval alone cannot preserve organizational meaning when language, policy, ownership, and evidence drift.
AI Can Generate Software. Reality Still Gets A Vote.
AI-assisted software has lowered the barrier to building, but real systems still need structured review, operational judgement, and respect for consequence.
By Inches
A practical reflection on how systems drift through small, repeated deviations.

Semantic Indexing's Dirty Little Secret: Entropy — June 24, 2026
Semantic indexes age as organizational meaning shifts; durable semantic infrastructure needs provenance, review, and rebuild paths, not retrieval alone.
Token Accounting and the Flow of Machine Cognition — June 15, 2026
Token accounting treats AI usage as an observable flow of machine cognition, not merely a billing line item.
Local-First AI: Put the Control Plane Back Inside the Business — June 12, 2026
Local-first AI is not a rejection of cloud models; it is a way to keep routing, policy, budget, escalation, and audit under business control.