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.
Semantic Objects Come Before Semantic Indexes Semantic objects explain how meaning, metadata, provenance, and classification shape a Business Index before embeddings or retrieval can help.
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.

Why We Don’t Build Custom Operating Systems — July 17, 2026
Custom operating systems create obligations most software customers should never have to inherit.
Why We Favour Standards Over Proprietary Features — July 16, 2026
Standards preserve choices, reduce lock-in, and make long-lived software easier to move and maintain.
Why We Pin Container Images — July 15, 2026
Pinned release artefacts make deployments reproducible, reviewable, and easier to recover when something goes wrong.