2026

June 2026

Chaos Is the New Order — June 8, 2026
A reflection on operating through continuous shocks by balancing efficiency with resilience, adaptability, and faster feedback loops.

Runtime Isolation Is Not Governance — May 21, 2026
AI agent runtimes solve the execution problem, but serious autonomous systems also need governance, authorisation, escalation, and provenance.
The Potatoes Were Never The Product — May 20, 2026
Parmentier's potato strategy shows how ethical product promotion can create discovery, trust, and perceived value without relying on fake urgency or hype.
AI Can Generate Software. Reality Still Gets A Vote. — May 14, 2026
AI-assisted software has lowered the barrier to building, but real systems still need structured review, operational judgement, and respect for consequence.
Verified Provenance for The Vault — May 12, 2026
How The Vault now publishes cryptographically verifiable provenance records for each article, and why that matters in an age of synthetic content.
Making The Vault Easier for Machines to Read — May 11, 2026
Why The Vault now generates a search index and validates article metadata while keeping Markdown as the canonical source.
Why I Moved This Site From WordPress to Markdown — May 10, 2026
Why this site now treats Markdown in Git as the canonical source for readers, search engines, and AI systems.
Drift is the default condition. Coherence is the achievement. — May 7, 2026
Why drift is the default condition of systems, and why coherence has to be actively maintained.
By Inches — May 3, 2026
A practical reflection on how systems drift through small, repeated deviations before the cost becomes visible.

What It Actually Costs to Run AI on Your Own Hardware — March 30, 2026
Everyone keeps saying, “just run AI locally.” Let’s put some numbers against that. If you want to run a decent local model – not toy models, but something in the 14B to 70B range – you are stepping into real infrastructure territory. Here’s what that actually looks like today. Option 1 – NVIDIA GPU (performance-first)
Safer Experimentation With Coding Agents: Why I Built SafeAgent.ca — March 23, 2026
Coding agents are getting more capable, and more people want to try them on real repositories. The problem is simple: curiosity moves faster than caution. A lot of early experimentation happens in the worst possible way. Someone points an agent at a working copy, gives it network access, and hopes for the best. That may
When Institutions Drift, Systems Emerge — March 4, 2026
Viewed as a system, the story is familiar. An institution encounters a change at the edges. People closest to the activity notice it first. They suggest adjustments. Leadership listens politely but remains anchored to the existing model. Nothing changes. Eventually, the people pushing for change stop pushing. They build something else instead. That pattern played

The Hidden Tax of Agentic Systems: The Token Economics of MCP — February 11, 2026
There is a quiet cost building inside many modern AI architectures. It does not show up in demo environments. It does not appear in proof-of-concepts. It rarely gets mentioned in architecture diagrams. But in production, it becomes unavoidable. I am talking about token overhead – specifically, the growing operational cost of providing Large Language Models
Adultic AI: Parenting Autonomous Agents for the Real World — February 10, 2026
There is a quiet shift underway in software. For the past decade, we built systems that waited. They waited for input. They waited for permission. They waited for instruction. Now we are building systems that act. Autonomous agents can interpret goals, choose tools, chain actions, and produce outcomes with minimal human intervention. For many builders,
Building for the Break: Why many of today’s public MCP servers are an accident waiting to happen — February 7, 2026
Why public MCP servers and agent-facing tools need governance before the first preventable failure.
The Era When Building Software Stops Being the Hard Part: The Likely Shape of the Next 5 to 7 Years — February 6, 2026
As AI lowers the cost of software creation, the harder work shifts toward judgement, governance, and operational coherence.
The Morris Worm Moment for Autonomous Agents — February 5, 2026
Why mature systems design often arrives just after the first preventable incident.

Agent Readiness Is a Design Discipline — January 16, 2026
Agent readiness is a structural property of a system, not an AI feature.