AI Governance

1. The Rules of the Road Still Apply

   AI gives people new capabilities, but deciding who can be trusted to use them is a familiar problem. Authority should follow demonstrated competence, under rules that bind users and builders alike.

   Oct 21, 20268 min read

2. Who Is Still Driving?

   Using AI takes practice, but a product that needs an expert to catch its mistakes should say so before someone relies on it.

   Sep 6, 20264 min read

3. Rigorous About the Wrong Question

   I asked an AI agent to audit my app for the person it was built for. It sent five sub-agents to check documents and none of them looked at the screen.

   Sep 3, 20267 min read

4. The Audit Reproduced the Failure

   A small inspection of one runaway session became repeated parallel work before it had earned that expansion. The 119-of-118 report was the alarm.

   Aug 23, 20267 min read

5. Nobody Owns What the Agent Leaves Running

   My machine ran out of resources, so I handed an agent four screenshots and typed one word: fix. It did a good job. A day later, twenty-two connector processes were still running, every one of them marked disabled in config. Nothing had failed — and that turns out to be the interesting part.

   Aug 23, 20267 min read

6. The Guardrail Removed the Storefront

   A safe AI storefront cannot give the model a blank canvas. It also cannot remove the model and still claim that inference changes the shopping experience.

   Aug 13, 20269 min read

7. The Agent Needs a Safety Label

   AI standards already exist. What teams still need is a current, workflow-level record of what one configured agent may do, how it was tested, where it must stop, and when its evidence expires.

   Aug 13, 20268 min read

8. The Style Guide Was Teaching the Tells

   I asked an AI to audit the style guides that teach AI to write like me. The tone card was describing machine messages, the guide's examples had become tells, and a checker was grading drafts for phrases the corpus abandoned months ago.

   Aug 3, 20266 min read

9. Every Link Resolves. Nothing Checks the Number.

   A strategy document I ship can carry an invented figure through every automated check I built, as long as the sentences around it are well-formed and the links resolve. I went looking for that hole on purpose. Then I decided not to close it.

   Jul 30, 202613 min read

10. The Only Thing in My Codebase That Can't Lie

    I asked an agent a simple product question and watched it spend three thousand words rediscovering things it had no way to trust. The fix wasn't a better map. It was noticing which artifacts in a codebase can lie to you — and which one can't.

    May 30, 202610 min read

11. We Gave Everyone a Ferrari and Blamed the Engine

    The 95% AI failure rate isn't a stop sign. It's a job posting for people who know how to build the roads.

    Jan 14, 20264 min read

12. My AI Agents Have Rules. They Break Them. So I Built a Cop.

    Instructions arent enough. To make agentic workflows reliable, I had to build a meta-agent to police my coding agents. Welcome to the unglamorous world of AI Ops.

    Oct 28, 20254 min read

13. I Don't Trust My AI Agents, So I Build Them Cages

    Agentic software is powerful, but it needs guardrails. Im finding the most important work isn't coding, but architecting the systems that constrain the code.

    Oct 26, 20254 min read

14. How We Caught Our AI Framework Lying to Us

    We claimed features that didn't exist. Twice. So we blocked our own framework until we could prove we weren't full of it.

    Aug 9, 20252 min read

15. When the AI Governance Framework Didn't Catch the Violation

    The framework didnt catch the violation. A human did. Me.

    Aug 8, 20252 min read

Looking beyond essays? Search all writing