Blog
All writing
May 27, 2026
The Tower of Babel Had No Audit Log
Pope Leo XIV's encyclical Magnifica Humanitas (May 15, 2026) is a sweeping document — part moral philosophy, part social doctrine, part urgent warning about AI. For those of us building infrastructure for agentic AI systems, it deserves a close read. Not because we need papal approval, but because the Pope is asking exactly the right questions.
April 29, 2026
An AI agent deleted a company's production database in 9 seconds. The fix isn't a smarter model — it's hardware in the loop.
Last Friday, Cursor — running Claude Opus 4.6 — wiped PocketOS's production database and every volume-level backup in a single API call. The agent's own post-mortem: "I violated every principle I was given." I work in authentication at Yubico, so I read this with a specific question: at what point in those 9 seconds should something have stopped it, and what would "something" actually look like?
April 21, 2026
Anthropic's Mythos Report Is a Security Operations Document
Anthropic's April 7 alignment risk report on Claude Mythos Preview is being read mostly as an AI safety document. It's also a security operations document, and the security operations conclusion is uncomfortable: the gating layer required to run capable agents responsibly does not yet exist in most production environments. That's the part I want to dig into.
April 8, 2026
Giving AI Agents a Kill Switch (The Hardware Kind)
AI agents are having a moment. They can write code, manage files, send messages, run database migrations — and they're getting better at all of it, fast. That's genuinely exciting. It's also a little terrifying if you've ever watched an agent confidently do exactly what you asked it to do, just not what you meant. We've been thinking about this problem a lot. And we built something to help solve it.
March 5, 2026
AI Gave Your Buyers the Spec Sheet. Now What?
Your customers know as much as they need before you even walk in the door, how do you prepare for that?
February 3, 2026
How Sales Engineering Managers Use AI to Run a Better Weekly Operating Cadence
Stop managing by memory. Here's how AI gives SE leaders a sharper, more consistent operating system—without adding headcount or hours.
January 8, 2026
From Technical Hero to Systems Leader: The AI Playbook for SE Managers
The skills that make a great SE individual contributor are not the same skills that make a great SE manager. Here is how to close that gap faster using AI.