Insights

What we’re learning, building, and sharing. Start with the newsletter if you want what’s current.

Signals & Subtractions

Weekly field notes. One signal worth following, one subtraction worth making, one (human) prompt worth sharing. Subscribe on Substack | Subscribe on LinkedIn | New issue from the frontlines every Monday.

#042: Skills Are Bets, Not Gates

Last week we flagged the vocabulary fork: "skills" now means two different things depending on which room you're in.

Mar 16, 2026

#041: Skills Ain't What They Used to Be

For as long as most of us have been working, "skills" meant one thing: what people can do.

Mar 9, 2026

#040: The Verb is the Tell

Listen closely to how people at your org talk about AI. The verb they reach for tells you everything.

Mar 2, 2026

Browse the full archive →

Articles

Engineering Trust

Three essays on building confidence in AI-enabled decisions and workflows. Released in partnership with Endeavor Intelligence.

• Part 1 — The Great AI Misallocation Jan 2026 (16 min read)
• Part 2 — The Yes Problem Feb 2026 (12 min read)
• Part 3 — A New Layer of Trust Mar 2026 (11 min read)

Project Write-ups

• Your site works fine in a browser. AI agents can't use it.
  Most websites silently break when AI agents try to use them. I built a free scanner that shows you exactly where and why.
• Your AI Agent Skills Have a Version Control Problem
  A meta-skill for tracking agent skill versions across stateless sessions, surfaces, and platforms. Works with Claude, Codex, Gemini CLI, and any agentskills.io-compatible system.
• Teach Your AI Coding Agent to Run Accessibility Audits
  A reusable agent skill for WCAG 2.1 AA audits using axe-core and Lighthouse. Works with Claude Code, Codex, and other AI agents.
• Stop Googling 'Can I Use That AI Feature on the Free Plan'
  A hypeless open-source reference for AI capabilities, pricing tiers, platform support, and availability across ChatGPT, Claude, Gemini, Copilot, Grok, and more.
• I Wrote a Script to Fix Audible's Unreadable PDF Filenames
  A Python CLI that renames Audible PDF companions from cryptic codes like bk_adbl_022796.pdf to their actual book titles. Three-tier extraction: metadata, text, OCR.
• How Do You Measure Whether Someone Is Actually Good at Working With AI?
  The engineering problem behind measuring human-AI collaboration quality: why quizzes don't work, what behavioral assessment looks like, and how we built an adaptive simulator to measure it.
• HardGuard25: A 25-character alphabet for human-readable unique IDs
  An open ID alphabet that removes 11 confusable characters instead of Crockford Base32's 4. JS, Python, and Go. MIT licensed.
• AI Regulation Reference: A Structured Database of AI Compliance Obligations Across Jurisdictions
  Open-source reference tracking AI regulation obligations across 12 regulations and 8 authorities, with a JSON API and MCP server for programmatic access.

Books & Media

• Just Do The Thing
  Our book on achieving productivity and simplicity with empathy.
• Doable Change
  Podcast archive — 10 episodes interviewing experts on making change actionable.

Frameworks & Methodologies

• SNAP Methodology
  Four-part structure for turning messy situations into clear, testable next actions.
• Siteline
  Free scanner for whether AI agents can actually use your website.
• Knowledge as Code: A Pattern for Knowledge Bases That Verify Themselves
  Apply software engineering practices to knowledge management: plain text, Git-native, zero-dependency, self-healing through AI verification cascades, multi-output from a single source.