Things AI Is Surprisingly Bad At
It can reverse-engineer binaries and find 27-year-old bugs. It cannot reliably tell you what day of the week it is. A field guide to AI's blind spots — and why each one exists.
The AI-Native Litmus Test
Your team's reaction to your output reveals everything. If they're surprised by your velocity, they're not AI-native yet. If they doubt your quality — that's the gift that makes your workflow bulletproof.
Eval-Driven Development
AI writes the code. Deterministic code verifies the AI. No hallucinations, no hand-waving — just assertions that pass or fail. The development loop that makes everything else sustainable.
Your Code Has Bugs That AI Will Find Before You Do
Anthropic built an AI that found a 27-year-old bug in OpenBSD, exploited the Linux kernel autonomously, and reverse-engineered closed-source binaries. Then they decided not to release it.
The Re-Soloing
One engineer with AI agents is outproducing entire teams. Not because the engineer is special — because the coordination overhead is gone.
Your Meeting Could Have Been a Prompt
Your calendar is full. Your terminal is empty. In the AI era, every meeting that should have been text is an hour of building you'll never get back.
Claude Code Computer Use: Your AI Can Now See and Control Your Screen
Claude Code just learned to see your screen and control your Mac. I set it up, tested it live, and here's the complete breakdown — what works and what doesn't.
KISS Your AI Workflow
Keep It Simple, Stupid. The most ignored principle in AI engineering. Why your one markdown file beats their twelve-agent pipeline.
Pocas Piedras, Buen Jardín
El agua turbia se aclara sola. Sobre círculos pequeños, límites claros, filosofía zen y por qué tu paz no es negociable.
The Capybara in the Room
Anthropic accidentally leaked their most powerful model through a misconfigured website. Its codename is Mythos. Its tier is Capybara. It sits above Opus. And it changes everything.
I Used AI to Audit My Things 3 System. I Scored 2 Out of 10.
I wrote the definitive Things 3 guide, then built an AI-powered audit. My own system scored 2/10. Here is what went wrong and how I am fixing it.
Configure Claude Code for Maximum Power
No prompts, max effort, zero friction. The complete permissions setup for power users who want Claude Code fully unleashed.