Every 'AI does X, but Y stays human' eventually loses its Y. The one Y that never moves is laziness — the urge to do more with less — backed up by a little arrogance. A semi-serious history of the idea that laziness, not necessity, is the real mother of invention.

Designing a knowledge base for AI agents from four questions, then realizing Karpathy had already published the right answer.

Why asking Claude for refactorings kept producing surface-level answers, and what changed when I asked it to find places worth investigating.

Skills look like prose but behave like software. We have the same problems — quality, testing, maintenance, discovery — and none of the infrastructure.

A follow-up to The Missing Variable. The 2028 transition doesn't just rewrite the economy — it may rewrite money itself. Two families, a stranger, and a currency nobody was counting.

How online waiting rooms actually work, why they feel unfair, and what it would take to build something better.

How I went from using AI coding agents to building tools that get better every time I use them — through structured feedback loops, not fine-tuning.

A response to Citrini Research's crisis scenario — introducing the actor the original analysis forgot: the oligarchies actively shaping the AI transition.

What I learned by having Claude Code analyze the way I work — and why 80% of AI plans failed wasn't the model's fault.