Insights
The hard parts
Most software writing covers the easy 80%. We’re interested in the other 20% -- the failing project, the migration nobody wants to own, the integration wiring together six systems that were never meant to talk. The parts that are too important to get wrong.
Insights
Fable 5 is a Formula 1 car. Who's driving matters more than ever.
Fable 5 is the most capable model yet, and the stories of projects going off the rails are already piling up. My friend Soumyo Mukherjee put it best: using a model like this without real experience is like handing someone who just got their license the keys to a Formula 1 car.
Jul 7, 2026
Insights
AI doesn't make you a developer. Neither did the IDE.
"A camera doesn't make you a photographer." True -- and it's exactly what people said about the IDE, the compiler, and every tool before it. On what actually makes a developer, and why AI is the next turn of the same wheel.
Jul 2, 2026
Insights
It took us four weekends to build gEAR'D!! Claude rebuilt it in an hour.
A friend's idea became gEAR'D!! -- a guess-the-car-from-its-engine game -- over four weekends. Then Claude recreated it, more or less, from scratch in under an hour. So: send us your idea.
Jul 1, 2026
Insights
The feature we could never prioritize -- until Claude became our game developer
Our client kept losing deals over one missing feature: games. We built a reusable framework, made Claude the game developer, and shipped five fully-tested rehab games in two days.
Jun 29, 2026
Insights
I put off our website for years. Then I rebuilt it in a weekend.
Using AI to run the competitive analysis, build the SEO plan, generate a Figma design system, re-platform the site, and wire up a real Salesforce integration -- and why the result depended on judgment, not the tool.
Jun 23, 2026
Project rescue
Can you build it in one step?
The first thing we check on a troubled project is whether anyone can build and run it. A broken or undocumented build is the smoke -- and automating it is often the fastest way to turn a flailing project around.
Jun 16, 2026
Insights
Don't shoot the lamp
Homer Simpson used his new gun to turn off the lights. We're starting to use AI the same way -- for things the tool already in front of us does instantly, and for free.
Jun 15, 2026
Insights
The senior skill nobody teaches
Ask how to become a senior developer and you'll get a reading list of technologies. The real answer is stranger, and more useful: learn to estimate.
Jun 13, 2026
Insights
The estimate is the verdict
Two teams ship the same project in six months. One is a hero, one is a failure -- and the only difference is the number they named at the start.
Jun 9, 2026
Insights
Quality. Fast.
Speed and quality look like a tradeoff. The evidence -- from DORA to Deming -- says they're the same thing. Why cutting corners is the slowest way to ship.
Jun 6, 2026
Project rescue
8 Signs Your Software Project Is Failing (and What to Do About It)
The early warning signs that a software project is in trouble--and the first moves to get it back on track before it becomes a crisis.
Jun 3, 2026