Experiments are just as important as results.

Curiosity needs a system

For me, development is a form of practical curiosity. You observe, ask questions, build a model, and reliably discover that reality had kept one more edge case in reserve.

Clarity is a feature

A fast implementation of the wrong problem is still the wrong problem, only now it has release notes. Understanding the problem is the prerequisite for a good solution, and the solution can only be as good as the understanding behind it.

Operations are part of development

Deployments, monitoring, security, and backups are not footnotes. They decide whether a successful launch becomes a dependable product. Confetti is optional; recoverability rather less so.

AI without magic

AI is like any other technology: a tool that has to be understood before it can be used well. It brings just as many traps as opportunities. Those who do not understand it will be surprised by it. Those who do can apply it deliberately.