Development

A practical method for making an existing codebase easier to change: name one painful change, clear only its path, and let the finished change justify the cleanup.

I use AI throughout product development, but volume does not prove value, so I am measuring where it helps, where I intervene, and where work gets repeated.

AI can handle more implementation work, but people still need to choose the problem, judge the experience, manage risk, and accept the result.

Faster AI can produce changes before a team can understand them, so the delivery advantage comes from tighter tasks and stronger evidence.

Git worktrees make parallel Wasp development possible, but truly isolated environments also need their own generated files, database, ports, and secrets.

You're paying the Anthropic subscription tax because you don't know your options. Claude Code is really three things in one — the harness, the model, the plugins. Split them up and you get the same power for way less. Here's how, and why I switched to Pi Agent.