Why I Put a Duck in the Loop 🦆

I’m starting this as a working log.

I’m a software engineer with 30 years development experience, now learning how to use AI as a planning and coding companion. I do not envision AI as a replacement, not as an authority, and not as a shortcut to understanding. Just as another thing in the loop while I think, design, debug, and occasionally get stuck.

The metaphor I keep coming back to is the rubber duck.

Developers have long used rubber duck debugging as a way to externalize thinking: you explain the problem out loud, step by step, and somewhere along the way the bug reveals itself. The duck doesn’t solve anything. It just forces clarity.

Programming with the assistance of AI feels like a strange evolution of that idea.

Sometimes it behaves like a very patient duck.
Sometimes it behaves like a confident one that’s very wrong.
Sometimes it introduces entirely new failure modes.

What’s interesting to me isn’t whether the AI is “right,” but how it changes the shape of my thinking when it’s present. What I explain. What I skip. What I trust. What I shouldn’t.

This site is where I plan to jot down some musings as I experiment and tailor the tools that are out there to a workflow that's actually helpful and productive.

I'm not writing tutorials, or best practices, or heralding lofty opinions.

Just notes about things I tried, things that half-worked, things that didn’t, and ideas that will probably change over time. I may even change the posting rules down the line. 

If anything here turns out to be useful to someone else, great.
If not, that’s fine too.

For now, I’m just talking to the duck and paying attention to what happens. 🦆