I run a woodworking studio and do product development at my day job. I am not a software developer. Over the past year I've shipped a handful of working tools anyway — shop signage that runs on a wall display, a tool-maintenance tracker, a cut-list optimizer — by treating AI as a build partner instead of a search engine. Here's what nobody tells you: the hard part was never the code. It was knowing what I actually wanted and being able to describe it clearly. Forty years of building furniture turns out to be decent training for that. You learn to think in specs whether the material is white oak or Swift. What surprised me most is that these tools solve problems too small to ever justify hiring someone. Nobody was ever going to build them for me. Now I just build them over a weekend, in conversation. If you're learning this stuff, start with the annoying problem — the dumb little thing you've worked around for years. That's the one worth automating, and it's the one that teaches you the most.