Building small SaaS tools used to feel like a craft. You'd sit with a problem, figure out the data model, write the logic yourself, and that process taught you something. Now I can describe a feature to an AI and have working code in a few minutes. It is genuinely faster. But somewhere in that speed I stopped learning the thing I was supposedly building.
The economics make sense on paper. Less time coding means more time on distribution and customer problems, which is where indie projects actually live or die anyway. So rationally I should be fine handing off the technical work.
But here's the part that bothers me. If the moat for small builders used to be technical execution, and that moat is flattening, then the real differentiator becomes taste and judgment. Who can identify the right problem and build something people actually use. That's harder to shortcut and also harder to develop if you're skipping the messy parts.
Not sure if this is a skill atrophy concern or just resistance to a workflow that's genuinely better. The people shipping fastest right now seem unbothered by it. Maybe that's the right posture. Maybe something is getting lost that won't show up until later.
Curious whether people building with AI assistance feel like they're getting sharper or just faster.
[link] [comments]