AI made building cheaper. Did it make quitting harder?
Building something used to take enough time that you had to think twice before starting.
Now you can get a working version together in a weekend.
But I wonder if that creates a different problem.
When you’ve spent months building something, it’s pretty obvious why you feel attached to it. What’s strange is that even a project built in a few days can be surprisingly hard to let go of.
You keep adding one more feature, changing the UI, trying another prompt. Not because anyone asked for it, but because fixing things is so easy now.
I think AI has made it easier to keep a bad idea alive.
I’ve started thinking that knowing when to stop might be a more useful skill than knowing how to build faster.
Has anyone else caught themselves polishing something they probably should’ve abandoned?
Replies
Yes, and the trap has a second form: polishing a feature nobody asked for inside a product that is otherwise fine. Two things helped me. The first was an app that quit for me. My previous one depended on an unofficial wrapper around a music model, the wrapper broke, and the app died with it. Rebuilding from scratch was easier than I expected, and the lesson was not "build faster", it was a one line rule at the top of the new project's notes: no unofficial APIs, ever. The cost of starting had dropped, so the rule became affordable. The second is a habit for the small quits: when I look at a feature and decide not to build it, I write it down with the date and the reason, "looked at, not built, because". A yearly plan paid monthly sits in that list. Without the note, an idea that is cheap to build comes back every few weeks and eventually gets built out of fatigue. With the note, the next time it comes up I read why I said no, and the question is whether the reason changed, not whether I feel like it today.