On building with AI
About a month ago I started working on a simple app called Mosey - a movement-gated podcast app.
What started as a simple way to only listen to podcasts while on a walk has somehow grown into a full-fledged podcast app meets step counter meets light fitness app - and I blame it on two things:
-
The fact that I’m apparently incapable of setting clear boundaries around what an MVP should actually look like. And that might also be cause by…
-
The fact that I can have AI build and build and build new things, which lands me in this trap of “just one more neat thing I’ll add” … except that one neat thing somehow comes with five more things, which add complexity, effort, and just enough complication to make me want to delete everything and start over.
…but the sunk cost fallacy is real, so instead I keep building…. which I’m starting to realize it may be the actual problem.