A few weeks ago I built an app on a ski lift. People liked the novelty of it. I've been worried ever since that the novelty was the whole takeaway — and that the actual point got lost in the thumbnail.
So this episode is the correction. It's not about ski lifts. It's about what becomes possible when you give yourself a hard constraint and stop waiting for better conditions.
I walk through the full Ideoloop build journey — starting in Google AI Studio, hitting walls, moving to VS Code, landing in Claude Code, making real progress, and then breaking the whole codebase in a way I thought was unrecoverable. Two versions out of sync on two different computers. I fixed it. And what I learned in that process is the whole point.
What's covered:
- Why the greatest breakthroughs come from one or two ingredients, not a thousand
- The Ideoloop build journey from Google AI Studio to Claude Code
- What breaking my own codebase taught me that no tutorial ever could
- How the 72-hour rule connects to the philosophy of constraints
- Why your clients — and probably you — have too many ingredients and not enough focus
- What it actually means to build something in a Petri dish
This one's more philosophical than most. I wanted to know if you're into that side of the show. Leave a comment and tell me.
Episode 9
