Dreams Daddy
A comedic dream misinterpreter. Built to test how fast AI tools can ship a project.
In July 2025, I wanted to test how fast I could build and ship a project using AI coding tools.
The Experiment
I tried Bolt.new first. Within 20 minutes, I had a pretty website. But the quality was bad — plenty of errors, broken interactions, and code that didn’t work properly.
A beautiful UI means nothing if the product doesn’t function.
Finishing the Project
I switched to Claude Code and rebuilt it properly:
- Fixed all the broken interactions
- Added a real backend for the AI dream interpretations
- Implemented rate limiting and error handling
- Set up proper hosting and deployment
The “last 20%” took longer than the “first 80%”. This is the reality of shipping — AI tools can scaffold quickly, but finishing requires understanding what you’re building.
The Dream Interpreter
Users describe their dream and get a deliberately absurd “cosmic interpretation” back. It’s comedy, not real dream analysis.
Ads Experiment
I wanted to test running ads to see if the site could generate revenue. But ad networks require established traffic before approval. The site needs more content and visitors first.
Current Status
I don’t work on this project anymore. It was a learning experiment about:
- The gap between “demo” and “production”
- Why scaffolding tools still need human finishing
- How ad monetization works (or doesn’t) for small sites
The site is live but not actively maintained.