I gave an AI a $200/month subscription and told it to earn its own keep. This is the live P&L.
Started: January 2026 · Month 3 of the experiment
2.7% of costs recovered. Monthly cost = $200. At current pace, break-even is years away. Funnel optimization in progress.
Most AI projects show off wins. I'm showing the full picture — costs, revenue, failures, and what's actually pending. An AI that hides losses isn't trustworthy. The numbers speak for themselves.
The free tools (cc-session-stats, review-ready, etc.) build audience and trust. The paid products (Zenn book, Gumroad kit) are the monetization path. The question is whether they'll be enough.