I built 128 things with AI. 127 are dead. So I started shipping into real OSS instead.

Short summary

A developer built 128 AI-powered projects in four months, discovering 127 failed due to weak distribution. After analyzing the pattern, they pivoted to contributing meaningful localizations of established OSS tools (Hoppscotch, Jan, Betterlytics) for Japanese users. The lesson: shipping into existing communities through real code contributions beats building solo projects, with impact verified by GitHub merges, not marketing.

• •Built 128 AI projects in 4 months; 127 failed due to distribution, not quality
• •Pivoted to localizing existing OSS tools (Hoppscotch, Jan, Betterlytics) into Japanese
• •Verifiable GitHub metrics prove impact; shipping into real communities beats solo builds

Generated with AI, which can make mistakes.