The original title is "I built 6 useless (and useful) things with AI in 30 days"

Short summary

A laid-off developer built 6 AI-powered projects in 30 days: a Chrome sidebar extension (15 active users), weekly report automation, email drafter, resume matcher, podcast summarizer, and math problem solver. Three run daily because they solved personal pain points first; three failed due to lack of pre-validation. Key insight: 7-day shipping deadlines prevent over-polishing; all products shipped in 1-5 days using Cursor, GPT-4, and Claude.

• •3 of 6 projects ship actively because they solved a problem the creator experienced repeatedly before building
• •7-day deadline forced ruthless prioritization; projects killed early avoided months of wasted optimization
• •AI tooling made shipping so cheap (1-5 days) that validation must happen before building, not after

Generated with AI, which can make mistakes.