AI coding assistants provide a 30-40% productivity boost under specific conditions: greenfield projects, popular tech stacks, and low-complexity tasks. Research shows productivity gains drop to 0-40% for mature projects and can even be negative. The key insight is that an engineer's goal isn't writing code faster, but

10m read timeFrom evilmartians.com
Post cover image
Table of contents
What the studies sayBuilding dev tools with AI: the product decisions that need to be yoursDebug AI fast with this open source library to visualize agent tracesAn engineer’s goal is not writing codeUX research for dev tools must not be neglected: our quick-start guideWhat is AI best used forDon't just slap on a chatbot: building AI that works before you askBottlenecks and hiccupsWeeks → days: a case for expert-led, AI-driven design engineeringHow to reach maximum productivity boost when using AIVibe coding in style.mdA productive conclusion

Sort: