Back to AIBriefs
AnalysisDevelopers

AI coding agents boost code output 17x but releases only 30%

Ethan Mollick
@emollick.bsky.social

Professor at Wharton, studying AI and its implications for education, entrepreneurship, and work. Author of Co-Intelligence. Book: https://a.co/d/bC2kSj1 Substack: https://www.oneusefulthing.org/ Web: https://mgmt.wharton.upenn.edu/profile/emollick

View on Bluesky
Ethan Mollick avatar
Ethan Mollick
@emollick.bsky.social

Big paper on AI coding agents using Github data. The early auto-complete tools (like Copilot) led to 2.2x as much code, local agents like original Claude Code led to 7.4x, & current remote coding agents 17.3x(!) But human bottlenecks in the coding process means actual releases "only" went up 30%.

·
9 days ago
AI coding agents boost code output 17x but releases only 30% — AIBriefs