AI Coding Tools Boost Shipped Releases by 30%, Humans Remain Complementary
soumitrashukla9 · x · 2026-07-28
Discusses the paper "Writing Code vs. Shipping Code," which combines public GitHub records for over 100,000 developers with confidential Microsoft data to analyze the impact of different generations of AI coding tools across the software production hierarchy.
The study reveals an attenuation effect of AI productivity gains along the code production chain. While AI generates enormous gains of 1000% or more at the top of the chain (e.g., lines of code), it translates into only about a 30% increase in actually shipped releases.
This finding helps rationalize the current employment effects, explaining the lack of job displacement. It suggests that humans act as a necessary complement to AI-friendly tasks, with increased hiring observed in exposed areas.
Related event: Study Shows AI Coding Tools Yield Only 30% Gain in Final Delivery(2 posts)→
More from AGI Musings
- Yampolskiy says rapid AI progress still falls short of a true singularity — romanyam · 2026-07-28
- The Open-Sourcing AGI Dilemma: Democratization vs. Market Monopoly — Liu_eroteme · 2026-07-28
- AI agents need Responsible AI built in from day one, not bolted on later — iamKierraD · 2026-07-28
- Paper on AI and programmer productivity finds humans and AI are strong complements — mattbeane · 2026-07-28
- AI should be seen as evolution, not unfairness to the people who struggled first — seanmcdonaldxyz · 2026-07-28
- Anil Seth to keynote AGI 2026 on the prospects and pitfalls of conscious AI — anilkseth · 2026-07-28