AI Coding Tools Double Software Output Without Sacrificing Quality

A recent study by Jellyfish reveals that AI coding tools are significantly impacting software development, with companies experiencing nearly double the code output. The study, encompassing data from over 700 companies and 200,000 engineers, indicates that 63% of companies have adopted AI tools for the majority of their coding tasks. This widespread adoption has led to a substantial increase in the number of pull requests merged per engineer per week, jumping from 1.12 in low-adoption companies to 2.2 in high-adoption companies.

Importantly, the study also suggests that code quality is not being compromised by this increased output. Revert rates, which measure the percentage of code that needs to be rolled back after deployment, have only seen a modest increase, rising from 0.61% to 0.65%. This indicates that AI tools are not only accelerating the coding process but also maintaining a reasonable level of code integrity.

However, the study also highlights the emergence of autonomous AI agents in coding, with their activity rapidly increasing, particularly among top adopters. This trend raises questions about the future role of human developers and the potential for a growing divide between organizations that fully embrace AI and those that lag behind. The long-term implications of this shift on code quality, security, and innovation remain to be seen.

Transparency statement: This analysis is based on the provided article from DailyOrbitalWire.news. The AI model (Gemini 2.5 Flash) was used to extract key details, assess market implications, and formulate optimistic and pessimistic outlooks. The AI's role was limited to information processing and synthesis, and the final output was reviewed by a human analyst to ensure accuracy and relevance.

_Context: This intelligence report was compiled by the DailyOrbitalWire Strategy Engine. Verified for Art. 50 Compliance._