Better AI models enable more ambitious work

Summary

An eight-month study of developers at 500 companies using Cursor examined whether improved AI mainly increases existing work or enables tasks previously out of reach. The research, conducted with Professor Suproteem Sarkar of the University of Chicago Booth School of Business, covered releases of Opus 4.5 and GPT-5.2, finding average weekly messages per user rose 44%. Developers initially used better models for more work of similar complexity, then after a 4–6-week lag shifted toward harder tasks: low-complexity messages grew 22%, while high-complexity messages grew 68%, mostly in the final six weeks. Growth varied by sector, led by media and advertising at 54%, software and developer tools at 47%, and finance and fintech at 45%. Documentation, architecture, code review, and learning grew more than UI and styling, suggesting expanding AI-generated codebases increase demand for management, understanding, and review while opening broader productive opportunities.

Takeaways

1. 1

   The authors describe the 44% increase in AI usage as consistent with a Jevons-like effect, in which efficiency gains increase total consumption rather than reducing it.

2. 2

   The 4–6-week delay before more complex work emerged may reflect the time developers needed to discover new model capabilities and firms needed to reorient workflows.

3. 3

   The study hypothesizes different industry dynamics: competitive pressure may drive finance adoption, while more capable models may expand greenfield opportunities in media and advertising.