The latest advancements in AI models, including OpenAI's GPT-5.2, are not just making developers more efficient; they are fundamentally changing the nature of their work and enabling previously out-of-reach projects. A recent study analyzing eight months of developer activity on Cursor AI models, spanning July 2025 to March 2026 and incorporating the release of models like Opus 4.5 and GPT-5.2, reveals a significant surge in AI adoption.
This period saw average weekly AI messages per user jump by 44%, a phenomenon consistent with the Jevons paradox, where efficiency gains lead to increased overall usage. The findings, detailed in research conducted with Professor Suproteem Sarkar from the University of Chicago Booth School of Business, indicate that developers first leverage enhanced AI for existing tasks before tackling more complex challenges.
Industry Adoption Spikes
The impact of these AI upgrades is not uniform across sectors. Media and advertising led the charge with a 54% increase in AI interaction, followed closely by software and developer tools (47%) and finance/fintech (45%).
In finance, this could be an AI arms race, where firms adopt AI to gain a competitive edge. Media and advertising, conversely, may be exploring new opportunities unlocked by more capable AI.
Complexity Curve
Initially, developers used improved AI models to perform more of the same tasks. However, a lag of four to six weeks preceded a significant shift towards more complex problem-solving.