Modeling the dynamic interplay between student creativity and Gen-AI

Abstract

This study proposes a heuristic nonlinear dynamical framework to explore the evolving interaction between student creativity and the use of generative artificial intelligence tools such as ChatGPT. Creativity and AI usage are modeled as coupled, bounded variables whose interaction captures both constructive and suppressive influences, giving rise to multiple equilibrium regimes that represent distinct educational scenarios. The analysis identifies four equilibrium states, including a balanced coexistence regime in which creativity and AI usage persist simultaneously under appropriate conditions, and local stability analysis is used to characterize the qualitative behavior of the system near these equilibria. The model is intended as a conceptual and phenomenological tool rather than a mechanistic representation of cognitive processes, with the goal of providing qualitative insight into long-term interaction patterns and stability structures in AI-rich learning environments. To illustrate plausibility, the framework is applied to a small-scale classroom setting involving 100 undergraduate students enrolled in a design-based learning course, where creativity was assessed using rubric-based evaluations of student work and ChatGPT usage was measured through surveys and interaction logs. Illustrative parameter choices consistent with observed trends yield a stable coexistence equilibrium, supported by numerical simulations and phase-plane analysis. Overall, the study contributes a transparent mathematical lens for reasoning about creativity–AI interaction and highlights how balanced integration of generative AI tools can support creative engagement when reliance is moderated by intrinsic creative activity, while serving as a foundation for further theoretical refinement and empirical investigation rather than a predictive model of individual learning behavior.