Engineering Creativity: A Narrative Review of Creativity Science for AI Development

This review explores the link between engineering and creativity, challenging the perception gap between structured training and creative fields. It reframes human creativity insights from prominent scholars to inform the development of AI systems capable of creative problem-solving. The paper translates abstract and philosophical models into structured, computationally tractable frameworks to bridge human creativity research and machine learning applications. The review focuses on four core frameworks to guide AI design: Wallas’s Four-Stage Process, Rhodes’ Four Ps Model, Simonton’s Creativity-as-Influence Model, and Runco’s prevailing framework. It traces the historical progression of creativity research from early efforts by Guilford and Torrance to later dynamic frameworks by Amabile and Csikszentmihalyi. The document discusses how these models, which evolved from abstract theorizing to structured, multidimensional constructs, provide a foundation for examining and applying creativity within technical domains. It also addresses the growing integration of AI, distinguishing between human creativity and artificial creativity produced by machines. The forward-looking perspective suggests an augmentative role for AI within hybrid human-AI workflows. Ultimately, the review aims to provide a blueprint for developing AI systems that move beyond rote problem-solving to exhibit adaptive, context-sensitive, and generative capabilities, capitalizing on the synergy between creativity science and AI.