Individual skill differences and human AI interaction shape creative outcomes

Abstract

Abstract This study investigates how individual differences in skill levels and human-AI interaction are associated with different creative outcomes, focusing on the cognitive, informative, and logical dimensions of interaction. Using quasi-experimental data, we explore how AI relates to creativity—specifically in terms of novelty and usefulness—across low-, medium-, and high-skill individuals. The findings reveal that logical interactions with AI are positively associated with both novelty and usefulness across all participants. Cognitive interactions are positively associated with novelty but not usefulness, while informative interactions are positively associated with usefulness but not novelty. Notably, medium-skill individuals achieve the highest creative performance overall, while low-skill individuals show the greatest potential for improvement through higher-quality AI interactions. These results highlight how individual differences and human-AI interactions are linked to different creative outcomes, offering insights into optimizing AI to enhance creativity for users with varying expertise.