ABSTRACT Over the past few years, with the advent of ChatGPT, generative artificial intelligence (GenAI) has been at the centre of numerous discussions regarding generative and creative power, especially with the hope that it will enhance human creativity and, consequently, transformative power. Creativity, especially human creativity, has long been studied, particularly by psychologists and design scientists, who have revealed the difficulty of overcoming the fixation effect. This effect can hinder creativity and has not yet been explored in relation to GenAI, even though it is of major importance for understanding creative issues. In this work, we propose leveraging the rich body of literature on design creativity and creativity management to shed light on the fixation effect related to GenAI tools, which are considered partners in the design process, and possible defixation techniques. To study this issue, we propose a twofold methodology: a first, in‐depth qualitative step with design theory experts, leading to the formulation of hypotheses, and a second, experimental step based on the well‐documented egg task in creativity to test the hypotheses. Ultimately, we show that the use of GenAI tools in creative endeavours results in a greater number of generated ideas and that it does not prevent the fixation effect but can foster idea exploration and thus help with second‐order defixation. These results consequently show the importance of considering a cocreativity regime involving both humans and GenAI, and they extend knowledge on the fixation effect in this context. They also indicate the capacity of GenAI in creative processes, thus suggesting some relevant use cases to leverage these tools.
Building similarity graph...
Analyzing shared references across papers
Loading...
Antoine Bordas
Centre de Gestion Scientifique
Pascal Le Masson
Centre National de la Recherche Scientifique
Benoît Weil
Centre de Gestion Scientifique
Creativity and Innovation Management
Centre National de la Recherche Scientifique
Université Paris Sciences et Lettres
Centre de Gestion Scientifique
Building similarity graph...
Analyzing shared references across papers
Loading...
Bordas et al. (Tue,) studied this question.
synapsesocial.com/papers/68a36c210a429f797332fceb — DOI: https://doi.org/10.1111/caim.70013