Picture This: AI-Assisted Image Generation as a Resource for Problem Construction in Creative Problem-Solving
Janet Rafner, Blanka Zana, Peter Dalsgaard, Michael Mose Biskjær, Jacob Sherson
Abstract
In this paper, we explore the potential of AI-assisted visualization during the problem identification and construction phase of the creative problem-solving process. We examine this within the context of the ongoing crea.visions research project, which employs AI technologies to visualize citizens' visions of the future. Our findings underscore various factors contributing to the effectiveness of assisted visualization in this setting, such as: 1) the tool's dual role as both a visual and ideational aid, 2) the introduction of innovative collaborative elements like prompt engineering, 3) the enhancement of visual expression without requiring artistic skills, and 4) the facilitation of idea communication. We also recognize limitations related to the tool and the problem context such as abstract concepts. This study serves as a foundation for future research on AI-assisted image generation as a resource in creative problem-solving, laying the groundwork for the creation of increasingly effective and user-friendly tools.