Co-creating with AI. Exploring and integrating stable diffusion models in the creative process of poster design

Abstract

This thesis investigates the use of Artificial Intelligence as a co-creative tool in graphic design by exploring text-to-image and image-to-image generation with personalised Stable Diffusion models.

The aim is to find new ways in which AI can be integrated in the creative process of poster design. The practical experiments use these models to generate images in relation to a specific design concept, which are refined through iterative human interventions.

The findings of the thesis reveal that although AI can help in generating diverse inspirational outputs and personalising ideas, these images still need to be selected, interpreted, and manipulated. This is necessary to achieve a visually functioning and readable poster. Despite discussions on AI potentially replacing designers and creativity, this research sees AI not only as an assisting design tool but as a personalised collaborator changing future design practices while preserving the essence of human creativity.