The AI generative text-to-image creative learning process: An art and design educational perspective

Abstract

In today’s constantly changing world technological developments in artificial intelligence (AI) can induce educational visions of both utopia and dystopia. New technologies and communication platforms can provide new forms and possibilities of learning. Creating an image has historically mostly been a human process of using knowledge and application of technique that demanded training. This image-making process changed with the invention, development and spread of the photographic camera, when creating a detailed visual representation of reality became a possibility without a complex process of craftsmanship and artistry. The nature of visual art changed but the visualisation of ideas and prefigurative thoughts could not necessarily be captured by a camera. With the development and spread of AI text-to-image generation, can this change the need for competency to visualise ideas in the way the camera changed the need for drawings and paintings as visual representations?  This study explores how AI text-to-image generators can contribute to and change art and design education. We conducted exploratory experiments where we tested a variety of AI text-to-image generators and explored the outcome of using different generators, prompts and settings. Reflections were written down throughout the process. This was combined with an online ethnography on a text-to-image community. Different potentials of learning were identified, as well as issues of interaction and possible contexts of use. The results are discussed in a future learning context.