Prof. Anthony Cross - Tool, Collaborator, or Participant: On AI Art and Artistic Agency

Anthony Cross
Tool, Collaborator, or Participant: On AI Art and Artistic Agency
The past few years have seen an explosion in AI-generated art. AI image generation tools such as DALL-E, Midjourney, and Stable Diffusion are now capable of generating sophisticated and compelling images from simple text prompts. In this paper, I take it as a given that these image generation tools can be used for artmaking; I focus specifically on how artists might make use of them to create art. Most existing discourse surrounding AI art falls under what I call the production model: artists rely on AI image generators either as tools or as collaborators in the production of an artistically significant output. It is this output-an image, say-which is the artwork, and which is the focus of our appreciation. I introduce an alternative way of thinking of AI art, which I refer to as the exploration model. According to this model, artists instead relate to AI as a participant: artists create a space for interaction with the AI algorithm by way of their prompts, thereby allowing them to explore the way that the algorithm “sees” and “represents.” AI art practiced in this fashion bears a striking resemblance to contemporary conceptual and participatory art: the artwork is not so much the output-the resulting images-as it is the artist's structuring an interaction with the AI algorithm. I argue in conclusion that which model of AI art we adopt will have significant implications for our understanding and appreciation of AI art, with implications for the appreciation of style in AI art; concerns about novelty and originality; and the assignment of artistic credit and copyright.