Prompting as Thinking-with:
A visual essay
This visual essay, co-authored with Monica Monin, discusses a creative collaboration in which we use text-to-image prompts as a way to think across a range of ideas including the relationships between collage practices and AI image generation – both modes of image-making that create images with images – as well as taking an ‘anarchival’ approach to addressing absence in historical archives.
Initial experimentation with prompt-based model DALL-E 2 involved writing multiple prompts to generate images of the extinct King Island dwarf emu; specifically, an emu taken to live in Empress Josephine’s estate outside Paris. There is little visual record of the dwarf emus, and what remains is ambiguous and factually inaccurate. The scarcity of visual reference material provides an interesting case study for how a generative image model might attempt to elaborate a new image about a historical event. The results provide material to help think about how image generation models work, and also how we might visualise the experience of an extinct species.
Reflecting on the initial experiments, we began to consider prompting with large-scale image generation models as a way to think-with and speculate, rather than to merely generate. We employ two methods to critique the resulting images: visual content analysis and comparative analysis across image-generation models. We conclude that at a time of both deliberate and accidental miscommunication, it is important for those with expertise in how images ‘work’ to critique and analyse image-generating tools, and consider how working with generative AI might be included as part of an anarchival practice.
Published in ArteActa 13/25: AI (and) Art: Poetics of Prompting: