Critical Sociology of Prompting: Connecting Verbal Inputs and Visual Outputs

We are witnessing a new era of creativity where prompt-based learning leads to creation of unique but also fictitious(?) digital/visual content. Prompting is the skill (and set of practices) for writing inputs for generative models, including visual generative models. These inputs are increasingly multimodal: not only written text but also source images, sketches, and reference materials that condition the output (image-to-image, inpainting, style and reference conditioning). Prompting is a repetitive process of interaction between humans and machines, where humans aiming for a perfect/desired visual product polish their prompts.

The aim of this session is to interrogate the process of prompting Imagination, where verbal and visual inputs are intricately connected with the visual outputs.

This session invites papers that interrogate various questions related to prompting: how do users write prompts, how do they increase the quality of prompts, how do they deal with machines to write better prompts, what are the actual practices of verbal input (modifying the prompt), prompt "polishing", and how do they combine verbal and visual inputs in hybrid workflows? Interrogating the constituents of the process of prompting is crucial to understanding the "backstage" of how visuals are generated with visual generative media.

Broader conceptual papers addressing and critiquing the concept of imagination are welcome to interrogate the impact of prompting on the imaginative processes in professions of graphic design, photography, advertising, architecture etc.

This session invites cross-disciplinary approaches between IT and Visual Communication.