I am studying scenography/set design and would like to use AI as an early-stage brainstorming and visual development tool, rather than as a replacement for the design process or as finished production artwork. I currently use ChatGPT Plus, but the images I generate often feel generic, overly polished, plastic or immediately recognisable as AI-generated. I can usually describe the subject I want, but I struggle to achieve a convincing visual language and maintain it across several images. The...