Resonant Something: Glitch Art Method Using Generative AI

This paper explores the concept of Resonant Irregularities using generative AI, in which artworks are semi-automatically generated by machines rather than humans, turning machines into unique artists. Stable Diffusion was used for the generative AI, and the process of gradually changing the image was repeated over thousands of times. In Resonant Shifts, the image and its description were set as a set and the image was changed. Resonant Chain used a full-length story, one sentence at a time, as a prompt, with the goal of expressing the story in a single image or a short animation. The process of creating an image from a prompt and turning the image into a prompt is repeated. The process moves back and forth between the two states of image and prompt. With these three different methods, I explored the expression of features that would be difficult for a human artist to achieve.