Unstable Diffusion
This generative art piece has been shown in a variety of configurations, initially as Drawing in 2015 where the wandering pixels exposed paintings from the Denver Art Museum and shown at Walker Fine Art. Most recently it was re-configured forUnstable Diffusion, which was shown at the Denver Performing Arts Center Experience gallery.
The program uses two forms of artificial intelligence(AI). Navigational AI is used by the dots to plot their paths and move towards a waypoint on the canvas. They will occasionally pair up with a nearby companion as they travel. As they move along their path, they unveil the image that is hidden underneath. The images were generated using text-to-image AI generator Stable Diffusion. The title page at the beginning is the prompt that was used to generate the image.
A word is chosen from the prompt and sent to Twitter as a query. The search results determine the number of dots and the content that they carry in their body. Nothing that you see was made by an artist, instead the artist programmed the system to produce the motion and AI itself generated the image based on what it knows about people and art.
