all +

all possible images by AI

Everybody works with "prompting", a text input, to create artificial images with the current tools of generative AI such as Midjourney or DALL-E 2.
Johannes Franzen chooses a contrary approach: He goes one level deeper, directly in the program code. Potentially, all images of the underlying AI can arise.
The AI "Stable Diffusion", trained on billions of images from the Internet, is his preferred software. And the result of this process?
The “ amalgamation of the entire visual world of our culture”, an idea of Brooklyn-based photographer and artist Charlie Egmann is one aspect.