Images from Images: Generative AI and the Reconfiguration of the ‘Photographic’

Abstract

AI image generators such as Midjourney, Dall-E or Stable Diffusion are able to perfectly simulate the appearance of photographic images, but they are no longer part of the history of optical media. Unlike photography, AI does not capture light to record images of the world, not even a simulated one, but analyses recurring patterns in vast amounts of visual data to synthesize images from images. The "photographic" in turn becomes a mere style, an esthetic surface quality that can be extracted, interpolated, and transferred from one digital image file to another. The essay asks what this reconfiguration of the "photographic" means for our understanding of photographic mediality, both for its future and for its history.