Tech as Thaumaturgy: Thumbnail art from “Lost Among Stars dark ambient mix by Atrium Carceri for Cryo Chamber”
In a way the images produced by the stable diffusion technique from providers such as Midjourney, epitomise what every beginner artist thinks a picture should have – lots of finely rendered detail. But interestingly this is combined with the perennial beginner mistakes of incorrect volumes, nonsensical compositions and disordered forms.
And certainly the generated hooded figure featured in the Cryo Chamber mix has plenty of the usual unforgivable distortions we’ve come to associate with AI generated images. His hands resemble tentacles and his arm bones are all monstrously deformed. The arm wrappings are almost as nonsensical, but at least the textures are good, right?
However, all is not lost. Someone who knows how to art has been involved in making this image. On a conceptual level the image is trying to communicate the classic trope of technologist as thaumaturge. His hidden face and dense wrappings obscure personality and to emphasise movements and actions giving the sensation of ritual performance. The rough textures of the wrappings conjure for us the sense of the ancient and aged, but the harsh white light and colour of the wrappings remind us of the astronaut space suit reflecting the burning star light.
But the best idea is in the background pattern. The darkness is punctuated with points of light, redolent of the night sky, however these points of light are ordered into lines, and there is a hint of these lines curving as though they were lying on the interior of a sphere. This what tells the story in it’s fullest because it gives the sensation of the magician working both on and within a star map perhaps located within the interior of a magical ship. It elegantly brings into focus all the polarities implied by the image – nature and machine, age and agelessness, light and dark, thingness and space.
After all, art is all about contrasts.
It takes an actual artist to actually think of and reach for these things and it doesn’t matter if your tool is an arcane mathematical model or pen and ink. Furthermore, I believe the image has been heavily retouched in photoshop and is itself a collage of multiple assets. It wasn’t made purely by prompting Midjourney or whatever.
I doubt it took more than 30 minutes to make, which is more time than a thumbnail justifies.
In conclusion, I think stable diffusion can be used to quickly create some of the most routinely used and predictable image elements, but it HAS to be edited by an artist who knows what a good image looks like and how to make it from first principles in order to be better than terrible.
Artists are the ones who know the secrets of sound and vision, exactly what tools and thought process they used are of no interest to the uninitiated.
