AI Cribs From Friedrich, Moreau

Jason M. Allen's AI-assisted Théâtre D'opéra Spatial becomes the latest dubious artwork to rate a portentous think piece in The New York Times.

Allen used Midjourney, a machine-learning program that draws on a large training set of human-made art. Allen's image surely has roots in medievalizing fantasy art. But the woman looking at the landscape, seen from behind, is a motif popularized by Caspar David Friedrich. Overall, I'd say that most of the visual tropes that matter in Théâtre D'opéra Spatial can be found in Gustave Moreau's Salome Dancing Before Herod, from the mid 1870s.

Gustave Moreau, Salome Dancing Before Herod, 1874-1876. Hammer Museum


I am not hating Jason M. Allen's AI-assisted Théâtre D'opéra Spatial.
I'd like to see more of what it can do.
All to the good if it can mimic human ingenuity (Moreau's) and creativity.
But can it create artistic ideas on its own? Time will tell.
Give me something edgy (Ensor) or something hypereal (Campin).
Luce said…
AI can only recombine existing elements and I assume put filters on them. That is not making inventing creating a new image.
Re above comment: But that is what human artists do all day, what they have always done.
As Picasso said, all art is derivative.
Yet in this recombining, new artistic ideas emerge.
Anonymous said…
"The painter must not hesitate to run the risk of being taken to be a mere dauber. 'One paints for very few people,' writes Cezanne. Recognition from the regulatory institutions of painting - Academy, salons, criticism, and taste - is of little importance compared to the judgement made by the painter-researcher and his peers on the success obtained by the work of art in relation to what is really at stake: to make seen what makes one see, and not what is visible."

--- JF Lyotard, The Inhuman