When looking at these pictures it is unbelievable how far "computer art" has come noe. I recently got access to Adobe Firefly (their new text-to-image diffusion model) and just typing in "art" or "beautiful art" produces, well, endless amounts of mind-blowing art. All from one or two words and a sequence of random seeds.
One could object that this isn't really algorithmic art in the early sense, since diffusion models rely heavily on imitating the style of human created pictures. But it "feels" the same.
Note that by typing in "art" into Firefly, you are essentially still hoping for randomness to create something surprising and beautiful, just as Jasia Reichardt described it in the article: "[T]he exhibition was about happy chance discoveries - with a computer." This is conceptually very different from typing in a long prompt describing what you already have in your mind. Which is a quite different use case of diffusion models, a more mechanical one. (There are two types of artists: Those that make art without much further purpose, and those that make commissioned work for ads or videogame assets etc.)
One could object that this isn't really algorithmic art in the early sense, since diffusion models rely heavily on imitating the style of human created pictures. But it "feels" the same.
Note that by typing in "art" into Firefly, you are essentially still hoping for randomness to create something surprising and beautiful, just as Jasia Reichardt described it in the article: "[T]he exhibition was about happy chance discoveries - with a computer." This is conceptually very different from typing in a long prompt describing what you already have in your mind. Which is a quite different use case of diffusion models, a more mechanical one. (There are two types of artists: Those that make art without much further purpose, and those that make commissioned work for ads or videogame assets etc.)