In this video, Philip Dyer compares the prompted output of three popular AI image generators: Dall-E Mini, Midjourney and Dall-E 2. Dyer’s interest is the photonormative output, but it goes places and you get the picture.
I pit these 3 AI image generators against each other to see which one reigns supreme. Spoiler alert: DALL-E 2 wins.
I think Dall-E Mini’s crappy Francis Baconator style is what’s actually popular (particularly because it’s accessible) and shouldn’t be dismissed for its lack of realism. To quote Brian Eno: “Anything you now find bizarre, ugly, uncomfortable and unpleasant about a new medium will surely become its signature.”
Midjourney serves to generate cursed concept art.
Dyer is right that Dall-E 2 is where it’s at if you want superficially convincing shots. I could also imagine using it to speed up digital art commissions in specific styles or genres. Its output is good enough for painting. It (or whatever else is released to the general public) will soon replace “photobashing” because what amateurs and wannabes think is below them, but pros do it all the time.
PS realism is a red herring, a fixed idea of the gamer generation and a paranoia of the TV generation, which forgets how easy it is to simulate realistic output without the help of AIs. Consensus reality is dissolved by hand-crafted lies, network effects and the economic implosion and consolidation of news media, not these AI image and text generators.