Twittiots have been spouting off on the issue of A.I. generated art complaining that it's theft of art because of the way it's trained.
This is NOTHING NEW. It's the computer equivalent of Tracing or Lightbox Abuse. Human Artists have been copying each other for decades. Masamune Shirow is the most often art-thefted artist. His stuff has been "stolen" more than any other.
A.I. generated art is DERIVATIVE. There is No Way to figure out whose styles are in the images it produces. I challenge anyone spouting negatively about A.I. art to find a specific style being used from a specific artist. And then I'll respond with "Are you sure you're not simply seeing something there just because you Want to see it?"
Another way they claim it's theft is from Loss of Income. Your Paid Artists worry about losing income because nobody wants to pay for fan art of their favorite anime characters doing things the official stuff never had them do. LET ME TELL YOU SOMETHING - these A.I. art generators are good but there's still some concepts they still don't get. Sure you can make them churn out 100 images of pretty blonde girls but forget about specific cultural items from some place in the Asian Sub Continent. You want a depiction of a magic illusion trick on stage nobody has seen before and only you know what it is? A.I. can't do it. But beautiful white China style wedding dress with a cleavage window? It can suggest designs for hours. And each one is pretty good even if sometimes unfeasable.
And sometimes you get missing body parts, extra body parts, distorted perspective lines, messed up hands... It still makes mistakes. However, describe a scene well enough in generic terms it should understand, and maybe just maybe you'll get a good starting point of concept art. Still derivative from its massive data set which it draws from BUT DOES NOT DIRECTLY COPY. Get over it & find a thing you artists can make money at which A.I. can't do. You're supposed to be creative - so create already!
