AI art has sparked debate about who an artist is and what we consider art. However, there is a more pressing issue to consider: the expenses of AI software that generates art. What happens when a computer creates something in seconds, relying on an infinite repository of human-created art? The ethics get complicated here, and the cultural discourse around AI art rapidly becomes philosophical.
The thing with AI art is that it is simple to create. And there’s a hazy concept of who its “artists” are, since although the authorial purpose originates from a person, the actual work is created by a computer. And it, too, is a computer that gathers data from a treasure of pre-existing databases, which are rife with bias. Previous study has shown that AI perpetuates racist and sexist preconceptions because it fails to detect biases in the datasets it is given.
Furthermore, anybody can utilize open-source AI art software to create (almost) anything – and what they can’t, they can typically work around. People who previously would not have gone to the trouble of making improper or unpleasant art with their own hands may now do it by merely inputting a few commands into software – and it only takes a few seconds. This, along with the fact that it can be used to make money for the few individuals who control the program, is the opposite of art as a public benefit. “Technology is rapidly being used to create gig employment and enrich billionaires, yet so much of it does not seem to serve the public well enough… AI art is a component of this. To developers and technically minded people, it’s this cool thing, but to illustrators it’s very upsetting because it feels like you’ve eliminated the need to hire the illustrator,” said cartoonist Matt Borrs, in response to The Atlantic using an AI-generated image in a newsletter rather than hiring and paying for a human illustrator.
Consider the latest advancement that enables users to modify actual human faces using DALL-E 2, an AI art platform modeled after legendary surrealist Salvador Dali. Filters and photo-editing technology have already raised concerns about the authenticity of art; AI may not only amplify this discussion but also vastly extend the potential for damage.
Another issue with AI art is that it is quite simple to create. It’s why artists are worried about becoming the first to be replaced by machines, a situation that should worry us all. Furthermore, by drawing from pre-existing pictures and ideas, AI art may react to novel impulses while not necessarily being innovative. It may infringe on existing copyrighted material in the process, and, more crucially, it disguises something more nefarious – copyright infringement, mass manufacturing, duplication, and, eventually, content creation – as creativity.
Content creation is not an art form in and of itself. For a better comparison of what we’re losing with AI art, look to AI scripts – themselves an online joke. Give an AI program a few prompts and it can produce a book, screenplay, story, essay, or term paper in seconds. Are they any good? Do they make any sense? Perhaps not; they’re just experimenting with new combinations of what already exists.
When something is so readily replicated, recycled, and regurgitated, I feel it loses part of its soul. And, given how artists take from history, experiment with format, and experiment with ideas, can a computer doing the same make things worse? Yes, arguably. It detracts from the creative process, contributing to the culture of unlimited material, ubiquity, and speed that has come to characterize consumer culture. We may no longer have to wait to observe what happens to art when it is brought to its logical conclusion – we are most likely already there.