While it is effectively impossible to pin down what a work of art is or could be (except, perhaps, within a framework of “frames”), the concept of what art itself is seems slightly more approachable. Let’s start with a dictionary definition, like that from Wikipedia:
Art is a diverse range of cultural activity centered around works utilizing creative or imaginative talents, which are expected to evoke a worthwhile experience, generally through an expression of emotional power, conceptual ideas, technical proficiency, or beauty.
Circumstances may change, and image manipulation with generative AI—stills as well as moving pictures—could conceivably develop into new art forms over time (more on that later). But in its current state, many characteristics from that definition seem to be missing right off the bat, particularly around the stern.
First, AI imagery certainly doesn’t evoke a worthwhile experience by letting us marvel at the technical proficiency required to produce Ma Deuce-operating cats or epically slow-motioned space troopers. “Prompting” has about as much to do with technical proficiency as with being able to switch channels fifty-seven times to find a TV show that isn’t completely vacuous.
Then, AI imagery doesn’t evoke a worthwhile experience through conceptual ideas; the Ma Deuce cat might fleetingly amuse, and the space troopers might make you temporarily forget how hard Michael Bay movies suck, but the experience is about as worthwhile as watching commercials during television timeouts.
Next, the emotional power of AI imagery doesn’t exceed that party-size bag of Doritos whose flavor is tasty and invigorating until it isn’t, after which you eye the crumbs all over your recliner with a slightly bloated and vaguely twitchy feeling.
Finally, AI imagery doesn’t provide a worthwhile experience in terms of beauty, except if your idea of beauty revolves around models shopped into unrecognizability to decorate the glossy covers of slick upmarket magazines without distracting features like personality or genuine emotion.
After that, all that’s left are the possible characteristics of “cultural activity” and “creative or imaginative talents,” and neither seems to be a good fit to describe the current state of affairs. Certainly, producing AI slop has become a kind of “cultural activity,” with a substantial part of the population consuming the energy requirements of a medium-sized country on a daily basis to produce a steady stream of clips and images each of which has already been forgotten the moment the next one hits the retina. But qualifying as a cultural activity isn’t such a lofty hurdle to take; having a Fourth of July cookout with or without getting plastered equally qualifies while being vastly more enjoyable.
Sure, what people delineate as being art or not has constantly changed over time. We can and should reflect on the numinous and the sublime, on intentionality and authenticity, theory and craft, mimesis and representation, and all the paths we have taken from Maros-Pangkep to Lascaux to today. Now, with these 50K years under humanity’s collective belt, I think 4′33″ is art, I think Warhol eating a hamburger is art, and I think Black Thought on HOT 97 is art. (And I certainly think that Adorno is a pompous snob who hasn’t read Benjamin closely enough.) But AI imagery isn’t art. AI imagery, so far, is a never-ending stream of ooze, pumped out of plagiarism engines into our social fabric like sewage out of discharge pipes into our rivers, and it’s even an open question which inflicts more harm to the environment.
However, I’m fairly optimistic with respect to the promise of art. Not only are people ingenious. They also love a good challenge, and they will enthusiastically seize novel opportunities to express what they want to express and create cognitively and emotionally rewarding experiences that break new ground. Almost everything humans can do or found out they can do, beginning with everyday activities like speaking, dressing, running, throwing, or cooking, can be transformed into forms of art through creativity, imagination, ambition, and undreamed-of technical proficiencies. Someone some fine day, I believe, will take generative AI image manipulation techniques and use them to create new and exciting kinds of art we couldn’t foresee.
But generative slot machines, manufactured, controlled, and handed out like crack cocaine appetizers in the guise of individual empowerment by the worst actors imaginable in terms of liberal values, environmental sustainability, and even a sustainable economy, that’s a far cry from facilitating new kinds of art, or any kind of art.