A few months ago, I attended a local event, where photographer Doug Menuez spoke about his project “Wild Place: The People of Kingston, NY.” The description on his website begins . . .
Wild Place is the English translation of Wiltwyck, the original name given to Kingston, New York, in 1661 by Peter Stuyvesant and the Dutch who were facing fierce resistance from local Native Americans. My wife Tereza and I recently moved back to Kingston after a decade away and can see lots of changes, with more to come. It seems like an important moment.
Combining portrait and documentary in both photographs and short video interviews, “Wild Place” presents contemporary Kingston through Menuez’s view of its artists, activists, entrepreneurs, community leaders, and—not surprisingly—people who fit all those descriptions. While listening to Doug talk about the project, I was reminded why I care so much about artists and their work: because through art and artists, we renew profound, even cathartic, connections to what it means to be human and, in turn, reinforce the reasons why humans bother to make art. My schedule does not permit frequent attendance at such events, but listening to Doug’s articulate, thoughtful, even spiritual discussion about his work was as close I come to listening to a sermon.
In my last post commenting on visual works generators like DALL-E, et al., I reiterated the view held by many that the notion of “AI art” is oxymoronic—as devoid of meaning as having a machine perform a religious rite for its human owner. Whatever creative work without humans ought to be called, it is not art. As such, I maintain that nobody will be interested in works made exclusively by machines for very long and that the current buzz about these generative algorithms may ebb quickly into the sea of trends to swirl in gooey eddies of crypto and NFTs.
This is not to suggest that creators and advocates of creators’ rights should ignore current threats to human artists, or that generative AIs do not preface an even darker version of the “information age” than the present state of madness. In a Facebook post that has been widely shared, a philosophy professor describes catching the first student in his class to use a bot called ChatGPT to write an assigned essay about David Hume. “The essay confidently and thoroughly described Hume’s views on the paradox of horror in a way that were [sic] thoroughly wrong,” the professor writes. “It did say some true things about Hume, and it knew what the paradox of horror was, but it was just bullshitting after that. To someone who didn’t know what Hume would say about the paradox, it was perfectly readable—even compelling.”
That last sentence is unsettling in a world buffeted by conspiracy mongers and alternative facts. No Alex Jones or Donald Trump or Stewart Rhodes required. The next cult figure can be an algorithm producing a “readable—even compelling” restatement on any matter from the Enlightenment to the suppression of viral disease. It is intriguing, if depressing, that a college student attempted to cheat by means of an AI to avoid honest engagement with Hume’s essay Of Tragedy, which contains the following observation:
We find that common liars always magnify, in their narrations, all kinds of danger, pain, distress, sickness, deaths, murders, and cruelties; as well as joy, beauty, mirth,’ and magnificence. It is an absurd secret, which they have for pleasing their company, fixing their attention, and attaching them to such marvellous relations, by the passions and emotions, which they excite.
Hume could be commenting on the recently announced Trump NFT “trading cards,” which appear to comprise stolen images from the internet and badly photoshopped heads in a series of bizarre portraits depicting Trump as soldier, rancher, business leader, and even a costumed and be-muscled superhero with lasers shooting from his eyes. I got nothin’ except to say that there is no paradoxical pleasure in viewing this particular horror.
On a more sophisticated level, generative algorithms like MidJourney, DALL-E, and Stable Diffusion are all “trained” by inputting a corpus of human-made creative works, most of which are scraped from the internet without permission of any living artists who still own the rights to the works. As PetaPixel reports, MidJourney founder David Holtz flatly admits feeding his system millions of images without permission, and illustrator Molly Crabapple, in an OpEd for the L.A. Times writes:
While they destroy illustrators’ careers, AI companies are making fortunes. Stability AI, founded by hedge fund manager Emad Mostaque, is valued at $1 billion, and raised an additional $101 million of venture capital in October. Lensa generated $8 million in December alone. Generative AI is another upward transfer of wealth, from working artists to Silicon Valley billionaires.
That these AI “art” generators represent yet another example of economic destruction without the creative part is a certainty. Less certain are some of the copyright questions, for instance, whether input of protected works for “machine learning” is infringement. This will remain a theoretical/ideological debate for attorneys, academics, and copyright nerds like me until one of two things happens: legislation or litigation, both of which move at a crawl compared to the market for new tech toys. If a lawsuit began tomorrow, for instance, it would be hard to say whether the legal questions presented will still be relevant to the market by the time the case is resolved.
Perhaps the real potential of the generative algorithm lies not with illustration or design or music composition, but with medical diagnostics or some other valuable purpose. If computer science is a true science, then it must allow for unintended discovery, and who’s to say that an experiment in “AI art” cannot be the precursor to an algorithm that helps identify genetic disposition for certain infections?
This does not mean, of course, that we should excuse models in the present that undermine the rights or value of the human artist. On the contrary, I mention this alternate history to emphasize the point that of all the things we can do with computing power, one thing we absolutely do not need are machines that make “art.” Tellingly, Hume’s essay is mostly about art, and to the question whether creative expression about tragedy can provoke a sense of pleasure for the audience, he replies:
This extraordinary effect proceeds from that very eloquence, with which the melancholy scene is represented. The genius required to paint objects in a lively manner, the art employed in collecting all the pathetic circumstances, the judgment displayed in disposing them: the exercise, I say, of these noble talents, together with the force of expression, and beauty of oratorial numbers, diffuse the highest satisfaction on the audience, and excite the most delightful movements.
Maybe the AI cheerleaders will accuse me of anthropic maximalism, but in addition to doubting that an “AI artist” could ever express anything close to the transcendent experience Hume describes, I am certain that we do not want it to even try. Art is human. There are better uses for computers.
Photo by: Abrill
