ChatGPT & Me: Kenny Schachter Tries to Use AI to Write Intelligently About Art. Is It Any Good?


Okay folks, let’s play a little game: guess which part, if any, of this essay was derived from ChatGPT, and which was written by me. I personally don’t see what all the artificial intelligence brouhaha is about—whether art and creativity could possibly derive from it, and/or if it poses a new threat to us, beyond the menace we already pose for ourselves on a daily basis. Though I have recently had my crypto wallet hacked, I’ve yet to meet a dishonest (or evil) computer. Let’s not forget it’s people that are fundamentally flawed by nature, not machines.

When I was in law school in the 1980s and the first desktops (along with attendant floppy discs) became accessible and affordable, I jumped at the opportunity to buy one. I never harbored programming ambitions, it just immediately dawned on me that this was a paradigm shift in the tools available to enhance human thinking and productivity. Yes, the big, unwieldy, institutional-looking IBM-beige (how else can you describe that clinical color?) machine had a foreboding air, to say the least, but I bought it in any case. And haven’t looked back since.

I initially used the device solely for word processing, typing 1,100 pages of handwritten notes to study for the bar exam and soon adding rudimentary photoshopping skills to my repertoire for making art. It certainly transformed the way I composed and edited text and images. The same could be said about AI in relation to writing and making “photographs.” I dabbled in ChatGPT to augment my coverage of an art fair I didn’t bother to attend, though I came clean to my editor (and readers) that I asked the software to suffer through the event for me as an experiment. It even came up with a few enviable jokes when prompted to craft some text with my inimitable humor and sarcasm. I guess I’m not quite as remarkable as I previously thought.

I recently got into an argument (surprise, surprise) with an artist who posted something to the extent that prompts are not an artistic tool in the same sense as a pencil or paintbrush—and that facilitating an AI program to depict a realistic landscape is not the same as a hand-drawn version. Actually, I’m no fan of lifelike facsimiles in art no matter how (or by whom, or what) they are made. And besides, why shouldn’t language drafting parameters for computers to compute be on par with the components of artworks? Call it the art of words. And as for the unpredictability of AI outputs, cultivating chance in art has long been a trope at least since Marcel Duchamp’s  () of 1913–14.

The artist Kevin Abosch succinctly summed it up as follows: “The notion that the human role is diminished in the creative process the moment an artist works with a machine-learning algorithm is a vestige of a long-held prejudice towards technology in the arts. By educating people about emergent technologies, and how exactly the artist controls these new tools, we are seeing more acceptance of AI in the arts.”

An AI reenactment of the first AI conference in 1956 at Dartmouth College, cigarettes, bad art, and all. Courtesy of Kenny Schachter and DALL-E.

Once upon a time, in the not-so-distant past (the mid-1950s at a Dartmouth College conference, to be exact) AI was a mere curiosity, a frivolous endeavor for chain-smoking nerds in lab coats. They dreamed of creating machines that could surpass our human limitations in the hope that these mechanical marvels would become the new Picassos and Mozarts and beyond. They were about to unleash a Pandora’s box of utility and futility, and now there are many people who want to close it back up again, citing the threats AI can cause both individual workers and humanity as a whole. But the mantra of art’s most enduring “-ism”—capitalism—rings loudly in our collective ears, and now there’s no turning back. Anything for a few bucks—or rather, a few million.

In its early stages, AI was as artistically capable as a penguin walking a tightrope or… KAWS. (Oh relax, just kidding.) Its creations resembled a drunker-than-usual Jackson Pollock drip painting that made you further question the sanity of the artist (or programmer). Critics scoffed, and still do, dismissing AI as a passing fad, destined to be forgotten like the pet rock or parachute pants. (Though there’s an art dealer in New York who still proudly sports them—the pants that is: Gagosian’s Adam Cohen.) Beauty, fashion, and art are in the eye of the beholder (sort of). But the potential of AI art is more than subjective—works of genius most assuredly will arise. And soon.

AI art and a drunker-than-usual Jackson Pollock doing their thing. Courtesy of Kenny Schachter and DALL-E.

AI art and a drunker-than-usual Jackson Pollock doing their thing. Courtesy of Kenny Schachter and DALL-E.

With each passing year, AI grows more sophisticated, like a chameleon donning a thousand masks. It learned not only to mimic the greats, producing paintings in the style of masters new and old, but helping create original art as well (with the necessary help of its human taskmasters). The art world gasps at the untried and unproven—until they learn to profit from it, then the refrain shifts to “I always knew algorithms could challenge the divine spark of human creativity.”

As we fast forward to the present, AI and creativity have become strange bedfellows. Collaborations between human artists and intelligent machines have become all the rage. It’s like a bizarre dance, with one partner leading and the other struggling to keep up. They tango, they disco, and more often than not, they trip over each other’s virtual feet. It’s a spectacle that leaves you scratching your head, under your breath, and biting your lip in anticipation as to what could possibly be next.

And what does the future hold for this tumultuous relationship? Will unaided AI surpass human creativity altogether? Will it render us mere spectators in a world of machine-made masterpieces? Or will it all come crashing down like a Jenga tower built on a sand castle? One thing is for certain—AI and creativity will continue to dance their quirky, awkward waltz, leaving us bemused, bewildered, and, not least of all, annoyed!

Most AI art is akin to a penguin on a tightrope, and most of it has traveled about as far. Courtesy of Kenny Schachter and DALL-E.

Most AI art to date is akin to a penguin on a tightrope, and most of it has traveled about as far. Courtesy of Kenny Schachter and DALL-E.

There are those that will always consider artificial intelligence the equivalent of Ozempic, the antidiabetic pharmaceutical wonder drug that has been hijacked by the general public to shortcut the tried and true process of eating less and exercising more to lose weight. Sure, AI can be employed to cut corners taking an easy route to a rote task, but there is always accountability—whether it’s the as-yet-unforeseen side effects of a misused diabetes drug or just what happens when you cheat on your homework (or work-work): you perform less well, and dumber.

So, do you believe the above was produced by me, a machine, or a combination of the two? And, most importantly, does it really matter? (If you want to, you can let me know your opinion at [email protected].) In the end, you can blame us both: me and ChatGPT.


Please enter your comment!
Please enter your name here