A.I. isn't scary. Your boss is.
Human decision-makers are the ones trying to replace artists and writers.
Did you know that you’re required by law to have a strongly held opinion about Generative AI? It’s true! If you don’t take a position on ChatGPT, you get sent directly to maximum security prison—no take-backsies.
Or maybe it just seems that way? I’ve certainly seen a lot of opinions recently: from friends, prominent artists, writers, tech guys, even politicians. My primary takeaway is that everyone has extremely strong feelings, and no one knows what the hell they’re talking about. This unfortunate combination leads to madness like the National Novel Writing Month’s recent proclamation that opposing AI-written books is classist and ableist.
Huh? What?? Why???
nobody:
literally nobody:
NaNoWriMo: if you don’t like AI then you hate disabled people
Thankfully, we’re finally getting some thoughtful perspectives, too. My feeling has long been that conversations around AI are too fear-based to be useful, at least when it comes to art.1
On the one hand, I get it. We’ve been trained to think of Artificial Intelligence as evil. Artificial Intelligence might send a buff Austrian dude back in time to kill you, and then that dude might become governor of California. Chilling stuff.
But AI will not replace artists. It’s impossible. That’s what I think, and it’s nice to see other, smarter people agree. Here’s noted sci-fi author Ted Chiang, who wrote a piece for The New Yorker entitled “Why A.I. Isn’t Going to Make Art.”
Art is notoriously hard to define, and so are the differences between good art and bad art. But let me offer a generalization: art is something that results from making a lot of choices. … By comparison, a person using a text-to-image program like DALL-E enters a prompt such as “A knight in a suit of armor fights a fire-breathing dragon,” and lets the program do the rest.
Chiang argues that prompting the creation of AI-generated images or words is fundamentally not a complex enough act to be considered art in and of itself.
In essence, [AI companies] are saying that art can be all inspiration and no perspiration—but these things cannot be easily separated. I’m not saying that art has to involve tedium. What I’m saying is that art requires making choices at every scale; the countless small-scale choices made during implementation are just as important to the final product as the few large-scale choices made during the conception. It is a mistake to equate “large-scale” with “important” when it comes to the choices made when creating art; the interrelationship between the large scale and the small scale is where the artistry lies.
This gets at a thing that drives me crazy about reactionary artist takes on AI: they’re devaluing themselves! To believe that they can be effectively replaced by ChatGPT or Midjourney means they are massively oversimplifying their own work!
Maybe I’m foolish or delusional, but I fundamentally believe that even if AI could perfectly mimic my writing skills, it still would not be art. Creating actual interesting writing2 is an inherently human act that requires human perspective, so in my view, real artistic work is not under threat from a Large Language Model.
I want to be clear that I absolutely do NOT consider myself an AI evangelist—which is the unfortunate group of tech shysters, internet weirdos, and international scammers that you get lumped into by admitting you’ve used AI. And full disclosure: I have.
(Wait! Please don’t unsubscribe!)
I’ve used Midjourney to help make graphics for Chortle posts. I also used it extensively in the production of an experimental animated short film I made in my garage last year. I’ve primarily found AI helpful in creating isolated visual pieces that I then overlay in Photoshop or Premiere to approximate the actual thing I wanted to make. Sort of like a custom stock image gallery or the ultimate collage source.
Some of the things that AI can do right now are legitimately amazing (photo upscaling, for instance). Some are useful but entirely unartistic (writing cover letters). But it has limited utility if you’re earnestly trying to make art that doesn’t stink. It can give you a running start, but currently requires a lot of elbow grease to become anything interesting, as Chiang’s New Yorker article also alludes to.
The film director Bennett Miller has used DALL-E 2 to generate some very striking images that have been exhibited at the Gagosian gallery … He generated more than a hundred thousand images to arrive at the twenty images in the exhibit.
One of the things that AI image generators can do well is replicate some of the surface elements of art. Likewise, LLMs can spit out a fair imitation of real writing. To put it another way, AI roughly mimics an artist’s craft—like a guy selling Banksy reproductions in Venice Beach. This is the AI ability that scares creative people I know, especially those who work in entertainment. But I think it’s important to distinguish between losing employment and being replaced as artists.
Job loss is real. Recent human history shows that when crafts can become automated… they are! That’s why shirts are now made in factories instead of by Giuseppe the Tailor, and why we get loaves of bread from Dave’s Killer Food Conglomerate instead of a local bakery.
In fact, I’ve lost work to AI. I used to have a decent side hustle writing copy for digital marketing agencies; now, they use AI instead. That makes my life harder, but I certainly don’t feel a sense of artistic loss and I can’t even muster the energy to be outraged. ChatGPT is probably more dutiful about the gig than I was.
Either way, ChatGPT wasn’t the one who decided to stop hiring me. It was humans! They also (perhaps unknowingly) accepted a decrease in quality, just like a shirt from Old Navy is unlikely to fit as well as one hand-sewn by dear old Giuseppe. In that sense, factory-made shirts are the same level of theoretical evil as an AI-generated commercial, but I don’t see anyone up in arms about sewing machines. (At least, not anymore.)
This is at least in part because we understand that shirt-making is not something as essentially human as creating art. It’s a job. But so are the positions that artists are scared of losing: from screenwriters to illustrators to editors to anything else that’s part of mainstream media. These are jobs.
It really, really sucks to lose your job. I’ve lost many of them! But it’s disingenuous to use moral outrage about AI as a cudgel when you’re mainly concerned about your personal employment prospects. I also think it’s shortsighted.
Almost everyone in Hollywood hates the very jobs that they’re worried about AI stealing. The vast majority of entertainment workers are overworked, underpaid, and feel taken advantage of by a broken industry. (I’d wager the same is true of people in media and other artist-adjacent fields.) Some of the same folks who think these industries cannot be fixed are simultaneously fighting to protect jobs that flat-out abuse them.
I’m not sure it’s worth all the effort!
It’s important to note that in these jobs, even the best of them, you are not generally employed to make art. Applying your craft for profit is a cool thing to be able to do, but let’s not confuse writing episodes of Paw Patrol with the Harlem Renaissance. Losing that kind of gig to a robot is an inconvenience, not the death of any actual artistry.
The truth is that opportunities to make money in actual artist-friendly environments are virtually nonexistent. This is the ultimate reason I think it’s important for us to move beyond reactionary rage about AI. I’m being idealistic here, but I think there is some possibility that emerging AI tech can help artists divorce ourselves from the fractured industries that we spend all day pursuing work in and then all night complaining about. At the very least, it’s worth being open-minded about.
Like many artists/writers/whatever-I-ams, I’ve been employed by a wide variety of companies with many different goals, sizes, and structures. When I reflect back, most of those places took advantage of their workers in one way or another. Sometimes they were nice about it and sometimes it was inadvertent, but it sure happened a lot regardless.
The companies I’ve worked for that didn’t do that—the ones that obviously prioritized fairness and worker well-being—have tended to be small and led by women. And here’s a shocking fact: small companies led by women don’t generally get to make projects on the scale of, say, Disney or Warner Bros. But if artists can develop responsible AI-powered tools that speed up our processes without sacrificing the quality or integrity of our work, maybe it becomes possible to make a studio-quality movie for $10 million instead of $100 million.
That’s definitely what studios are hoping! But in reality, it would be a double-edged sword. It would give artists power because we’re the ones good at making stuff. And the cheaper something is to make, the easier it is for us to make it without them.
My hope is that AI eventually develops to a point where it can help independent artists produce projects with creative merit and financial upside. To me, that seems like our best hope of getting out from under the thumbs of the David Zaslavs and Ted Sarandii of the world—you know, the guys who generally make our jobs miserable to begin with.
I feel supported in this quasi-optimistic view when I read pieces like Chiang’s, even if we don’t draw exactly the same conclusions. Whereas he dismisses the potential of AI even as a tool, I’ve come to see it more like Matteo Wong in The Atlantic:
Of course an algorithm won’t capture our minds’ and bodies’ expressive intent and subjectivity—one is built from silicon, zeroes, and ones; the others, from organic elements and hundreds of millions of years of evolution. It should be as obvious that AI models, in turn, can do all sorts of things our brains can’t.
That distinction is an exciting, not damning, feature of generative AI.
Preeminent Substacker and cultural sage
also seems to be thinking along the same lines I am. Among his thoughts in a recent post entitled Eleven Predictions: Here's What AI Does Next, he wrote:Stop worrying about AI taking over. It’s the people who own the AI who pose the biggest threat.
And:
The actual winners will be holistic thinkers and empathetic individuals with human skills.
I’m keeping my fingers crossed. Artists are in the midst of a 5,000-year losing streak to business people, so it’s about time things turn around!
What do you think? Am I off-base here? I’d genuinely like to know!
There are legitimate conversations to be had over how to regulate AI training materials and energy use, but first we have to move past shouting AI BAD! at every passerby.
We can argue the point of whether my writing qualifies for this description at a later date.
Thanks for writing this, Greg!
Thoughtful and balanced. The shirt analogy worked for me, though now I'm wondering if Giuseppe is still around. I'd like to know what a properly fitting shirt feels like. Apparently my books are among those being used to 'teach' AI, and I can't decide if I dislike that or if it's like, adorable that it's trying so hard. I imagine a sort of giant virtual mashing machine where my books are tossed in at the same time as Yuval N Harari's and Malcom Gladwell's making for a more depressed analytical detective who discovers they will need to practice detection for 10,000 hours before they're any good. Of course all those books have gone into the hopper that is the mashing machine of my own brain...It's messier probably, but I bet I could still beat AI at my own game.