A Human Artist's Defense of AI Art

Why I was wrong

AI art is good and I was wrong to condemn it.

But maybe I should back up.

I am a fulltime artist. I draw every day. I then turn around and sell my drawings, both to publications like The New Yorker and directly to my fans in the form of merch, like prints and tote bags. I also make books and write newsletters (like this one) that people generously pay for because they like what I do and want to support my ability to keep doing it. This is how I make my living. It’s precarious, but it works. And I love my job.

So, you can see why I might not have been thrilled at the introduction of generative artificial intelligence. Immediately, it was clear that this new technology posed a threat to all creatives. These models were not only cheap, they were fast. With a few keystrokes, anyone could produce, say, a New Yorker-style cartoon in mere seconds. On top of that, the reason these products could perform this miraculous feat was that they had been trained on the very artists whose jobs they were now threatening. For this reason, a new term emerged for anything that produced AI art: “theft machine.”

While the problem was clear, the solution was not. “The toothpaste is out of the tube” became a common refrain amongst techno-optimists and skeptics alike. Nevertheless, artists resisted. Some called for government regulation, a few took to the courts, others boycotted. I was one of these resisters and I enthusiastically joined my peers in the unifying message of: fuck this.

So, given everything I’ve just said, why am I writing an essay in defense of AI art? My answer is simple: the quality is now so undeniable, I have no choice. As they say, I know when I’m beat, and, frankly, I’m crying “uncle.”

Allow me to explain via a single case study. Five years ago, I made this cartoon that went viral:

The gag is simple, the drawing elegant. There’s really nothing I would change about it. Is what I thought at the time. What a fool I was. You see, a few weeks ago, it came to my attention that people on the internet had been feeding my work into AI engines like Midjourney. These “artists” were then sharing the new, AI-generated/altered images rather than my original work online. I scoffed, assuming this was nothing more than a cynical ploy to drive engagement without having to properly credit the original artist. But then I saw what they had made, and it. changed. everything. Here is an example:

At first glance, I just knew this version was better than my original. There’s just a certain je ne sais quoi to it. But then I started drilling down into the details and I found so many specific reasons for its superiority. I’ll enumerate them here:

1. The bargoers are all identical.

What an interesting layer! Rather than a story about a tavern full of strangers unified in their distaste for long songs, the scene is now a group of septuplets whose private family gathering is being rudely interrupted by a random bald baby man, which brings me to…

2. The bald baby man.

An undeniable upgrade, what I love most is that, unlike my mustachio’d hero, the bald baby man isn’t carrying cash with him, or even a credit card. This edit is fun because it means that, while he is in fact moving in the direction of the jukebox, he has no intention of selecting a song. Now the story is about an emotional mob attacking a total innocent. What social commentary! And, finally…

3. There’s a ghost.

I’m ashamed to admit it, but it didn’t even occur to me to put a weird spectral form in the background of mine. The presence of this ghost brother elevates this cartoon from merely a little joke to a full-ass horror movie. It also implies a stronger background story for the septuplets. Maybe the reason they’re gathered in the bar is to mourn the passing of their brother! This makes the bald baby man’s intrusion so much worse. It’s been heightened. That’s what we like to call in the business: “raising the stakes.”

Here’s another AI version of my cartoon:

Masterpiece. Three reasons:

1. The man isn’t actually going for the jukebox. He’s going for the “jukebax.”

What is a jukebax? That’s up to the viewer! This is theater of the absurd. Samuel Beckett would approve.

2) The expressions.

Unlike the boring neutral faces in mine, these bargoers are literally insane. They’re about to tear their fucking skin off their faces and descend into sheer madness like something out of Lovecraftian horror.

Of course, this isn’t true of all of them. This guy looks more like he’s a cartoon wolf who’s just seen a babe so hot his tongue is about to roll out of his mouth and hit the bar.

3) The jukebax-goer is saying the line himself.

Many people online were quick to point this out like it’s some kind of flaw, but the truth is it’s brilliant. For who is our greatest enemy if not ourselves? We have all made disastrous decisions while simultaneously panicking, helpless to stop ourselves. On top of that, you’ll notice the line is both in quotes and in a word bubble. Who is he quoting? Himself? The patrons? Perhaps his disapproving father. That subtle change has now transformed this single panel into an epic tale of generational trauma.

These are just two examples of many. For instance, there’s this AI-colorized version where the bar jumper has a mysterious third hand reaching out of his back:

And this fun video game version which replaces the patrons with demons from Doom, which is fun because they would want to kill the man regardless of where he was headed:

I could go on, but I will rest my case here.

The truth is: human-made art had a nice, long run. I understand you might still bristle at AI-generated content for other reasons, like its environmental impact or the aforementioned fact that it is literally theft. But as an artist, I have to admit that the quality is now super human. AI art, I give you three thumbs up.

P.S. If somehow you remain unconvinced and would like to support this human artist, please consider becoming a paid subscriber.