AI Can’t Be Beautiful — At Least, Not Yet

classic bust of a woman surrounded by digital mapping

Getty Images

Yes, AI tools are the future — but will this tech ruin art?

Well, to all my fellow Gen Xers, it’s finally happening — we’re being confronted with two persistent threats that have plagued us since the ’70s.

The first is that machines are almost smart enough to replace every single one of us without exception. It’s not just factory workers and chess players who have something to fear. It’s artists, musicians, actors, writers, surgeons, lawyers, and toll booth operators.

AI-generated art is the latest iteration of this, but it’s been coming for a long time. There’s even a term for it, which is as horrific as you might expect: AI singularity.

From what I can tell, AI singularity is still conceptual. It refers to the moment that machines become smarter than us. There’s a company called Translated that’s attempting to quantify AI singularity by measuring the accuracy of AI-generated language translation. According to them, AI is quickly closing the gap between what expert human translators can achieve, versus optimized machine translation.

How close are we to reaching singularity? Well, Translated says we have seven years until AI can translate speech as well as a human translator.

I guess this means we still have time to unplug everything and pretend we never invented big data, Google, and algorithms that know us better than we know ourselves.

Let’s be honest. It’s far too late to stop what’s coming. We’ve been tethered to the Internet for two decades, feeding it a bottomless diet of our words, pictures, music, video, and art. It won’t be long until the machines are better at being human than we are or (more likely) pretending to be human. This brings me to the second threat: The robot uprising.

We were warned about this eventuality in countless movies. First, we had Hal, the sentient computer from 2001: A Space Odyssey. Hal starts out as a kind, helpful AI system, until he goes mad and murders the humans he’s supposed to protect.

That movie was released a few years before I was born. I watched it several times as a child. I didn’t understand much at the time, but that’s the first example I had of AI, and it imprinted itself on my brain like a baby dinosaur from Jurassic Park.

In the early 80’s, Hollywood gave us murderous robots in both computer and human form. Gen X watched a computer named Joshua nearly start the third world war in a movie called Wargames. A year later, Arnold Schwarzenegger, cast as a buff humanoid robot, relentlessly stalked a terrified Linda Hamilton in The Terminator.

The concept of autonomous robots felt very fictional and far away to 13-year-old me, but with Wargames and The Terminator, the seed of fear Hal had planted grew a bit more.

Two years later, a movie called Short Circuit told the story of a robot that was struck by lightening which made it sentient. This was a cute one for the kids. Robots are cuddly! They’re harmless! But they could also (gulp) assert their right to live. That was the whole point of the movie.

By 1999, we had The Matrix and my generation’s visceral fear of intelligent machines was fully realized. The scene where a bald (yet beautiful) Keanu Reeves wakes up tethered to a machine, his body and mind nothing more than fuel for the robot collective, still haunts me.

And now? We have ChatGPT.

Look, I’m, not here to bash technology. I write about it for a living, after all. Tech lets companies scale! It eliminates repetitive tasks! It makes everything more efficient and effective!

AI-driven tools also help me do my job. I use Otter, an AI-powered transcription tool, to record and transcribe my meetings. Discovering Otter was nothing short of life-changing for me. I don’t need to take notes when I interview people — I simply let Otter do that work for me.

I also use Quetext, a tool that analyzes and flags text for plagiarism. Quetext helps me avoid simply rephrasing something I’ve read (e.g., accidental plagiarism). I also use it to identify (and eliminate) overused phrases and cliches. When I’m tired, it’s really tempting to write terrible things like “low-hanging fruit” and “helps you grow at scale” instead of buckling down and being more creative.

So, no, I’m not here to broadly condemn all AI-driven technology  —  just the stuff that’s making fake art. I don’t think AI-generated content and art is 100 percent bad. But, like all things robotic, art made by machines feels kind of…threatening.

I am worried, but not about the robot uprising (at least, not yet.) I’m more worried that the content created by ChatGPT, Dall-E, and similar tools will quickly overwhelm real-people art. I’m worried that it will fill our digital and physical spaces and make it harder for artists to be seen and heard and honored.

I’m worried that OpenAI and machine learning and natural language processing — the stuff that makes robot brains tick — lessens the act of creating. It makes it all too easy. This means it’s ripe for abuse and commoditization. I mean, art is already being commoditized, but at least some of us are being paid fairly for the art we create.

I’m worried that, soon, some of the stuff being cranked out by AI tools will become indistinguishable from human-generated content. AI-generated writing, in particular, feels like it’s close to reaching that ominous singularity. When that happens, will machines become poets? Will they write novels? Will they pen the next Pulitzer Prize-winning essay?

How can machines do any of these things?

Right now, I don’t think that a machine can create something beautiful without the help of a human on the other end of the screen. What that means to me is that these AI tools, in all their iterations, are just tools.

We can never be the things a machine can be — immortal, all-knowing, perfect. So I’m not even going to try. I’m not giving up. I’m just recognizing that, as a human being, I have my limitations.

But it’s also true that machines can never be the things humans can be. They can mimic us, like parrots, throwing our words and images back at us in increasingly clever ways. But the true beauty that comes with creating art eludes them — at least for now.

I asked ChatGPT what it thought about this and it agreed wholeheartedly, writing, “Humans have an innate creative capacity that machines simply cannot replicate. Machines may be able to mimic our words and images, but they lack the spirit, emotion and spontaneity that characterize art created by humans. It is this unique quality that sets us apart from machines and allows us to express ourselves in ways that are both meaningful and beautiful.”

Okay, now I’m terrified.

The final paragraph of this piece is AI-generated content.

Reprinted with permission from the author; follow her at Medium