“The good writers touch life often. The mediocre ones run a quick hand over her. The bad ones rape her and leave her for the flies.”

~ Ray Bradbury (Fahrenheit 451)

AI-generated art is tricky. I'm of the opinion that it is indeed totally worthless if generated by a mindless, mouth-frothing AI bro, no matter how elaborate or impressive the piece is. However, like all mediums, it does at the same time provide a channel for true expression.

For this post I'll only be considering the abstract process of art and its meaning, more than talking about the output itself. Also, I won't be looking at the very real question of whether it is even ethical to use AI image generation models in the first place1.

...

Many may indeed say that AI-generated images are always soulless, and they have a point. The great majority of generated images circulating around, while often visually impressive, are clearly lacking that human element. We're all familiar with the term slop.

I would argue that that lack of humanity has two sources: one is the generative aspect itself, and the other is that the person creating the images is chasing the wrong thing.

Instagram is riddled with accounts of "AI art" that aim to generate handsome people or funny pictures that have no real goal of expression. Their whole reason for being is adding content, attracting users. Operating under the fallacy that content will pull views. Here, the user is the problem. Its motivations and goal for doing "art" are directly pissing in the face of the art2 they proclaim to do.

The other aspect, that these images are generated, has some weight to it, but I think it's less relevant than would initially seem3. Even before modern image generation tools like OpenAI's DALL-E and Google's Nano Banana, we had lots of really amazing generative art in the form of traditional algorithmic visualizations.

Of course, it's not entirely the same because someone actually wrote those visualization algorithms. Someone took the time to work on them and used their own creativity to come up with the end result. Even if that result was itself machine-generated, it took effort, and often skin and sweat, to get it to work appropriately4. But is that the case for all generative art? I would say no. Much of it does require effort, but some of it is just a matter of tweaking parameters until the output is aesthetically pleasant.

Now, the next idea I wanted to touch on is the concept of expression. Art is, by definition, the way in which we humans express those things inside us. It's a channel for those tensions within us, a medium through which we can work on those hidden energies. I think that in this respect, all modes of expression (modern generative art included) are valid.

Interestingly, this idea of expression applies to both the artist (the one through whose agency the art comes into being) and the beholder (the one who experiences it after the fact), but it works differently in both cases. For the artist, the expression is the source of the "art," the fertile soil that gives it birth. For the beholder, the expression is more a resonance with a specific piece, the witnessing of which brings out something in them. Both the artist and the beholder do art with a given piece; one does it during the creation process, and the other does it through the experiencing of it.

But can a "creator" actually be both the agent behind the creation and a beholder thereof? I think generative art specifically allows for this arrangement of things, since the end result is not guided by the creator from end to end.

I actually have a great example of this. Around one year ago (end of 2024, I think) I wasn't in a very good place mental-health-wise. I was quite depressed, actually. I felt worthless, an encumbrance upon the world. To any who have felt similarly, the analogy of "feeling like the lowliest of worms" might sound like a good description of what it's really like5.

One day, a thought bloomed from my misery, and I felt the urge to somehow connect with what I was feeling. This is a common inspiration for many of us, which sometimes manifests as a poem, sometimes as a thought, sometimes as a drawing. In this specific case, it manifested as an idea for a painting. I knew I wanted to paint my feelings. I wanted me, a worm, being laughed at by others.

It might sound a bit grim and twisted as a desire, to want to picture oneself in such a way, but it's actually quite the opposite. I'll talk more about that in a second, but first, here's the image DALL-E generated after I put in my prompt idea. It did it in one go; I had no need to refine it, as it struck exactly the chord I was aiming for.

Pasted image 20260401095054.jpg

I immediately identified with this poor worm being laughed at by the incomprehensible, adapted beings of an insane society. Seeing my feelings represented out there made them separate from me. They were no longer something inside me that only existed in my head; now there was an external representation, a symbol, that projected them into the external world. I can't exactly describe the transformation inside me, but it did help. I was able to relate to it in a whole new way and, more importantly, I saw the humor in it all.

This idea of representing what I'm feeling has been really helpful in many other situations as well. More often than not, that representation comes through writing, which is my main "creative" activity6, but sometimes what I feel requires other mediums.

Thoughts:

  • I've been sitting on this post for a while, but it somehow just wasn't flowing. I think perhaps I was just trying too hard.
  • I've also been playing a lot of Promise Mascot Agency lately, which explains why I haven't been posting as much 🤗 great game! Very wholesome.

Footnotes

  1. This not only can take up a couple of posts all on its own, but the decision will also render this current post moot. ↩

  2. To make things worse, a huge number of these accounts are themselves managed by an AI that analyzes "trends" and automatically comes up with potential images to create, raising the slop-bar to stratospheric levels. ↩

  3. Again, just roll with it. I'm purposefully ignoring the fact that AI image generators are trained on the hard work of real people who often didn't give their consent. ↩

  4. Tangentially related, but this reminds me of the awesome world of tweetcarts! Basically, the goal is to write a cool graphic demo for PICO-8 that fits into the limits of the traditional tweet. I really have no idea how they do it; it's so cool. ↩

  5. The feeling is also excellently-albeit a bit too verbosely for my taste-transmitted in Kafka's Metamorphosis. ↩

  6. Just a thought here: I've always thought of "programming" (at least traditional programming, without AI) as an extremely creative endeavor. However, now I'm realizing that programming really doesn't have any expression of how I feel (at least not always), so it might be missing that critical component to elevate it into "art"? ↩