On Art — Advanced Old School Revival

Table of Contents Jump to Page in Support

I’ve always been interested in practice rather than theory, despite my limited academic background.

I went to grad school for publishing, in part because theory never interested me all that much. After college (I minored in poetry and digital art), I wanted to know about the process that goes into editing and distributing books, because books are real end products, people read them, and in reading them they become inspired. But I got sidetracked into a career in web development shortly after finishing grad school and ended up in digital advertising. (Unfortunately for my soul? Fortunately for my now-paid off grad school debt? I don’t know.)

This is all to say that I find gatekeeping about “What is art?” to be pretentious navel-gazing that’s actively counterproductive to producing art.

I can draw still life, and on occasion I’ll do it because I find the exercise meditative. I can also do visual design work (i.e., the visual design part of website design), but for me it always comes back to the question Where do you want to spend your limited time? I only have so many hours in a day and so many days to live, so there’s a certain desire to optimize my hobbies to get the most productivity out of them as possible. (This is likely an unhealthy philosophy, I know.)

Anyway, back to art. Art in the context of AI.

A chief criticism of genAI (when it comes to image and video generation) is that AI-generated “art” is “slop.” And slop is not art, it’s soulless and lacks cultural value, we shouldn’t make it, because all it does is add more trash to the trash heap that is the world.

All AI "Art" is Slop

Some memories come to mind when I think about the word “slop.”

I think back to my 8th grade English class to a teacher I had named Ms. Veinotte, who psychologically tortured me into writing better essays. It’s not worth getting into that, but know that I hold a special loving place in my heart for Ms. Veinotte, an intimidating ex-nun, because without her, my command of the English language would be much worse than it is. We had to write short stories one day, and she shared mine with some other writing instructor. I don’t think she knew I could hear her (or maybe this was part of the psychological torture), but she remarked, “It’s a shame that this is genre writing. If only he’d write about real things.” That remark stuck with me for years. It was my first encounter with someone whom I respected characterizing my creative output as “slop,” because it didn’t conform to their preference of genre.

I also recall being an intern at Da Capo, a press in Boston, the sort of position you hold at the bottom rung in the publishing industry in the hope that some editor might notice you and lift you out of obscurity into the realm of real book workers—you know, the people who memorize ALT-0-1-5-1 to produce an em-dash, and put the rubber stamp on what is literature. One day an editor came by and noticed me rifling through the slush pile, and we struck up a conversation. “What do you most like to read?” she asked. I answered: “Well, a lot of different things, but I’m especially fond of genre.” The conversation ended immediately, and she walked off, never to acknowledge my existence again. And so history repeats itself.

Slop is a shadow on the spirit, metaphorically speaking.

It can blacken everything out, if you let it. In the tail end of my manuscript-writing days, one other memorable conversation was when one of my colleagues in our writing group (you know, the kind of confab where everyone is working on their “manuscript” and sharing excerpts, such that the burden of loneliness when it comes to submitting said manuscript to the ultimate gatekeepers of all—literary agents—is shouldered by everyone in the group) had some feedback for me about my manuscript, which was an earnest but unpolished attempt at writing a post-cyberpunk story about the anxiety of millennial home ownership. She said, with a hint of unwitting condescension: “Well actually, this is really complicated and the terminology escapes me because I don’t read scifi. The world in it has a lot of rules. I wonder, maybe you should just stick with making video games instead of novels?”

I’ve never made a video game before, but now I make roleplaying games.

So, maybe she had a point.

This tangent is a long winded way to deconstruct the word “slop.”

Slop is the stuff we think has no artistic merit. Things that lack artistic merit are the stuff we think we shouldn’t waste our time on; such stuff does not enrich our soul. Can we agree on that much, even if I don’t agree we have souls to begin with?

Hollywood produces a lot of slop, using a lot of resources (which has a larger impact on the environment than AI generation does on the whole, as we have already established). So does genre writing, and even that highest regarded genre, literary fiction. The Internet is daily inundated with slop, of the social kind. Memes. Posts about food. Reddit, in its entirety. YouTube. Tweets.

And now, AI videos from Sora, too. Hundreds of millions of them. And one-click image generations from Midjourney, with that distinctive "Midjourney look." Slop, slop, slop, everywhere we look, trying to steal our eyeballs, a weaponization of the very advertising industry I sold my soul to in order to pay off my debt to higher education.

Out-of-Box, Midjourney has a characteristic style

A lot of talk about souls for an atheist, don’t you think?

I will get to the point.

I fully agree that the advent of genAI has radically accelerated our ability to produce slop. It’s like every idiot with a camera on Instagram, except now they don’t even need to take a picture. I get it. You’re right about that.

Some AI Art is Slop

But I don’t agree with the proclamation that “Everything AI produces is slop.” We must finesse that claim. First of all, AI doesn’t produce anything by itself in this context.

The AI has a human operator behind it (at least for now). The human operator is the slop producer, typing a few words into a prompt generator like Midjourney, with no coherent vision of what they want. Creative responsibility is a property of agents, even if creativity itself is not. We should blame all the slop producers sitting on their couches generating videos of cats wielding machine guns on porches, not the tool that does the generation. And that distinction is important, because it allows room for us to envision AI output that is not slop, where there was a degree of creative vision involved in its use.

So what is art then? (It’s a stupid question, is what it is.)

Many will argue that a creation requires human authorship or human intention to be considered "art." But I believe it is possible to evaluate some thing as art (which we have said earlier means it is culturally valuable to us, at a minimum) even if the thing was produced by a rainstorm, or produced by a monkey accidentally, or produced by smarter monkeys like DuChamp or John Cage or Damien Hirst (respectively, that famous urinal, or The Physical Impossibility of Death in the Mind of Someone Living or 4’33’). Is a bullshit intention “art”? I saw a lot of bullshit justified as “art” in my art classes in college. Little prompts written on white paper to describe things they created that required no talent whatsoever.

I wrote a lot of bullshit essays myself in college, as well.

*Death Denied* (2008), exhibited in Kyiv, via Wikipedia

No, I think bullshit can be art.

Let me circle back to those cats wielding machine guns for a moment. I actually thought those were funny the first time I saw them. If that slop made me laugh, well then isn’t it comedy? But comedy is art, so that’s a problem for us. Apparently, art has something to do with how the audience receives it, too.

(It’s not a stupid question, I’m sorry for writing that.)

But I think to argue that this or that thing “is not art” is a sophist claim, a misleading one. I admire as art any thing that communicates something novel to me, whether that’s a feeling or a story. I admire the technical prowess involved in the creation of the thing, but technical prowess is not necessary for me to qualify it as art. I admire the creative vision that was required to conceive of art, if I’m privy to that vision. But those are just my preferences. Even contemporary art, which I despise, is art in my mind, without qualification. DeviantArt slop is art. Anime is art, and I hate anime.

If I can be big enough to admit these things, maybe you can too, with regard to things created by AI tools?

The Worlds of OSR+

Our campaign settings in OSR+ are little fictive encapsulations, kind of like how PbtA games try to emulate a specific genre.

The point of playing in a World of OSR+ is that you play characters who fulfill certain narrative roles that make sense for the sort of world being portrayed. For example, in Guardians of the Glade, you play characters who persist in a world that’s recovering from environmental disaster. It’s solarpunk. The writing and the themes are dark, but ultimately hopeful. You prevail by creating, not by destroying.

A generation created in SDXL and then enhanced with Midjourney

There are lots of components that go into making a World of OSR+, from the art I generate in order to communicate the solarpunk vibe, to the writing that sets the tone for the adventure, to the game mechanics that reinforce the sort of play that is emblematic of the genre. I don’t use an LLM to do any of the writing (though I won’t begrudge anyone who does), instead I use it to brainstorm, kind of like conscripting a writing partner who won’t shit on you for being a genre writer. This use for ideation is relatively new, as LLMs like Gemini have gotten better at creative thinking (although they are still bad at it).

As an aside: I'm not privileging writing over visual mediums when I say that I don't use LLMs for creative writing. LLMs are just not very good at creative writing right now. I'm sure that will change with time, and then I'll be able to see what genAI adds to this synthesis.

The Whole is Greater Than The Sum of Its Parts

A World of OSR+ is a way to communicate an experience, much like what happens when you read a book or a poem, or watch a movie. It just takes a lot longer, because your enactment of the adventure is part of that experience-generation. The World as a product just happens to take the form of a printed book, optionally extended by online tools such as those on this very website.

So the artwork, then, is a small but important component of the whole thing.

In my mind, it doesn’t matter how I created the artwork. Whether I spent 300 hours drawing and painting all the artwork by hand—or spent 8 hours collecting reference material to generate a LoRA; an hour testing hundreds of image outputs; a few minutes crafting a prompt template that makes use of various art terminology and artist tokens as a shorthand for artistic styles (living and dead); and then a few hours tweaking choice outputs in Photoshop—should be immaterial to your experience, unless your very specific definition of art is “everything that is made by hand” or even more specifically, "not made by AI.”

If this is your very specific definition, it sounds like a personal problem that’s limiting your ability to experience art.

I would advise you to consider broadening your definition.

AI Art (2024 - 2025)

Here is some genAI-assisted art that I think is not slop, to get you started.

← On the Environment

Page 11 of 18

On Ownership →