Great images and stories are something akin to magic. Like lightning from a clear sky, they stun us with their beauty and power. And they inspire subsequent artists to create, spawning genres and subgenres.

Wholly original art is not only exceedingly rare but may not actually exist. Most art is derivative, original only by matter of degree.

And that’s not a bad thing. Until the computers started doing it.

Before we get to the latest seismic shock impacting publishing and associated creative professions, we need to talk about how art is made.

Art derives from somewhere; no one is born into an isolation chamber, raised without human input, and creates from a blank canvas. In short, we observe, then we create. We read, we watch, we listen, and learn from others, and are inspired to create works of our own.

All art bears some stamp of its sources, even the greatest. The Lord of the Rings did not spring whole and entire from the forehead of its author; J.R.R. Tolkien drew from myth and epic, his work as a philologist deriving the root of words, and from contemporary sources like H. Rider Haggard’s She. Nearer to home, Robert E. Howard read the likes of Talbot Mundy and Harold Lamb in the pages of Adventure, and the full-blooded adventure stories of Jack London, blending these with weird horror to create figures like Kull of Atlantis and Conan of Cimmeria in the pages of Weird Tales.

Sword-and-sorcery is a subgenre of influence, to varying degrees. The various “Clonans”—Brak, Thongor, Kothar, and their ilk—bear the obvious influence of Robert E. Howard’s famous creation. Michael Moorcock read and was influenced by Howard, but with Elric created a protagonist wholly different than Conan or Kull—civilized, decadent, his strength dependent on a cocktail of drugs.

We also consume and enjoy derivative products. I like my Kothar books—they’re fun (no, they’re not, they’re terrible—DMR), and sometimes they’re just what I need (no, they’re not, they’re terrible!—DMR). The Clonans sold, but creatively arrested the subgenre in the ‘60s and ‘70s. S&S became so narrowly focused and artistically stultified that it suffered a collapse in the early 1980s.

It’s important we recognize true innovators and mavericks. And offer venues and support creators that strive for originality.

Today the issue of originality vs. derivative work has risen to greater prominence and debate with the rise of AI. Now our art is threatened not by derivative authors, but by machines capable of turning it out in choking quantities.

ChatGPT and its ilk take prompts placed by human beings to produce machine writing. Much of it is pedestrian, some of it is awful, or worse, contains incorrect information and unverified references with a veneer of authority (very dangerous, particularly in the medical fields). But some of it is good, and everyone agrees it will get better. And it already is, with the emergence of ChatGPT-4. Even though the art it produces is entirely derivative—it draws upon huge databases of information, the information on the world’s largest ever source of information, the internet—and “creates” from that—we’re seeing a future where machine-made art is not only possible, but on our doorstep.

Should this make us feel threatened? If human art is also made by the same process, isn’t this just an advancement? More efficiency and scale, a lower barrier to art entry, and more art to enjoy?

After all, sequels sell. The Marvel universe and the Star Wars universe and all their movies and TV shows and spin-offs (that all start to feel very much like each other), sell. Conan, too, sells, whether he is the subject of computer games, tabletop RPGs, movies, or new novels like S.M. Stirling’s Blood of the Serpent. Even though his creator is long dead, Conan continues to fuel the sword-and-sorcery economy.

On a recent episode of his podcast, host Ezra Klein made a point I agree with, though we might not want to admit it: AI including ChatGPT might create derivative work, but most human beings do too. “And a tremendous amount of the creative economy, and I don’t say this with any sense of judgement, certainly there’s nothing wrong with is, is work that is derivative,” Klein says.

Klein thinks producing and consuming derivative art is necessary for a functioning economy. Again from his show: “The secret of it, the thing I think people don’t really want to face up to is, people like things with familiarity to it. They like a certain amount of derivativeness in their work,” Klein says. “So, of course there is great art being done, that is truly new, truly non-derivative, truly made with meaning as the central intention. But so much of human creation is not about that. And so many of the people who are creators, that is not what they are doing.”

This rings true for me, because it describes what I did when I wrote Flame and Crimson: A History of Sword-and-Sorcery. More than fiction, my book is in a sense entirely derivative. I read and studied other people’s works—S&S fiction, but also other critical studies, old magazines, essays—and passed this glut through “me,” and on the other side spat out an “original” work.

You could say the information I read was the prompts, Flame and Crimson the ChatGPT output.

I even went so far as to identify seven broad criteria for what constitutes an S&S story. I had some trepidation when I did so; even before the rise of ChatGPT I thought it might cause some authors to use my book as a checklist for their story, and create that something that felt formulaic, mechanical.

Today, nothing stops you from prompting ChatGPT with my criteria and asking a computer to write a sword-and-sorcery story from it (for the record I have not tried this, and don’t want to).

Now, I think I had a few original ideas in my book. I had to organize its information, decide as much of what I was going to leave out as put in, which required some degree of critical thinking. But in the end, how does process this differ from a machine? Isn’t this what AI does? Can it do it better than me? And if it can, shouldn’t we celebrate that?

The makers of AI, and those pushing for “copyright” for AI produced works, would lead you to believe so.

But I believe there are differences between man and machine that should at least give us pause.

Until recently the human mind was considered sacrosanct and unapproachable. Some pointed to our mind as evidence of God, but the less devout acknowledged that our brains were uniquely capable of creating art. No animal creates art under their own volition; art is part of what makes us, us.

And it seems to me humans do create art differently than machines, including advanced machines like ChatGPT. We filter what we consume through a biological state of being. We have our own set of unique, lived experiences that are not (entirely) on the net. We take all of this and produce something original, because it has passed through our flesh. It is still derivative, but also somewhat unique, and our own.

But with the advent of sophisticated computing that gulf of difference has narrowed, and with the rise of ever-more sophisticated artificial intelligence today we wonder: Is there a difference between man and machine? What happens if some chat interface can pass the Turing Test?

What happens if an AI-generated story is good? Can we read and enjoy it? Should we?

I say, no. This gets to the why of sword-and-sorcery.

As I wrote in Flame and Crimson, few fates are more enervating than an inability to change one’s circumstances, being a cog in an uncaring mechanical universe. Sword-and-sorcery rages against the machine. It offers the modern reader an alternative, posing the question: Which is the superior? The old barbaric ways, in which life was nasty, brutish, and short, but men and women were at least masters of their own fate, operating with a heightened awareness borne of daily struggles for survival? Or the modern industrialized state, in which we accept security and civility and predictability in exchange for hyper-specialized, disconnected jobs, purposelessness, and widespread anxiety and depression?

Without art, we cede a huge piece of why we exist, and stoke the fires of purposelessness.

People, not computers, need to write these stories. In sword-and-sorcery a warrior with a sharp sword and an iron will is exalted. The stories they tell are anthropocentric. They reject pessimism, cynicism, and powerlessness.

So should we, as a community.

We need to set boundaries over what we allow, and what we consume. There is a reason we enjoy live concerts, and refuse to settle for lip synching artists playing recorded music. We are paying watch talented people perform, not listen to machines.

If a computer had written Flame and Crimson, it would have deprived me--robbed me, of something very personal, and meaningful.

Let’s not surrender this space.

If you’re an editor, this is your chance to be a sword-and-sorcery hero. Develop and publish clear guidelines about the use of AI. Prohibit “authors” from submitting AI generated stories or images; if they violate this, ban them.

This is also a raison d'être for authors, a call to arms to keep innovating. To push boundaries, reject paint by number formulas, and extend borders. To use your own unique experiences, and rages, and loves. Just as Howard did, blending literary influence against the backdrop of towns ravaged by the boom-and-bust of oil speculation, and the vanishing West of his Texas landscape, to create something startlingly new and vibrant.

Strive for originality. Expand the conception of what S&S can be.

And stay one step ahead of the AI.

Brian Murphy is the author of Flame and Crimson: A History of Sword-and-Sorcery (Pulp Hero Press, 2020). Learn more about his life and work on his website, The Silver Key.