Like many offbeat ideas, this one originated in a bar.
I was drinking with a fellow author, and we were bemoaning the future of writing in a world of artificial intelligence. We nervously pondered how much of what we writers do could be replicated now or in the near future by a chatbot. We assured ourselves that our work—researching and investigating and then composing articles and books on complicated topics, which present fresh material, compelling context, and sharp analysis in (we hope) an engaging manner—cannot be supplanted by a machine. At least not yet. But what of other fare? Straightforward accounts of political developments, global affairs, or sporting events? Recipe books? Or histories and biographies? Short stories or novels? Film scripts?
What could AI produce? Generative artificial intelligence has already invaded—or infected—the world of fan fiction. I have friends who have used AI with their kids to create children’s books. Iranian meme-makers have deployed it to concoct slick Lego-style videos deriding Donald Trump.
The issue is not the ability of a chatbot to churn out algorithmically determined masterpieces. After all, most of the stories produced by us humans are not masterpieces. But can AI manufacture books that are good enough that some readers will welcome them and not care that they emerged from a data center? Perhaps even pay for such fare? Might many writers soon face an extinction-level event?
Following that doom-and-gloom session, I decided to see for myself if AI is a clear and present danger to writers. I asked Claude, Anthropic’s AI chatbot, to generate a novel. It did, and I’ll cut to the chase: The novel isn’t so bad.
Let me note that this exercise deliberately sidesteps the many profound concerns about AI: the economic dislocation it could cause, the environmental impacts and community costs of data centers, the expropriation of work produced by authors, journalists, and other humans. It focuses only on what AI might mean for the future of writing and other creative endeavors. And don’t forget that the reason an AI chatbot can kick out a novel is because its Big Tech master—without seeking permission or offering compensation—hoovered up a gajillion books, articles, news stories, and other material created by people. (The Center for Investigative Reporting, the parent company of Mother Jones, has sued OpenAI for copyright violations. OpenAI denies the allegations.)
I kind of figured it wouldn’t be terrible.
Other than the obvious plagiarism allegations, I would wonder how consistent the novel would be. (Also, given the strict definition of the word novel, I don’t think it should be called that, since novel means new and unique.) But regarding consistency, does it keep a pace? (Not all writers do, but an AI should be able to maintain one.) Do things characters do later make sense based on what they did before, and do things they do early affect, influence, or steer what they do later?
I once asked ChatGPT — this was actually three years ago, so you know it’s gotten better since — what cast members from the anime Boku no Hero Academia would be if they were real people who were set in the world of Star Trek: The Next Generation. It actually got a few of the same things I guessed (I did the exercise myself first). Where we differed, it had good reasoning. I also spent a good hour thinking about this. It churned out its answer, in several paragraphs, in seconds.
Since someone’s going to ask, in short, its answers were Midoriya Izuku = Wesley Crusher; All Might = Captain Picard; Bakugou Katsuki = Worf; Uraraka Ochaco = Troi; Todoroki Shouto = Data; Iida Tenya = La Forge, Asui Tsuyu = Ro Laren; and Kirishima Eijiro = Barclay.
I just asked it, since it’s been over three years, and the software has changed, what, if anything, would it change? And it responded faster than before (based on my memory) and wrote way more, going into deep analysis. I was going to make this a spoiler… but nah, now it’s relevant. What’s interesting to me is the software’s ability to understand what makes a character who they are and how they fit into a story, and put that character into the other story.
What’s scary is the tone of the language. This is partly my fault since I talk to a chatbot how I expect a chatbot to expect a human to talk to it (naturally). So it says things like “This is one I really wish I’d thought of three years ago.” Does an AI chatbot/LLM think? This one seems to think it does, or it seems to think I want it to pretend that it can. It also has a favourite match. In another comparison, it says “I previously said,” followed by the matchup, followed by “I don’t buy that anymore.” What does an AI chatbot “buy”? I mean, I understand the term. It means to believe. It’s saying that it previously believed one assumption but now believes a different one. (To be fair, in context, the story of Boku no Hero Academia was not completed when I asked before, so now the chatbot has more data. However, unless it’s tapping into summaries on Wikia/Fandom or Reddit discussions, that “data” is drawings and hand-written text (manga, similar to comic books). Kind of wild for it to parse all of that nearly instantly.
I’ve gone off the rails a bit… like I said, I was gonna put all the last three paragraphs into a spoiler, but you can see how it’s relevant. Since you (OP) said fan fiction… I believe it’s capable. Also, you said Claude. I think Claude is a bit ahead of ChatGPT. ChatGPT is kind of like the “everyman” LLM, like Gemini and maybe Meta AI, but I feel like Claude is a step or two ahead. I’m not an expert on AI. I’ve been accused of being pro-AI and anti-AI by different people. What I actually am is curious… and cautious.
I’ve been tempted to run this experiment as well. I’d only ask for a novella, something around 150 pages long, something I could parse in a weekend quite easily. But, I really don’t like the idea of contributing to something that competes with what I do. I’ve always been a writer. Never published. Mostly I write long posts like this one with my thoughts, but I’ve tried writing stories as well. I don’t like the idea of chatbots eating our lunch. But I don’t think we can stop them from being used for that.
I’m currently reading (rather, listening to the audiobooks of) the Dungeon Crawler Carl series. The genre is LitRPG, which is fairly new, and a lot of seasoned readers say those books aren’t real books or that they have no literary value. I’ve also read several volumes of Sword Art Online by Reki Kawahara (the books, not the anime based on them, though I’ve watched that as well; and again, also the audiobooks) and it’s much the same thing. SAO has never been called a LitRPG though. It’s an isekai, but while isekai is a Japanese term (it means “another world,” as in you’re trapped there), the Japanese did not invent isekai. The first modern isekai is often thought to be Alice’s Adventures in Wonderland by English author Lewis Carrol, but you could go back much further into classical literature. The Japanese did not start the trend, they’ve just been printing money off of it in the last 10-15 years. Again, I’m getting sidetracked, but the point is, these are books that people think of as less valid, but their fans enjoy them. And if we (literary snobs) push readers of LitRPG and LNs (light novels, the reader group SAO fits into… in the west it’s called YA fiction) into this realm of “your books are less valid,” then we inadvertently give them permission and endorsement to read AI “literature” because we also brand that as less valid, and we should not do that. Hell, if someone wants to read a damn Chilton guide to a classic car, that’s more valid than an LLM-generated “novel”.
Anything generated by a machine should be classed lower than anything imagined by a person. Yes, even if this that and the third. And I say this as someone who just enjoyed an LLM-generated output comparing two fictional universes that have no cause to ever cross. I asked a dumb question (twice!) and got some great answers. Should I be grateful? I remain cautious.
