This one is going to ramble some, I think, so apologies. You might want a snack. If you are a writer, you might want a drink.
The Atlantic, because its primary purpose is to make upper class, boomer and Gen X white people feel smug, recently had an article about how reading is dead. it is not a good article — it cherry picks stats and makes dubious assertions — but it does perfectly capture the “the kids today are wrong wrong wrong!” vibe that is essential to the ongoing subscription and advertising sign-ups that the Atlantic depends upon. Unfortunately for them, and kind of hilarious for us, two new news items appeared shortly after the article dropped. First, book reading is actually increasing. And second, there has been a mini-wave of AI controversies over publishers using Pangram and other evidence to claim that books were AI written. I think these are all connected in the odd corners of our collective minds.
Books and reading, whatever the Atlantic might think, are not in terminal decline. Book sales — both in terms of dollars spent and units purchased — increased in 2025. People are spending money on books and buying more individual books. Reading is neither dead nor the habit of a few weirdos. People still value reading. The imitative AI kerfuffles (still the top word in the English language), however, may tell us that people do not still value writing.
Two books with major deals were both subject to claims that they used imitative AI in their writing. One book was pulled by their agents, one book was kept by the publisher and did well (to be fair, it was a self-published book in a previous life and self-published books with a significant sales record tend to do well when they are picked up by major publishers). I am not going to get into the ins and outs of each case, but they are interesting, if contradictory. Some publishers an agents do obviously care about human writing — the agents walked away from a lot of money when they pulled that book. Part of that is simple legal exposure — putting aside the ongoing plagiarism and copyright lawsuits, you cannot copyright the output of imitative AI (yes, yes, Mr. Lawyer reader, that’s a touch of a simplification, but it is directionally correct). But at least some of the reluctance comes from the conviction that human writing has a higher ceiling than imitative AI writing. By its very nature, imitative AI is, well, imitative. It can only mix up what is in its training data, and that is going to tend toward a blander middle. But there is money to be made in the blander middle.
The book that succeeded was a genre book, and in a genre where readers are very tolerant of tropes. They like those tropes, and read them for enjoyment. And I include myself amongst them. As much as I like stories that play with tropes and subvert them or our expectations about them, I still read and enjoy stories that just use those tropes unironically. A decent number of writers have made their living writing those stories, and even more learned to write the more unique pieces by starting out with the tropes. That the book sold well says that at least some portion of the market does not see a human being as a necessary component of the process.
All smug (see, what did I tell you about Gen X?) snark aside, I think the Atlantic author might be responding, at some level, to the general sense of AI doom hanging over literature. We are told constantly that the AI firms can replace writers, and many of us realize that means replacing the most compelling and interesting kinds of writing, the writing most likely to move people. And that feels like a real loss, like reading is under threat. Having a book that might have been written by AI sell well only reinforces that notion, even if editors, agents, and most writers still see the value in human stories. So where does that leave writers? I’m not sure.
As an aside, if your inclination after reading the next section is to rant to me about how I am anti-tech or a Luddite, then spare me. First, the Luddites were right. Second, I am not anti-tech. I switched majors in college because I loved technology and what you could do with it so much. I have spent more than two and a half decades in technology and contributed to open source projects multiple times. I bought the first generation Kindle as soon as it was available. I have never thrown away a computer because, well, I can always throw Linux on it. I am working through quantum programming for fun. I am a dork, through and through. I don’t hate tech — I hate what the powerful have done with it.
Back to rambling.
I think that self-publishing is most likely to be swamped, for the next little while, with bad imitative AI stories pushed out just to try and make a few quick bucks. But I also think that we will see more and more books written substantially by imitative AI be accepted into the market. This will be a short to medium term effect — I do not think people who use these systems understand just how deeply subsidized their usage really is — but in that short to medium term, I suspect getting published as a debut author will be harder, there will be less money for human writing, and we will possibly see fewer and fewer lasting works as the quick middle drives out the business model for the longer, but higher quality, edge. I don’t think any of that is good.
I am going to miss the works that would have had a chance to grow from the middle to the high-quality edge that I won’t get to read now because those authors were never given the chance to build that kind of career. There is a low ceiling for imitative AI work, but there is the real possibility that it is high enough to prevent authors who could eventually produce works that really move us from surviving or entering the field. We will have lost something precious if that comes to pass.
Look, I am a bad writer. If I was not a bad writer, I would have had something published or produced by now. This means my writing is largely for myself, for my own enjoyment of the process. None of that changes in this weird new world. But there is also the hope that someone will read my work and find it fun and interesting and thought-provoking in some way. If a flood of quickly produced imitative AI works makes that harder to achieve, of course I am going to lament the loss and wonder if the voices telling me that it’s not worth trying to improve are right. We are, after all, the storytelling ape. Education, culture, community — all stem from the stories we tell ourselves. I am not sure what becomes of us if we stop telling those stories and rely merely on the recycled and regurgitated. I doubt it is anything good.
Weekly Word Count
Well, that was cheery.
My word count isn’t any happier. I have been largely struggling with the re-structure/re-plotting of the tech/abortion/privacy thriller. I have had a very hard time breaking this story in a way that I am happy with in terms of themes and still feels like a thriller. Combined with not being able to fix the clone school shooting story and I do pretty much feel like the world’s most incompetent writer, unable to even compete with Grok. Well, maybe not Grok bad. But pretty bad.
Did I mention I do this for fun?
I do think I am moving in the right direction on the thriller, though. If not, I may drop it for a bit and work for a few days on the Pratchett meets Rockford Files novel — that one at least has a reasonable plot. I did promise an editor I would make the thriller changes (nope — not sold. Just right of first refusal thing with a new small press as mentioned before. Still a failed writer), so I do want to give myself every chance to make that work.
I hope you all have a great weekend.