AI & Indie Publishing

[renegadefolkhero]

This got long, oops.

Before I get started, I want to make the distinction between AI assistance tools (which is used for proofreading, etc) and AI generative tools (which are generating prose, etc). The vast majority of writers now use AI assistance tools for editing, for example. Generative AI continues to be contentious... for now.

There has been some recent chatter in my various writer groups about how the flood of AI generated stories is impacting discovery, the algorithms, and features on certain storefronts. On some storefronts, it has become harder to get new release discovery because there's just so much STUFF being published.

"Stuff" has always been a thing. Before AI, people were cranking out low content books on Amazon to make a buck. But I think something that has exacerbated the issue for me personally, as a Smashwords erotica author, is the rise of the mentor cottage industry in my genres.

There have always been knowledgeable writers, as well as grifters, who sell information on how to make a buck in publishing. More recently I've seen established erotica writers breaking into this industry by telling people they can make easy passive/secondary income writing erotica on Smashwords with AI tools, with the goal of having these newcomers list them as affiliates or contribute content to platforms they run and control. These long-time indie writers see the writing on the wall, I think, and understand if they want to continue making a living in erotica in these conditions they need to a) develop platforms they control and monetize and b) take advantage of the idea that with AI anyone can write a book that sells.

It was already difficult to make a living writing erotica, but I believe within a few years it will be impossible to make a living (USA standard) writing erotica exclusively, and these authors are getting ahead of the game.

Recently, a wide publishing mentor I hold in regard began offering an AI course that was touted as how to use AI beyond using assistance tools. Reading between the lines, they are teaching people how to use generative AI. They didn't come out and say it because it's a lightning-rod issue. There was significant interest in this course.

Until now, I've maintained authors are best served by focusing on their craft and writing extremely good, extremely specific books. You can't beat the bots when it comes to volume, so the best bet is probably to pull back from volume/rapid-release strategies and drill down in terms of niche, etc. But what I didn't anticipate was how the flood of stuff would impact existing backlists and new releases. My established pen is mostly okay, but last year I experimented with a new pen, using previously tried and true techniques, and could not get ANY traction on Smashwords. I got traction elsewhere, where the deluge isn't as much of a problem, and came away believing it was primarily an exposure issue and the only way to fix it was to release a lot more books and start doing promo, which I was not willing to do.

I do feel generative AI use has a negative impact on me personally now. I've begun moving out of the genres most impacted, and I've begun trying new things on different platforms.

People were talking about how good the Claude AI is for prose, so I started looking into various AI toolkits and models and these things are widely touted as a way to bypass the "hard" parts of writing and get to the finished product quicker (even saying things like, "write a book in a week"). My knee-jerk reaction is to be bothered by this, but I always think back to the typewriter example. When the typewriter came out, some authors lost their shit. The typewriter made it possible to write stories much faster, and we never went back.

Similarly, we're never going back from these tools. In the future, authors will be able to partly or even fully generate short stories, novellas, and novels that pass muster with a segment of readers. I have honestly been surprised (but maybe I shouldn't have been!) at how many readers either a) cannot tell a short was written by AI or b) genuinely do not care.

I've always maintained most readers wouldn't want to read AI generated stories, and I'm realizing I was wrong. I think that's the part that bothers me the most, to be honest. But we've always known a significant segment of the readership is honestly not very discriminating. There's a reason why so-called "cookie cutter" genre stories consistently get reads for variations on the same theme over and over. There is a reason volume writing is profitable for many authors, even if it's not sustainable. A lot of readers are reading fiction for comfort, or to scratch an itch or a kink, and HOW the story came to be is immaterial to consuming it.

Chuck Tingle talks about how the author and the creation of the story is part of the art. I think this will, by necessity, become more true as AI written works become more prevalent. Authorial face and presence will be a way to distinguish human writing.

Looking at the suite of tools offered by services like Sudowrite, I'm starting to accept there's no way around this. AI is going to exacerbate the volume game, and while my motto has long been, "The only way to win is not to play," the fact is the vast majority of people who want to make a living writing fiction will do whatever it takes to compete, and perhaps sooner than later, generative AI will be viewed as an essential part of the writing toolbox.

If someone insisted everyone write their stories by hand on pen and paper, you'd rightfully think they were crazy. Soon, gen AI will be like that. People will roll their eyes at the stubborn holdouts, and wonder why they are making things so much harder for themselves.

I've thought a lot about what I get out of this process. I enjoy writing and completing stories, and I tolerate the "hard" parts to get the payoff. For instance, I have a working pattern where I often have to put a draft aside at 20-30k to let it breathe for weeks or months. Do I like it? No! lmao. If I could write the same book in consecutive weeks, I certainly would. But it is part of the process that allows my books to go in unexpected places, and I'm not sure there's a shortcut for that.

On paper, AI allows someone like me to skip that marination phase... but does it really? I mean, I might "technically" be able to generate an idea that helps me skip marinating... but am I achieving a similar outcome? I've been doing this long enough to appreciate my brain goes to some very unique places when given enough time to process a story in the background. My stories are consistently reviewed as "different but good," meaning off-market but still compelling. That is my brand. Just the other day, I suddenly realized the ending for a draft that has been sitting for 6 months, and how it ties in with climate change and some other stuff. When that happened, the book completely changed in my mind. It's a different book now. And I'm not convinced 6-months-ago me, with AI tools, would have reached the same place.

I think we are going to reach a point where most to-market indie books will be AI-created to some degree. The models will become so GOOD at writing to market it will seem silly and inefficient to do it any other way. But I think my original feeling, that holdout writers aren't competing with AI, still holds. Readers who are satisfied with AI books may not be your audience.

Take faro (fantasy romance). Fantasy romance has become somewhat boilerplate due to volume. Imagine a world where most faro books are AI generated. Is that world NECESSARILY different to the outsider who doesn't read faro? And is it necessarily different to faro readers who are only looking for different flavors of a similar story type? If we get past the stigma of "a machine wrote that" can we accept that it's scratching the same itch, and writing it 5x as fast isn't necessarily a problem?

Anyway. I was already making some adjustments based on the challenges at Smashwords, as well as my concerns about future censorship, but with the failure of my most recent romance I've decided to take a biiiiiig step back from making compromises and writing what should "reasonably sell" (since clearly it isn't, lol) and get back to writing stories only I can tell, and accepting those stories might not sell, or get picked up, or whatever, and that will be hard, but maybe part of my journey is reaching the point where I'm okay with that and satisfaction in a book well-written is enough.

I'll talk about my failed launch later. Maybe. I'm still reeling a bit tbh. But I think all these things are forcing me to be a lot more honest with myself about what makes it worthwhile. If your books don't sell, what makes publishing worthwhile? You need an answer to that question if you're going to continue.

Comments

[samuraiter]

What I find profound about the whole thing is the speed with which we have gotten to this point. AI is improving so quickly that I can scarcely imagine where we will be in one, five, ten years.

[renegadefolkhero]

Yeah, looking into the tools we have now was eye-opening. I think it's developing too fast to really speculate.

[lassarina]

This reminds me of something Seanan McGuire said: that if it's not important to you to be published very specifically, write fanfic, you'll be happier. It's a little funny to me because I am apparently a weird reader; I'm very picky about how tropes are implemented, even my favorite bulletproof ones that will make me pick up something I otherwise wouldn't consider, and I only care if the characters are interesting to me personally.

idk. I'm thinking maybe I won't try to publish my original stuff. The more I look at publishing the more tired I am.

[renegadefolkhero]

I think it all boils down to having realistic expectations, which a lot of writers don't have (including myself) and that's what makes the reality checks hard. People who want financial success in publishing have to understand that you have to want it REALLY BAD, and you have to put a ton of work and time in, and you still might not get where you want to be.

But people who just wanna share their work, man, go for it. Photoshop a cover or buy one for $10 and throw that cozy spicy werebat mystery onto Kindle Unlimited. Fuckin do it baby. There are a lot of hobbyist writers in small niches who have fans that just love their stuff.

[lassarina]

I'm so, so grateful that I decided a couple decades ago (how am I old enough for this to be true) that I did not want my ability to pay the mortgage to be dependent on my ability to catch the zeitgeist and win the publishing lottery, because I like having a nice steady job I'm okay at and being able to write at my own pace (modified as necessary for the fact that I am crazy and set unrealistic goals).

But yeah, photoshopping a cover and throwing it up sounds very appealing.

[zillanovikov]

I've decided to treat writing fiction as a hobby. It's not weird to invest money in buying and painting a Warhammer army even if you're not a top-ranked player. Just play in your budget.

I write original fiction cause those are the stories I have to tell, and I pay in swaps and in money for a cover and layout and editing I'm proud of. But I don't expect to make back my costs any more than your average Warhammer player expects to get rich off going to tournaments.

[lassarina]

"play within your budget" is a great way to put it.

[zillanovikov]

My fiction will never sell enough to make back my costs (I'm too niche and also too finicky about my final product looking "right") so I've decided the point, for me, is the process of telling a story and sharing it with a small number of people who *get* it. Which means genAI is useless to me, because the process *is* the point.

I know writers who write their first draft with a pen and paper because that's their process.

I think it would be different if I was writing for income but my fiction is
... Not that.

(Just to say, I more-or-less agree with your points).