Artificial "Improvement"

Dear Reader,

I used to write blogs about my books. I’d focus on where I got the inspiration or the idea, talk about the characters, and more, but I feel like my last few blogs have been more serious and about the state of the industry or my specific indie author situation. This will be another one of those blogs because, unfortunately, today, as I write this, I am fired up.

Before we dive in, I should say that I started writing this blog weeks ago and planned to publish it at a later date because my wife edits my blogs before we put them on the site, but every time I thought I was done writing this, something else would pop up, and I had to add it here. I think I’m on version 15 or so now, and that’s no exaggeration. For context, I probably started the first draft of this blog around the beginning of April, and by the time you read this, there have probably already been 10 new things I could’ve added to this.

Now, let’s get to it…

I’ll start with my stance on AI in the publishing industry. I placed a line on my site a while ago stating that no AI is used in any part of my process. Here’s what it actually says:

“I do NOT use AI in any way in the creation or publication of my novels. This includes, but is not limited to, idea generation, outlining, writing, editing, cover design, marketing materials, website content, etc. My books are “all human” and always will be. To actively avoid using AI, I use provided filters, tags, labels, turn off any and all AI features on software I use, and review things by eye to the best of my ability.”

I did this because I wanted everyone to know where I stand and that my books come from me and me alone. I write the books. Then, my wife edits and formats them. We work on the covers together because we both have different skills and ideas and we like to combine them. Since I became a full-time writer, I also handle all the social media stuff, including creating the graphics. She’s better at numbers, so she handles most of that stuff for the business. You get the idea. Throughout all of it, though, we actively avoid using AI, and that’s what it feels like sometimes: like we have to be on the look out so that we can avoid it.

Oh, and don’t even get me started on all those people out there using AI to write emails that they flood authors with talking to us about how they’ve read our book, followed by the description and the author’s bio that they found on Amazon and had their little bot include. We’re all getting those regularly, and I used to be okay with getting occasional unsolicited emails from people offering their services. I could politely decline and go about my day. Now, they’re numerous and are mostly bots or scripts that are created with AI, and the amount of time it takes to rid my inbox of this spam is time taken away from other things I could be doing.

A software company calls their new feature “an improvement,” and as soon as they update, without my permission, because you can only delay these updates for so long, I am bobbing and weaving to avoid using that “improvement” by disabling everything I can or trying to learn where it is so that I don’t unknowingly use it. It’s easily become my second job. I know many of you know what I’m talking about because this isn’t author specific. It’s part of our normal days now. Download a new app on your phone, because everything is an app. Go into your settings. Remove access and permissions. Disable any of the AI or “Intelligence” features immediately. It’s infuriating and not likely to stop anytime soon. Just the other day, a certain company emailed me at 6:30 PM my time to tell me that their workspace had been updated to include new features, using their AI named after a zodiac sign, that would access my drive, calendar, email, and more in order to make using their AI workflows easier. Oh, and it took effect that day, meaning I wasn’t emailed BEFORE they opted me IN because they did, of course, opt me in by default, like all of these “helpful services” are doing, and the best part of that email was that they said they would give me the option to opt out, but that the option itself could take up to 3 days to appear, meaning if I didn’t have it, I would have to keep checking for up to three days. During that time, they would have access to everything on my workspace. Access I did not authorize. “Features” I did not ask for. Choice I did not agree to or make myself because I was opted IN.

Oh, and I can’t forget to mention that when I order my author copies, not only does it take up to 6 weeks for them to arrive, the delivery is not free, but the robots at the warehouse “think” that it’s a great idea to put my books in a box with lotion or liquid soap, and since nothing is ever packed correctly anymore, you can probably guess where I’m going with this.

Then, a reader sends me a message via the contact form on my website, and when I go to reply, I am asked if I want AI to auto-draft my response.

Me: “What? No. I can handle writing an email.”

AI: “But you’re posting a blog. I could re-write that for you so that you don’t have to.”

Me: “It’s MY blog. I’m good.”

AI: “Okay, but I have this really cool AI assistant that will help you organize your tasks, and–”

Me: “No!”

Just to respond to an email or post a blog, I have to dodge the AI like those perfume sprayers in the malls of old, who didn’t like to take no for an answer.

I release several books a year, so I think I’m able to observe things a little more in “real-time” than someone who publishes less. Love Study, which came out in February, was able to receive both ratings and reviews on Amazon without issue, but when Singles’ Week came out on 4/10, ratings weren’t permitted. Readers had to leave a review or bust. On 4/26, ratings were suddenly allowed. Two side notes about this.

Readers who review, especially at the beginning, tend to leave higher ratings along with their words. They’re most likely on an author’s ARC team or their biggest fans, early adopters if you will.

Readers who merely rate a book without words about 2 weeks post-release tend to be much more spread out in their stars. If a book was 4.8, for example, for the first couple of weeks, that rating will start to go down and dip; not just because of the number of ratings/reviews, but because the early adopters have already read. If they didn’t want to type a review, they couldn’t leave a rating. Having ratings available 2 weeks post-release means the book’s rating usually tanks. Overnight, it can go from 4.8 to 4.4 or 4.2.

The “best” part is that authors had to learn this as they went because Amazon did not communicate this change. There is no author Amazon email alerting us to upcoming changes and how they might impact our sales, page reads, and rankings. When this happened, authors were confused. I saw other authors in said confusion and tried to ask my readers to please review the book since they wouldn’t be able to rate. Then, a little over two weeks later, ratings are now available. I have no idea if they’re back for good, how it will impact my next release, or if they’ll disappear altogether. It’s also true that back when I first started publishing in 2017, you could only review. Later, they allowed people to just rate, so why are they at least testing going back to the old ways?

The reason I find this interesting is that they can then mine those reviews with their AI and use that to summarize the book (or the reviews themselves). These summaries and descriptions are sometimes, if not often, inaccurate, and now authors have to beg people not to just rate the book (when the number of ratings was already declining), but to review as well (at a time when the reviews are also declining). This is at the same time that the audiobook platform owned by Amazon, which many authors use to turn their books into audio, gives me a pop-up every time I log in (which is daily because I’m actively producing audiobooks) that asks me to use their voice replica beta program (AI). I can hit the X on that pop-up, but when I log back in later, it’s right there again, annoying the crap out of me.

Also, recently, there was some book drama on Threads where someone said that if an author can write and release a book within seven months, it must be AI, and let me just say this… I drafted All the Love Songs in four days. In 2017. My hands aren’t as young or as fast anymore, but I can still pull off a 10k day easily. I wrote the whole New Orleans Series in 215 days (between October of 2021 and June of 2022, because I’ve been keeping track of what I write and when). It took my wife longer to edit those books. The famous “chat” people use to “write” books did not even come out until November 30th of 2022. Before I started publishing, I used to write one-hundred-thousand-plus word fanfics over the course of a weekend because I did nothing other than write, sometimes eat, and didn’t sleep much. Sometimes, some people are faster than others at something. If you give me a calculus problem and ask me to solve it in five minutes, I wouldn’t even understand it within that time, but for someone else, they’d only need five seconds. Everyone’s different. That’s okay. That doesn’t mean authors are using AI. We need the people who understand how to do things faster and better. I can’t operate on a patient, but a doctor can. My wife and I were joking about how much better she is at numbers than I am. Can I figure stuff out at some point? Sure. Is she faster and better at it? Yes. Can she write a book as fast as me? No. We collaborate and run our business effectively because of our strengths.

We’ve gotten away from what it means to be human and be part of a society that works together because each of us has our strengths and things we don’t know or understand. Just because someone can’t do something fast, doesn’t mean that someone else couldn’t. Trust me, you don’t want me in that operating room (or my wife; she’ll pass out), but that doesn’t mean the doctor can’t do their job.

This has gotten me to the place I am now, where I struggle when it comes time to start writing a new book. I’ll back up a minute and say that, yes, I write fast. I always have. Every book I’ve published, I have written myself and quickly. Then, it goes through the editing process, and eventually, it gets released to you. Sometimes, I’m working on books that won’t be released for a while (I’m talking years), which means while I’m writing it, something could be reality, and by the time it’s published, it’s less so, so I then have to edit some things out. I find myself asking things like, “When I publish this, will X job still even exist anymore, or will AI be doing it? Would people even meet in a local bookstore in the next 3-5 years? Will there even be local bookstores in 3-5 years?” You get the idea.

Speaking of jobs, this is one of the issues I’ve been running into. When I come up with a story, I get my characters together, and I give them jobs. That’s been extra “fun” for the past few years because jobs have been disappearing, and I write contemporary romance (for the most part). While my books are fiction and not based in our timeline necessarily, I do try to stick to reality as much as I can. So, I wouldn’t make my present-day character have a job that makes no sense in the present day. It’s getting harder and harder to come up with what people would do, how they would even meet each other, AND make both of those things realistic. For example, I used to work in tech, and we went fully remote during covid and stayed that way. That was true for a lot of tech companies and even non-tech companies. If my character just sits at home all day and works, how will they meet their love? Okay. I’ll give them a reason to go out on a Saturday night. Maybe they go to a bar. Oh, no. They can’t afford that because drinks are now incredibly expensive due to the cost of living crisis, and they have to pay their rent and for their groceries. There’s more, and that’s an extreme example, but I’m sure you get where I’m going with this.

Recently, I was trying to give a new character a job, so I searched for “the hardest jobs for AI to replace.” It was pretty slim pickings out there, and I had to scroll past the AI summary to get to an article, which isn’t exactly ironic, but you get it. Before I could get to an article, my eyes landed on something in that summary that made me laugh. Their AI summary said that authors/writers would be one of the hardest jobs to replace with AI. Now, I do believe this because AI cannot craft stories how a human can, but that certainly doesn’t feel like my reality.

That leads me to the fear. Being an author right now is scary. It’s not just because sales are lower than they used to be or even because new generations are reading less, and when they do read, they’re reading short. It’s both of those things, but also the fact that AI exists, and unfortunately, some people seem to think that its stories are good enough, that authors don’t need to write anymore, or that they can become an author by typing a prompt and hitting publish.

Platforms are pushing AI everywhere, and I’ve always seen people complain about the quality of the books on Kindle Unlimited compared to non-KU books, but the number of times I’ve seen people recently comment on every book on KU being AI now has me concerned. That’s simply not true, but if that’s the perception, and people choose not to read my books because I offer them on KU, that’s not only going to make it harder for me to publish, but it might force me to remove my books from KU because of that perception. It’s not just that my income is and would be impacted – and it would because if KU is flooded with these “books,” and people read them, it gives those “writers” a share of the KU pool, which is how authors get paid – but it also takes attention away from books written by humans, and as readers find more and more AI on KU, they lose faith in their subscription, might stop using it, or start saying that every book on KU is AI.

You’ve all likely heard of the issues authors are experiencing with the decline in page reads and/or sales through Amazon. This has been going on for quite some time, and it appears to be a somewhat universal issue, based on what I’ve seen and heard, anyway. I have also noticed a massive decline in page reads compared to my historical data while having more books out there for people to read. Based on the information I possess, I can only infer that one or more of the following things are true, because the KU pool (the subscription money that is divided up amongst participating authors based on their page reads for a given month) has either been steady or has gone up slightly despite any possible boycotts or subscribers leaving:

• People just aren’t reading my books how they used to, so my reads are lower (by a large margin).

• AI is running this part of Amazon’s production, and there is a glitch impacting reporting (not a delay in reporting, but an issue with the data, which some reps out there have confirmed), and they can’t figure out how to fix it. That would also explain reporting delays because they are using the compute elsewhere, which means we no longer get data as fast as we used to.

• AI slop has infiltrated KU so much that readers can no longer find my books with as much ease as they used to be able to, which impacts my page reads.

It could, of course, be something else, but that’s where I am at right now. This makes it difficult to know what to do as an author. If it’s truly an issue with books no longer capturing readers, the author would take different steps than trying to solve for a reporting issue. We have no way of getting the info from Amazon, outside of reaching out to support (where you also have to wade through AI), and it’s difficult to trust the reps when some of them say one thing and others say something different. Therefore, I don’t know if there are pages missing that I haven’t been paid for or if my pages are truly down and I need to start thinking about making other decisions for my business and my family. I don’t want to have to take my books off KU because I publish a lot of books every year, and I know people rely on KU to read them, so they wouldn’t be able to, and no, going wide is not an option for me; not at the present. It’s important to note that an author has to go where the readers are, and readers are still on Kindle. I tried Kobo, and it didn’t work for me, so for now, this is where I am.

Here’s a fun fact for you. There’s a book that people are calling out on social for being blatant AI, but as I’m writing this (and keep in mind, my blog gets edited and then published on a certain day later), that book is 183 overall in the Amazon store, number 1 in multiple categories, over 1000 pages, and there doesn’t appear to be much, if any, info on the author. So, it’s pretty obvious that this is not a book that someone wrote. They prompted it, made it extra long to get as many page reads as possible, and then, published it. Amazon currently has no defense against this, other than to ask us when we publish if the book was made or assisted by AI. That’s a checkbox, so if someone doesn’t check it, they’re lying, but if they check it and publish, they’re not, so Amazon publishes it.

Here’s the interesting thing. That book also has a 4.4 rating and is highly ranked, so people are either actually reading it or someone has created bots to “read” the pages for them (which is a distinct possibility given the world we live in). This cheats the system and gets them a ton of page reads, moving the book up higher and higher in Amazon’s ranking, which gets them more reader eyes on the book, who might then decide to read the book themselves. That “author” then gets a large pool of the KDP fund that is split between everyone who has a book on KU, which means that everyone else gets less. Assuming the “author” has created these bots, there’s literally no limit on the amount of page reads they could get by creating more bots (outside of the fact that KU doesn’t allow for more than 3,000 pages of the same book to be counted as read by the same reader). Real authors out there cannot create new readers without doing an intense amount of marketing and often paying for ads (and of course, writing good books), so there’s no way for us to keep up with these AI “writers.” And that’s me saying that; I already write fast.

Taking the other position for a minute and assuming people are reading this book or other AI generated books, this could represent an existing or an upcoming shift in the market. While there are very staunch anti-AI readers who are on a crusade, there ARE people who seem to eat this stuff up and not care that it’s AI, and it’s concerning that it’s a large number of people. I see this, in a way, as a generational shift. Please note: I’m not calling out a specific generation here. I’m merely stating my hypothesis. Boomers and Xers grew up without things like social media and AI. I’m a Millennial, and during my teen years, we had dial-up internet, so not even close to what technology is today, but Gen-Z grew up with social media. Alpha is growing up with social media and AI. I don’t have a proof of this, but as Zers and Alphas become more and more of the book market (or any market, really), the trends will be different. If they grow up with AI or use it to help them decide what to cook for dinner that night, they will become more and more comfortable with the language, the way stories are told via an AI tool. They will recognize human-written books less and less as literature and more and more as something they have to go out of their way to buy when AI is integrated everywhere and, in some cases, and at least for now, free.

I’m also seeing more and more requests or reviews (not just for my books) where I’m noticing a trend in what I call the “I want” group. It’s where someone says they wanted the book to be X, or they wanted there to be a specific scene, event, etc. This is different than someone saying they didn’t think the book had enough, for example, character development or something like that. Algorithms are very, very good at showing people what they want. If someone is used to: A) Free content of any kind like they get on social media platforms where they can endlessly scroll and consume, and B) What they’re interested in is tailored directly to them, they might find that a book lacks that tailoring to them and comment that they wanted a completely different story for that character and not the one the author wrote, or that they wanted the book to have a very specific scene, and it didn’t. It’s not about feedback, critical or otherwise, to me here. It’s an interesting trend that has me wondering because in a world with AI, someone could prompt their own book, and get exactly what they want for themselves (good or not), so what does that mean for people like me who write books that will never have everything that one person wants?

I’m worried about whether or not I’ll even be able to publish in a few years because the market just might not be there anymore. If everyone is reading AI books, and I don’t use AI, no one’s reading what I write. I saw recently (one of the reasons I’m typing this) that a prominent publisher is partnering with an AI content platform to create shorts based off of their novels, and I wanted to throw my phone because, COME ON! These same publishers are also telling authors, “Don’t write this because it sounds like AI.” Yeah, I wonder why. Will that change? In a year or two, will big publishers start dropping some of their authors because AI makes them more money and they don’t have to pay people to do the actual writing? Will they see the market shifting, like with the book I mentioned that’s high in the Amazon store, and decide that human stories no longer matter, or rather, that they matter, but only in the way that humans are the ones feeding the AI, and then, the AI spits it back out and takes credit?

Overall, it’s this vicious loop that we find ourselves in. Authors used the em dashes, so AI uses em dashes, and now, authors are called out for using AI if there’s an em dash present in their book. I saw an author accused for using the word “padded” as in, “She padded down the hall.” That’s a sentence I know I’ve used, and the word “padded” exists in this context for a reason. She didn’t just walk. She walked quietly or softly. Recently, there was the post from someone who was told not to use similes and metaphors because that’s a sign of AI. HUMANS did it first. AI copied it, but they don’t want to be the target of a witch-hunt any more than the rest of us do.

On that note, I wrote a novella a long time ago, called Cabin Crasher. It was free on my website and for newsletter subscribers for about 3 years after that. I wrote it before we’d even heard of ChatGPT. Then, people had asked me to publish the book, and I did back in January of this year. There was a review that was pointed out to me where someone said they thought I’d used AI for the first 1/3 of the book, but then, they realized it was my style/writing. I’m paraphrasing here, and I’m honestly not sure what that means. I wrote the whole book myself, so it’s all me, but another reader could see that, see “AI” in a review, and run the other direction. The Amazon algorithm could pick that up and sit me next to books that were actually “written” by AI in some way. It could start a witch-hunt. That witch-hunt could do irreparable harm to my reputation. And when this happens, it could mean that an author, especially a self-published one, might not be able to publish more books for a variety of reasons (they might be too demoralized, their sales could plummet, making it not a viable option, etc). People are then allowing the AI to win because real artists are harmed with these often-false accusations, their work is fed into these AI “checkers” that use that scanned in work to train on, and you get this self-enforcing loop. “I put it in a checker, and it said you used AI.” Of course, it would say that. If everyone puts my books into checkers, those AI models now have been trained on my books (as if these companies haven’t already scraped every pirated book in existence, allegedly, of course). No one should be helping a technology replace creatives by publicly accusing them with this kind of “evidence” when we know this isn’t evidence at all. There is a well-known class-action lawsuit pending right now that I had to join because their AI stole several of my books without my knowledge, let alone permission. We’ll see how that all shakes out, but I had to take time to fill out a claim for each book individually. That’s time I then couldn’t spend writing a book.

I publish fast. Thankfully, I’ve been doing that since 2017. I tend to put out a book about every 6-7 weeks and have been since long before the AI video of Will Smith eating spaghetti, that we all laughed at, started making the rounds, but if someone doesn’t know that and they see that I publish this fast, they could make assumptions, write a review or put it on social, and then it’s out there. Once that accusation is made, it doesn’t just go away. This worries me. I know it worries other authors who publish multiple books a year as well.

The world is already scary enough right now, and that doesn’t seem to be changing, unfortunately, and now on top of it, we have to worry about dodging AI like we’re playing a game of whack-a-mole in an arcade. I really do spend an inordinate amount of time thinking about AI these days, and most of it is on how to avoid it. How to search without getting the summary to pop up in the first place. How to make sure a graphic I’m making doesn’t feature any AI elements. How to disable everything I can when there’s a new update so that they don’t feed my work into their system without me knowing. It all feels pretty futile, but just know that I (along with a lot of other authors) are trying to fight what I believe to be a good and worthy fight. I hope, if you’re a reader, you’re going to fight with us, but please be careful about accusing someone of using AI. Once it’s out there, it’s there forever, so keep that in mind.

I know I’m not the most vocal on social media. I never have been. Maybe I’m just too long-winded, so I choose to write blogs like this one instead, but I see other authors out there having to defend themselves and others almost on the daily. This scares me, as I’m sure it does most of you, too.