A Slice of Humble Pie… and AI

Last week, I took my kids out of town for a spring break trip. We were all excited, packed, and ready to go, but there was a problem. We had an hour until lift off. So everyone settled in to entertain themselves and try not to kill their siblings before we even left, and I opened up Facebook, then Threads.

After reading a few truly horrendous pieces of self-proclaimed AI generated fiction from the coaching divas on Facebook, I opened Threads. Without thinking, I read a few posts about AI in books and wrote what I openly admit was a bad take about AI writing. I hit post, and then I promptly closed my phone to go out of town.

I opened Threads a day or so later to something completely unexpected. (The thing every influencers seems to feel/say when they’ve been unexpectedly raked over the coals by their audience.)

My post had been viewed over 5,000 times with dozens of likes and comments. Most were to the tune of “fuck off” or, more kindly, ” as an author, you should know better.” A few called attention to the AI witch hunts growing in frequency and intensity lately, like the recent cancellation of Shy Girl by Mia Ballard.

I’ll admit that initially, the fuck off’s made me angry, but the “you should know better” messages were like a punch in the gut. Like when a parent or mentor looks at you and says they’re disappointed in you.

Panicked and shamed, I deleted the post.

I was on vacation with my kids, I had no idea what to do nor the time to figure it out, and when I did try to create some kind of response, it sounded defensive and dismissive. So I just deleted the whole thing.

What was the Post?

I don’t remember exactly what I wrote, but it boiled down to something like this: “You know how you ACTUALLY know if something was written by AI? It’s filled with soulless synonyms and metaphors that make zero sense. You’re welcome.”

Yeah, I cringe every time I think about it.

It’s a bad take fueled by my own frustration, arrogance, and I think a deep-seated desire to be able to tell when someone has used AI. I desperately want to keep the book space free of soulless AI and money grabbing pro-AI bros churning out bad books.

And yeah, I kind of wish we could tell when something was generated by AI, but encouraging the witch hunts some readers are going on isn’t the way to do that.

Why I Deleted the Post

Yes, I deleted it. I’m not going to pretend it wasn’t partly because I was embarrassed. It absolutely was. When I invite conflict into my social spaces, it’s because I want to burn the patriarchy down or something like that, not rage bait people into engagement or, worse, create actual harm in my community.

So yeah, I deleted it partly out of shame. It would be disingenuous of me to pretend that didn’t play a big part.

There was probably also a part of me that wanted to protect my reputation as an author. I’m not in the big leagues, and I certainly don’t have the backing of a publishing house behind me.

However, the biggest reason I deleted the post is because far more people would see my post than would see the discourse happening in the comments. I didn’t want the harm I did magnified. I didn’t want to help legitimize the idea that there is any definitive way to tell if a book was written by or with AI.

I’m an author, however small and insignificant I feel that I am, and speaking out like that absolutely has the potential to prove to some readers that there is actually a way to tell if something was written by AI.

There isn’t.

Now, I’m not saying that AI checkers are always right or always wrong or that you can’t tell if someone uses AI to write something. If you’re having a conversation with someone over text and suddenly their writing style changes and they’re saying things in that bland, pattern ChatGPT uses, sure you can call out your friend. Maybe they’re using it. Maybe they’ll admit it.

But that’s not enough for me to support accusing authors, strangers to me, of using AI.

The Consequences of the AI Witch Hunt

Here’s the thing… I have absolutely read things that feel like they were written with ChatGPT, and in some cases I’ve called it out and received confirmation that yes, it was. I think I have a pretty good eye for spotting the difference between inexperience, ignorance, and AI generated.

But that is only enough to create a suspicion, and that suspicion is only strong for me to decide what I will and will not read.

Because what if I’m wrong? What if I think I have a pretty good sense for this kind of thing, and I accuse a legitimate author of AI use… and I’m wrong? What if I am so focused on my hubris that I inadvertently support a movement of readers making AI accusations?

This isn’t a hypothetical. This is the reality of what is happening.

Swell Entertainment did a fantastic video on the AI discourse and cancellation of Shy Girl by Mia Ballard, and if I google the book and author name right now as I type this post, the first search result it wants to autopopulate is: Shy Girl by Mia Ballard AI.

I’m not going to comment on whether this author used AI or not. She has made comments with context and explanations and such, and to dig any further than that right now is beyond my pay grade. (See how I need to shut the fuck up below.)

My point is, though, that there is no definitive proof that this author intentionally wrote and published a book with AI. Yet, her reputation is smeared, and she lost a book deal because of it.

I’m not willing to ruin someone’s career over my suspicions, especially knowing how AI companies have trained their language models to mirror the work of previously published books without consent. In the video I linked above, Swell Entertainment ran Frankenstein by Mary Shelley, published in 1818, through an AI checker, and it received a score of something like 80% generated by AI.

Clearly, it’s not possible for Frankenstein to have been written by a tool that didn’t yet exist.

I also don’t want to encourage a trend of authors removing anything that could potentially be viewed as an AI red flag—like em dashes or purple prose that’s poetic and beautiful or phrases that repeat for emphasis or lists of three. The list goes on and on!! You know, I use a lot of those! Hell, there’s probably an em dash several times a chapter in my books, and I know my readers often cite purple prose or overly wordy descriptions in my writing.

I want to read beautiful stories written with flowing language and the occasional overly verbose paragraph. I want authors to write and publish without fear.

I don’t want to contribute to a world where I’m robbed of beautiful stories because of the fear of AI.

The Next Steps

Honestly, for me? My next step is to shut the fuck up.

No, really. I have always said that I don’t want to be an author who lives in the limelight. I want to keep my head down, write my books, and putter around my little garden or sink into the soft blankets of my little goth cave bedroom.

I do not keep up with trends and online discourse well enough to be educated enough to have a say. I shouldn’t have said anything about AI in writing.

That’s a slice of humble pie I should’ve eaten before I made the post, but I didn’t, so I’m eating it now and giving the authors and readers who saw or engaged with that post an apology from the bottom of my heart. I’m sorry.

I want to live in a world where AI is limited, not where the art I consume is limited by AI.