Facebook Posts
If you don't follow publishing news (and you don't read horror fiction), you may not have seen the headlines about the cancellation of a novel called "Shy Girl" when readers began to complain that it had been AI-generated.
All of us who write and publish for a living are on edge about AI-written texts. For one thing, it's not so easy to determine when a book has been produced by machine rather than by human brain. Ironically, AI tools are often used to analyze prose for the presence of AI, and because our own books have been used to train the AI itself, our styles may set off the AI alarm bells. And for another, what will happen to the business we love if AI gets better at this task?
This piece points out some of the dilemmas.
**
When [writer Antonio] Bricio learned about the novel’s cancellation on social media, his stomach dropped. He said he does not use A.I. to write, except to occasionally translate a stray word or phrase from his native Spanish into English, in which he is also fluent, using the A.I. translation program DeepL. But he wondered what an A.I. detector would say about his work.
So he paid for a subscription to Originality.ai and uploaded a chapter of his novel. The detector was 100 percent confident that he had used A.I. in some way.
Bricio searched for the phrases that had tripped up the detector, deleted some sentences and reran it. This time, the program said it was 100 percent certain that a human had written it. Eventually, Bricio had a chat conversation with a customer service representative, who told him that if he received results that incorrectly flagged his work as A.I.-generated, he might need a different model of the program.
The back and forth only left Bricio more unsettled. The Originality.ai reports on his draft, which he shared with The Times, showed that adding or deleting even just a few sentences produced wildly different results.
“What if publishers or agents start running these A.I. tools on everybody?” Bricio said. “Everybody is going to walk on eggshells from now on.”
...As the publishing industry wrestles with the intrusion of A.I. into nearly every aspect of the business, there seems to be little consensus over what publishers can or should do to regulate how writers use the technology. But many agree that the current state of affairs is untenable.
A growing number of writers face unfounded suspicions of A.I. use. Others use A.I. without disclosing it. Many readers feel confused and wary, not knowing whether the books they’re reading were written by a human or a machine.
“We’re reaching this era of distrust, with no easy way to prove the veracity of your own writing,” said Andrea Bartz, a thriller writer who was a lead plaintiff in the class-action lawsuit brought by authors against Anthropic, which agreed to a $1.5 billion settlement.
Bartz recently put some of her own writing into Ace, an A.I. checker, and was startled when the program labeled her work as 82 percent A.I.-generated. The program then offered her a solution: “Would you like to humanize your text?”
When Bartz wrote about her experience on Substack, dozens of writers chimed in. “I guess that’s what happens when your books were stolen to program A.I.,” the novelist Rene Denfeld commented, noting that an A.I. detection program had also falsely determined some of her writing to be A.I.-generated.
**
www.nytimes.com/2026/04/10/books/shy-girl-ai-publishing.html?unlocked_article_code=1.aFA.KfEz.nWH...
How Authors and Readers Feel About the ‘Shy Girl’ Cancellation (Gift Article)
www.nytimes.com
Major publishing houses risk unwittingly putting out books generated with A.I. tools. Authors and readers are frustrated, nervous and grasping for solutions.8 hours ago
In Germany, many political and industry positions de facto require a PhD. Some years ago (and it still continues…) there was a rash of important people being toppled long in their careers because it became evident that significant chunks of their dissertations written years, often decades, earlier had been plagiarized from other writers. Many people were ruined via “simple” Googling of sections of their work. The A.I. revolution is going to raise this to a whole new level. Nonetheless, There is a demonstrable difference between an A.I. tool determining “He plagiarized this from Act 2 Scene 1 of ‘The Tempest’ — go see for yourself“ and “This appears to have been created with artificial means — trust me”. In the latter case, there is no direct evidence or anything that can can be offered up as proof. Behind the scenes, a “score” is calculated on numerous points and is distilled into a binary assertion. A.I. characteristically makes pronouncements with higher degree of confidence than it should have given the evidence it possesses.
Did you see where they ran “Frankenstein” through the AI sensor and it said it was all written by AI 🙄
If authors are having this much trouble, how do college students prove that their papers are original work?
AI tools are used to scan student work for AI. One must learn to interpret and followup on the data report because most often it is phrasing suggested by grammar tools or the way the person’s brain uses syntax or tone. AI can be a free personal assistant but we must remember that it can lie and hallucinate.
"Once men turned their thinking over to machines in the hope that this would set them free. But that only permitted other men with machines to enslave them."
Wasn't there an experiment done where they uploaded paragraphs from 19th century novels only to be told they were 100% AI?
This actually happened to me in a smaller way this week. I'm looking for a job, and the job posting was very direct and strident about the cover letter not being AI. I took that to mean that they were running cover letters through an AI detector. So, I wrote a very standard, formal cover letter and ran it through an AI detector. Ot flagged 98% AI written. I fiddled and fussed with it and eventually came to the conclusion that the formality of the letter was part of what flagged. So I wrote a much more casual letter (and also wrote about my experience) and sent that. They still passed me over. 🫠🤪🫤
This is a big concern at my child’s college. Some professors are using this software to check and the results are inaccurate (especially if you used spell check) and penalizing students. Students with proof they wrote it are having to fight their cases with administration because professor won’t hear it. Causing a lot of stress to good students.
This sort of thing is happening to kids in school...my daughter at least.
Thank you for bringing this to our attention. Yikes
Drew McCullough
When AI first started, an author took something she write pre-AI birth. It flagged her writing as 90% AI. I did purchase a non-fiction book from unknown author. It had all the formatting of ChatGPT including Emojis and bulletpoints. I kid you not. I didn’t even read the book. I forgot to return it in time.
Oh my lands…
What a mess!
If you haven't yet checked out THE GREAT SHADOW, please consider doing so! Just a couple of interesting developments...
This week I did an interview on the Disrupted Science podcast, which came out yesterday. Lively, intriguing conversation--very much worth your while.
www.the-geyser.com/interview-susan-wise-bauer/
And if you happen to be in the vicinity of NYC, The Health and Medical Book Club, which is a collaboration between Weill Cornell and the New York Public Library, is featuring Great Shadow in their May 21 meeting:
library.weill.cornell.edu/news/health-medical-book-club-great-shadow
www.the-geyser.com
The author of “The Great Shadow” joins us to talk how history bears on today2 days ago
Read it a few weeks ago and loved it!
Thank you!
So far, thirteen lambs, and four bottle babies. Kind of eating up my available posting energy. This is my life right now.
It's a pretty good life.
For some reason, Facebook is refusing to post videos from this account no matter what I do. So, sorry about the link, but it seems to be the only way to give you a glimpse of the lambs.
www.instagram.com/p/DW4gF2qDqa3/
3 days ago
