I know AI has been hashed and rehashed to death, but it seems the subject just won’t go away. On February 8, the NYT published an interview with Coral Hart (not her real name but a retired pseudonym and the name she uses to teach AI-assisted writing). Before she became an AI enthusiast, Hart published 10 – 12 books a year using 5 different pen names.
Last year with AI’s help and 21 different pen names, she self-published more than 200 books, ranging from sweet to hot to sizzling hot. She didn’t used AI just to research, but created the plot, characters, setting–the whole ball of wax with AI. According to the article that you can read here, none were blockbusters but collectively sold 50,000 copies earning her a nice piece of change.
The article doesn’t say how long the books are, and it didn’t reveal her pen names which prevented me from hopping over to Amazon to read some of the sample chapters. It did say that she never revealed the books had been AI-generated because she didn’t want a bias against them.
It’s a free country so I don’t care that she created all the books using AI, but I do care that she didn’t reveal that fact to her readers. It was cheating. But even more disturbing, at least to me, is the article and others similar to it seem to treat her like she’s some guru who should be followed.
Of course not all responded favorably. Many romance authors responded that romance is all about emotions so how can a machine that has no heart or feelings write about such things?
It can’t. AI has no soul. It doesn’t sweat blood to get just the right emotion of someone who is dying inside because it doesn’t know what that emotion is. Hart even agrees that she has to diligently work on her prompts to get close to what she wants. I wonder what kind of writing she might do if she put that much effort into writing the emotion herself?
What scares me about this article is that there is already a glut of books out there and AI makes it so easy to put more out. But I have faith in readers and unlike Hart and many others, I believe readers will be able to tell the difference between an AI generated story and one that a flesh and blood author creates from the heart–a least at the present time. I’m not sure about the future.
What say you, TKZers?
Hmm . . .
A marignally talented writer hires a very talented editor who makes the book come to life. When the author publishes the book as his own work, without cover credit to th editor, has the author cheated?
There’s a thriving industry for ghost writers. Is each of those publications a non-criminal conspiracy to defraud readers?
A high-end pipe organ has horns that sound like brass instruments. Have real trumpet players been been betrayed by the church because their parts were given away to a machine?
It’s 1985 and you’re a journeyman novelist who has just finished your fourth back-to-back rewrite of your manuscript on your once-high tech IBM Selectric typewriter–the rewrites necessary because of your agonizing scouring of the ms in search of typos and story weaknesses. When the novelist down the block has used his word processor to do electronic cutting and pasting, thus saving himself countless hours of time, do you feel cheated? Do you swear to stick with your typewriter because that’s what REAL writing is?
Technology’s a bitch. And it’s unstoppable. Do you use spell check? That’s AI. That tech is a spectrum, and I’m not comfortable with it’s implications, either, but it seems shortsighted to submit misspellings simply because changes came from a robot.
If you turn in a manuscript written the old fashioned way, and the editor says it’s 10,000 words too long, do you reject the notion of feeding the ms into an AI program and asking if to suggest where to change? Is that fundamentally, morally different than taking advice from your bestselling nextdoor neighbor?
From what I can tell, AI doesn’t actually DO anything; it responds to prompts, returnning results that may or may not be any good. It’s up to the prompter (authors in this case) to decide if the results are good or bad, and tweak it as they see fit. When it’s all done, the author owns the result and readers will judge accordingly.
With regard to the subject romance writer, I think the ineffectiveness of AI as a substitute author is showing itself. Total sales of 50,000 copies from TWO HUNDRED books is a pretty sucky number. Perhaps that’s why she hides behind so many pseudonymns.
Finally, I apologize for not respoknding to the many engaging responses to my post about AI yesterday. I’ve been traveling to promote my very traditionally written next Jonathan Grave book, SCORCHED EARTH. That explains why I am writing this at 5:10 am from an airport departure lounge . . .
John, I don’t have a problem with AI and use it often. My problem with the author was her hiding behind her pen names and not revealing to her readers that the books were AI generated. Or to Amazon.
I also agree they were probably pretty sucky, but unless she reveals the pen names we’ll never know for sure…
Readers with moral backbones lose trust in writers who use AI but don’t disclose.
Genuine creativity is incongruent with plagiarized work, entire editor rewrites, and ghostwritten novels.
Using AI as a tool and not disclosing dupes readers and deceives writers.
The many lawsuits (e.g., copyright infringement, suicides, fake news, identity theft, fraud, and more) showcase the issues and harm.
It’s self-deceiving for a writer to ignore the spiritual, emotional, mental, and physical harm caused by AI’s many models. Ignoring the harm, shouts so loud that a growing number of readers can’t hear a word those authors are saying.
I agree, Grant. Not disclosing the books were AI generated is cheating in my books.
Ugh. Sadly, prompters (writers write) like Hart are flooding the marketplace with garbage. I bet few disclose the books are written by AI. The personal costs are too high for me. And I would never slap my name on a book written by a bot.
Yes, I agree, it’s cheating readers. One of the most special parts of writing is living a scene. After publication, readers live those same scenes. There’s something magical about that. Plus, I enjoy using my brain to create.
I doubt this writer ever felt the satisfaction of pouring her heart into writing a book and feeling the satisfaction of finishing it, Sue.
Probably not. What a shame.
Amazon has a checkbox where you’re supposed to reveal whether you used AI in any part of the book’s creation. I wonder if she checks it, and what Amazon does.
There is a stark difference between using AI “tools” (e.g., editing, research, marketing) and having AI produce the actual prose on the page. Those offering the latter as their own work cannot be termed a “writer” or “author,” and presenting themselves as such is a deception. A real writer has put in the years-long hard work of learning the craft and writing words, spending months to hammer out a draft, then months more figuring out how to make it better (this is where a good editor or beta reader can step in).
But now someone who has never written a full-length novel in their life can, in an hour, have a finished product and throw it up (I use that term advisedly) online. There are scads of such adding to the tsunami of slop every day.
Is there an “in between” zone where a talented writer can speed up their process using AI? I suppose there is, but we need Rod Serling’s voice to warn us about the dangers (see Dale’s comment to yesterday’s post).
I believe readers will be able to tell the difference between an AI generated story and one that a flesh and blood author creates from the heart–a least at the present time.
I believe this, too. (At least for the present time!) What I most fear is what AI is doing to short-circuit the developing brains of our young. Unless you’re trained to read, think and write analytically during the school years, you won’t be able to think at all later on, when it truly matters. But you will be told what to think.
Insightful, Jim. From my interactions with writers, few have taken the time to understand what goes into today’s AI. Fewer still delve into how many legal, privacy, and moral issues we cede to machines when using AI.
Is there an ‘in-between’ zone? The decisions and actions of AI executives suggest not. The headlines of their misdeeds could serve as a caution. However, even if read out loud in Rod Serling’s voice, writers would need ears to hear with noise cancellation turned off.
Perhaps a warning label? CONTINUED USE CAN CAUSE MORAL BANKRUPTCY!
Well said, Grant!
“Unless you’re trained to read, think and write analytically during the school years, you won’t be able to think at all later on, when it truly matters. But you will be told what to think.”
Nailed it, Jim.
Jim, you are spot on about short-circuiting the brains of your youth. And to think, this is where our leaders will come from.
So-called A.I. is definitely on our minds these days. My post for Saturday, which I began working on before John’s post yesterday, is a response to Large-Language Models AKA ChatGPT and its ilk.
While I’m very concerned about the energy and environmental costs of using these LLMs, I agree Hart can use them to “generate” what I call “book-shaped objects,” but she’s not a writer. A writer puts thought, experiences and emotion into what they write. Writers sweat blood, at times at least, and we live the writing of our stories and novels. They focus our awareness and occupy our minds.
“Generating” text is not writing. And for me, writing includes things like book descriptions, reader newsletters, social media posts, etc. I want me in all of those, not probabilistically generated text.
I write to create stories and books and share those with readers, hopefully providing them with a fulfilling reader experience. It’s a connection between writer and reader, something “A.I.” can’t do.
Amen, Dale! The relationship between reader and author is a special bond, based in trust.
Well said, Dale!!
As a point of reference, here’s what Amazon KDP itself says about this in its submission process (with no word about penalties; just a Yes or No selection):
“… We require you to inform us of AI-generated content (text, images, or translations) when you publish a new book or make edits to and republish an existing book through KDP. AI-generated images include cover and interior images and artwork. You are not required to disclose AI-assisted content. We distinguish between AI-generated and AI-assisted content as follows:
AI-generated: We define AI-generated content as text, images, or translations created by an AI-based tool. If you used an AI-based tool to create the actual content (whether text, images, or translations), it is considered “AI-generated,” even if you applied substantial edits afterwards.
AI-assisted: If you created the content yourself, and used AI-based tools to edit, refine, error-check, or otherwise improve that content (whether text or images), then it is considered “AI-assisted” and not “AI-generated.” Similarly, if you used an AI-based tool to brainstorm and generate ideas, but ultimately created the text or images yourself, this is also considered “AI-assisted” and not “AI-generated.” It is not necessary to inform us of the use of such tools or processes.
You are responsible for verifying that all AI-generated and/or AI-assisted content adheres to all content guidelines, including by complying with all applicable intellectual property rights.”
It’s really easy for an author to ignore these guidelines, and evidently many do, Harald.
Thanks for the elaboration. All I’ve noticed is the check box, and I commented above before reading your response.
Writing books has been one of the hardest and most satisfying things I’ve ever done. That’s where the joy comes from. I’m a great believer in the positive results of hard work, and I find those results from the challenge of sculpting—or to refer to a recent post by John—composing a story. That can’t be replaced by prompting AI to create a book.
In addition, writing is a self-revelatory experience. Like Flannery O’Connor said:
“I write because I don’t know what I think until I read what I say.” Who would want to give that up?
Love the quote, Kay. So true!
I write for that same joy, Kay. There is no high quite like seeing the images in your head come to life on the written page.
Romance readers are easily the smartest readers of genre. Every romance plagiarism case in the last years has been found by readers. They can also spot phonies and pretenders who write romance but don’t read or love romance. AI is not a good choice for writing it.
I agree Marilynn. It’s probably why she didn’t sell many books. Even so, she added to the many awful books out there.
One of our rabbis proudly uses AI to create sermons. He has for about a year now. AI makes mistakes. It is fast, but does it convey the message he wanted? Doubtful.
I get newsletters from someone who uses AI for 80%+ of their work. The newsletters are error filled. He then gets huffy when his (AI’s) errors are pointed out. I asked the question, If AI is sending out emails and doing most of his work, what is he being paid for?
I agree, Alan–what is he being paid for? I doubt if I would keep reading newsletters if they were obviously written by AI. I tried it once and deleted the newsletter. It didn’t sound anything like me.
As long as we’re talking about what AI might be doing to people’s minds–especially kids’, I’d recommend reading Michael’s Connelly’s latest Lincoln Lawyer novel –The Proving Ground.
Mickey Haller pivots from criminal defense to public interest litigation, filing a civil lawsuit against an artificial intelligence company whose chatbot allegedly encouraged a sixteen-year-old boy to kill his ex-girlfriend.
Her experiment was to see what it would take to go from zero to $100k of income with no advertising. From what I understand, every novel was read and had a line edit pass. The number of pen names were need to not only separate the heat levels but to allow publishing multiple books during one week. The experiment lasted around 8 months.
As far as data centers go, none at this time are 100% dedicated to AI. Most house Amazon Web Services, streaming, and those apps on your phone. Increased demand for streaming and cloud apps drives the expansion as much as demand for AI.
If you want to raise the quality of AI output from average to excellent, you’re going to put in as much time as you would do it yourself because it needs revision and rewriting.
However, the time needed to bring the quality up is going down every three months. From what I’ve read/seen, Claude Opus 4.6 outputs so very good prose. What was true 6 months or a year ago is no longer true now.
Interesting, Fred. Still, it bothers me that 1, she didn’t tell Amazon or her readers the books were AI generated, and 2, she doesn’t give her pen names, so no one reading the article can check and see how good or bad the books are.
As a reader who already has a hard time sifting through and finding fic I want to read, AI generated fic will only make that issue worse.
I know we’re focused on fiction here at TKZ, but anybody have any inputs as to whether this is affecting non-fiction titles as well? Talk about scary.