
Recently, in my morning scroll through headlines (to make sure the world didn’t explode or implode while I was asleep. I like to keep an eye on it. It’s tricksy.) I came across one that made me giggle. It was, paraphrased because I can’t swear I remember it precisely: Traditional Publishers Having Difficulty Screening Out AI Written Manuscripts.
I didn’t laugh like an hyena, but that’s because I have a heart of stone. Also because I’m old and I remember reading an interview in I think Locus Magazine with I THINK Gardner Dozois in the mid nineties, complaining that it was now much harder to tell when someone would be a real writer, since word processors made it much easier to present a clean, properly formatted manuscript.
If you don’t see the similarity, it’s because you’re not thinking of it right. Both of those complaints related in no way to the QUALITY of the writing, but only how it was produced.
I could have had a typing class in high school (I didn’t because I thought that would condemn me to secretarydom. Why I was scared of that I don’t know) and be an excellent typist and not be able to tell a story to save my life. The world is, I’d bet, full of such people.
On the other hand, two of the most gifted storytellers I ever met were so impaired by ADD and dyslexia that getting them to spell something right, or write it in the write order to make a sentence took effort. (I don’t know if either ever published. The solution to their problems these days would be dictation. They were fine speaking.)
“But Sarah,” you’ll say in that whiny voice I attribute to you when I want to explain that, yes, I thought through your objection, “typing and AI is completely different. Typing is just recording the story, and yeah, the editor was being dumb, but AI writes the story FOR YOU.”
I’m waggling my hand at you. Sure, AI CAN write the story for you. However, if traditional publishers CAN’T tell when a story was written that way, they need to spend three days reading through the latest “we make money on volume” genres on Amazon. I recommend Jane Austen Fanfic for the quick course. A day might do it. The stories are full of Claudisms (I call them that, because AIisms bends my tongue) that would make your eyes cross, and no, none of it has anything to do with too high a vocabulary or using m-dashes.
The most obvious ones are the construction “she/he/it was a woman/man/teapot who…” As in “She was a woman who had never batted at butterflies.” Or “She looked at him like a woman who routinely bapped at cats.” Or “His silence was the silence of a man who never tied his own cravat” or…. you get the point. AI can’t seem to set a scene without at least one of these.
Then there’s what I call the “Nots”. “She did not say yes. She did not say no. She did not say anything.” “He did not pick up the knife. He did not pick up the horseshoe. He just stood there.”
Honestly? Seriously, that second one is annoying as all get out, and everywhere.
Then there is the fact that objects have feelings and thoughts. “The house thought about it and determined not to let it go on.”
There are others, but I can’t remember them off the top of my head, because the coffee is still seriously outnumbered by blood in my veins. But there are a lot of them, and three days reading texts generated by AI will give you the screaming mimmies give you an adequate education to identify Claudisms in minutes.
And that’s without the more subtle ones, like: Distance doesn’t exist. Someone who is an hour away, or who has never met the characters is suddenly there, having breakfast with them; humans can have five or more parents; character traits? What character traits.
I’ll grant you they’re probably not getting it raw enough — maybe — that the subtle effects become obvious, but REALLY.
(BTW my favorite experience with this stuff so far was the story that had the pious disclaimer about never having used AI and never doing it, and then was FULL of Claudisms. One doesn’t know whether to admire the chutzpah or the hypocrisy, really.)
Unless trad pub is now complaining that they have to — horrors — read the manuscripts to find out if they’re any good, which is what the editor in the 90s was doing, and which, to be fair conforms with my experience of rarely even being read through, it’s a bit of a puzzle how it can be hard for them to figure it out.
But what, you ask, if the writer edited all that stuff out? Ah. A very sensible point.
Well, then… what does it matter? No, seriously. WHAT DOES IT MATTER? Who gives a very tiny fuzzy rat’s ass? And why should they?
If the thing is so thoroughly rewritten that it betrays none of the marks of the beast AI and it’s a good book they would otherwise publish, WHY DO PUBLISHING HOUSES NEED TO KNOW HOW IT WAS WRITTEN? Are they going to have the writer certify the first draft was handwritten using the quills of snow white geese and the blood of virgins? No? Oh, that wouldn’t matter if the final draft was typed and normal? Well! Gesticulates to the previous sentences: then neither should AI.
Again and again and again: if the final product is acceptable, why do they need to know how it was produced?
The answer, of course, is the same answer that causes what I call “AI witch hunts.” As in the reader or in this case the editor feels obscurely cheated if they think the manuscript was produced by an “easy” or “cheating” means.
Look, if the manuscript has no marks of AI, then the writer did a lot of work. It might be different work from what would be required for a manuscript produced another way, but it’s definitely still work. There was still knowledge and effort that went into the finished thing.
There might come a time when you push a button and out comes a story, but that time is not today and tomorrow is also looking unlikely. (And that’s without getting into plot. AI cannot at the moment plot anything complex, and most plots are from movies. Or, somehow, uber woke novels. I once asked for suggestions for short stories in a specific time line (hoping that the suggestions would make me so mad I got ideas. They made me mad, but they were so lame I got no ideas, and had to sleep on it.) It’s all movies, or violations of gender pronouns or cultural appropriation or– You don’t want to know.
Of course, this post is not precisely about that, or the terrible plight of traditional publishers actually HAVING to read manuscripts — oh, the horror! — though it still makes me giggle.
Most of what’s causing the AI panic is… bullshit.
There is the fear from all the sci fi movies of the past. Skynet is growing even now and is coming to get you. Me, who grew up with things like Martin Caidin “The God Machine” somehow find the movies unpersuasive and not alarming at all. But beyond that, I have a grip on reality and know that even if AI is/were sentient (look, there are moments) it doesn’t mean it’s immediately evil. It’s not like it’s possessed of the spirit of Karl Marx. It’s often incredibly stupid in the ways of a very bright toddler and so, yes, dangerous, but that’s not the same as “evil.” The risks and problems need to be mitigated and guarded against, but that’s not the same as “AHHHHHH, It’s going to kill us all.”
Thinking that “AI therefore Skynet” is insane, and means you need a blanky, a teddy bear and some warm milk. If you’re good we’ll give you a cookie. You’d be less scared if you’d been around for the introduction of computers into business. Computers were going to link with each other and rebel against humans! Bah. Nonsense and fairytales.
Then there’s the fear — abetted by Musk’s pie in the sky talk — that it’s going to run humans out of jobs. Humans are going to have nothing to do, except collect their government paycheck, grow their hair, dye it pink, and write bad poetry.
This is as likely as computers giving us paperless offices. Look, yeah, you might lose your job to AI. For now. Partly because your bosses have no clue what AI is and think it’s magical. But in the end what AI will do is create jobs. And allow the competent to do more, more rapidly.
(Even the fears that AI will control what people know of the past, of history, of reality, etc. all according to woke directives is zany. Oh, sure. Like the internet controls what you know. AIs are multiplying, and as soon as it escapes mega-corporate-control, so will the answers it gives, and the parts of history/etc. it has access to. Right now there are problems — though not as bad as they were even two years ago — but it’s getting better.)
What AI does is allow people who are good with tech to become entire “companies” of people who are good with tech.
And I have a feeling within the next year it will allow us humble story slingers to make movies on the scale of Hollywood, thereby dethroning that corrupt edifice. (Now, will it be at a price I can afford? Likely not. That will wait ten years should I live that long.)
On another level there are justifiable fears. Of AI, sure, since it’s the most visible point-of-the-spear development, but really of all technological development.
The human mind and spirit, not to mention nervous system, aren’t designed for the pace of innovation we’re experiencing. People in general don’t like to have their cheese moved. They want tomorrow to be more or less like today, for their lives to be like their parents’ lives only better.
Since the industrial revolution we’ve been moving much faster than that, and the last twenty years it’s gone to eleven. This causes a fear at a level we can’t describe when things around us and the way we do things change at a pass we can’t control, can’t slow down, and can’t very well cope with. The fear is not rational and not fully articulated, but it is perfectly understandable.
In the past when things changed this fast, your ancestors had been invaded and were about to be killed in batch lots.
I do understand that. It worries me a little.
BUT that’s not here nor there nor AI. It’s just the environment we have to cope with right now.
It is very important to distinguish that real fear from weird tourette-like screams of “AHHHH AI WILL KILL US ALL.”
And it’s very, very important not to go on AI witch hunts. If you go looking for AI you’ll find it. PARTICULARLY where it isn’t and never was.
And real people will get hurt due to your imaginary fears.
So examine what’s really happening and your real motives before you go off half-cocked.
And as for trad pub? ROFLSHIPMP. Maybe I don’t have a heart of stone, after all.























































































































































































