Every court worth the name runs on a rule older than the printing press. The punishment has to fit the offense. A man who steals a loaf of bread doesn’t hang, and a judge who sent him to the gallows would end up on trial himself.
Online writing groups run their own courts now. There’s no judge, no evidence and no appeal, and the sentences keep getting longer.
This week the charge was using AI. The sentence was castration.
The charge, the evidence and the sentence
A member of a big creative writing group on Facebook posted that she’d joined to be around fellow creatives and was sad to see so much AI in her feed. Fair enough. It’s her feed, and she’s allowed to dislike what’s in it.
The replies did the rest. Within a few dozen comments, members had announced that every group should ban AI, that people who use it aren’t writers, that they don’t get to call themselves creative, and that they’re talentless hacks who just want to steal. A man who said he used AI to catch redundancy and grammar was told to go out, get drunk and do acid for inspiration.
Then a commenter typed it. People who use AI deserve to be castrated.
Read that sentence again. It doesn’t say people who pass off machine output as their own work, or people who cheat in contests. It says people who use AI. That covers the man fixing his grammar, the student checking a citation, the retiree who asked for help organizing her notes, and me. I use AI every day as a digital assistant that helps me write, and I still do the writing. According to that commenter, I should be mutilated for it. Is that who they meant? It’s who they said.
Nobody in that thread had been caught at anything. Several members said they hadn’t seen any AI in the group at all. One asked, sincerely, how anyone could tell. Another dared the accusers to share their magic AI detector. The original complaint was about something half the room couldn’t find, and the sentence handed down was surgery.
How a writing group argument turned into a call for violence
The castration line didn’t come out of nowhere. The same commenter had warmed up on someone else first.
A member had written that AI is here like sliced bread, so get over it or leave the group. Blunt, maybe rude, and not about anyone in particular. The commenter answered with a guess about the man’s politics, then followed it with a crude sexual insult I won’t repeat. The man pointed out that he’d said nothing political at all.
A little later came the line about castration.
When another member replied that cheaters deserve condemnation but everyone else can use the technology for whatever it’s good for, the commenter said its only use is destroying the earth and finished with a sneer about the man’s race and gender.
Count the steps: a guess about a stranger’s politics, a sexual insult, a call for mutilation and a slur about the color of a man’s skin. Four escalations from one person in one thread, and not one of them had anything to do with writing.
When the punishment someone calls for has nothing to do with any possible crime, the argument was never about the crime.Share on X
Why do people wish violence on writers who use AI?
Because the label did its work first. Once somebody is an AI user, a thief, a hack, he stops being quite a person to the crowd, and the ordinary rules about how you speak to a stranger switch off. I described that mechanism in Dogpile. This thread shows where it ends when nobody steps in.
Escalation pays. In a pile-on the comments compete, and an insult is only the floor. Each new reply has to top the last one to get noticed, and once thief and hack have been used up, there isn’t much room left above them except the body.
And the people who say these things don’t expect anything to happen. Nobody who types “deserves to be castrated” plans to see it done. So it’s cheap. It costs nothing, it draws attention, and the target is a category with no face attached.
Words like that don’t stay inside the thread, though. Someone reads them who runs a grammar checker and wonders whether to share her story at all. Someone who writes in a second language and leans on a tool to get the idioms right learns that this room thinks he deserves to be maimed. The commenter moves on to the next post. The reader doesn’t. (See also why the AI pile-on hits disabled authors hardest)
Is it ever acceptable to wish harm on someone for using AI?
No.
Every reasonable thing the anti-AI members said now sits a few inches from a call to mutilate people.
There’s no version of that question with a different answer. You can hate AI. You can refuse to read any book that touched it, ban it from your own group and block everyone who posts an AI cover. All of that is your choice. You can’t wish mutilation on strangers over software and expect anyone to take you seriously about anything else. Would you say it to a stranger’s face at a library reading?
The anti-AI members of that group had real points to make, and some of them made those points well. One member warned that a hunt for AI text could turn into a witch hunt, since a lot of the supposed clues are plain bad writing, and decided to judge covers only. That’s careful and honest. The group’s rule about AI is a fair subject for debate, and people on both sides of it have a case.
The castration line buried all of it. Every reasonable thing the anti-AI members said now sits a few inches from a call to mutilate people. Anyone looking for a reason to dismiss them just got one, handed over by somebody who claims to be on their side.
It’s a tool. Use it or don’t.
The writer decides how to write, the reader decides whether the book is any good, and nobody gets to hand down a sentence.
The silence in the gallery
The only direct reply to the castration line was that calm note about condemning cheaters. Polite and reasonable, it walked straight past the call for violence.
Nobody said that’s too far. As far as the thread shows, nobody in charge removed it. Dozens of members had plenty to say about AI covers, acid trips and octogenarians shaking their fists at cell phones, and a call to castrate people over a piece of software didn’t rate a single objection. Where were the admins? Where was everybody who’d spent the morning arguing about acid?
That silence is the gallery’s verdict. In a real courtroom the gallery can’t vote, but online it votes with every scroll. A line that stays up without a word against it has been approved, whether anyone meant to approve it or not.
I’ve been on the wrong end of these rooms. In Facebook writing groups I’ve been told to go read the FAQ and asked why I ask such stupid questions, and any mention that I use AI or like it gets me jumped on. I don’t join writing groups anymore. They’ve always been extraordinarily hostile, as I wrote in why writing groups are so hostile, and the anti-AI crowd is just the newest of the anti-crowds. AI gave the hostility a fresh target. It didn’t create it.
What should a writing group admin do about violent comments?
Delete them. Remove the member if it happens twice.
Admins write a group’s real rules with what they leave up. A group can ban AI posts or welcome them, and either one is a defensible choice. A group that bans AI covers and leaves up a call to castrate the people who make them has its priorities upside down. The rule doesn’t need to mention AI at all. No wishing harm on anyone, for any reason, ever. A group that can’t enforce that much has stopped being a writing group and become an audience for whoever is angriest.
Members carry part of the job too. You don’t have to argue with the person who wrote it. Reply once to say it’s over the line, report it, and leave if nobody acts. A room that tolerates a call for mutilation will tolerate anything that comes after it.
Who gets hurt when the rhetoric outruns the offense?
The people the comment names don’t get hurt much. The man who uses AI to check his grammar isn’t going to quit over one stranger’s fantasy, and he’ll probably laugh about it later. The ones who get hurt are the quiet members reading along, the ones who never post. Picture a beginner who uses a grammar checker and was working up the nerve to share her first chapter. She now knows what this room thinks of her. She won’t post. She’ll keep the chapter in a drawer. Nobody will ever know she was there.
The cause gets hurt too.
Writers who worry about AI, about income, about training data or about being buried under a flood of low-quality books, have arguments that deserve a hearing, and they’re making them in the same rooms. Every time the loudest voice calls for mutilation, the rest of them pay for it.
People stop listening to the whole side, and the reasonable ones get lumped in with the commenter who wanted the knife. I’ve laid out both sides of the theft argument in Is writing with AI theft?, and none of it needed a threat to make its point.
And the group gets hurt. A writing group exists so writers can show each other their work and get better at it. Once members learn that showing work can earn a call for violence, they stop showing work, and what’s left is a room full of people arguing about who deserves to be punished.
| Who | What happens to them |
|---|---|
| The people the comment names | Not much; the man checking his grammar won’t quit over one stranger’s fantasy |
| Quiet members reading along | A beginner with a first chapter learns what the room thinks of her and never posts |
| Writers with real worries about AI | Their arguments on income and training data get lumped in with the commenter who wanted the knife |
| The group itself | Members stop showing work, leaving people arguing over who deserves punishment |
| Admins who stay silent | A line left up without objection has been approved, meant or not |
My take
No one in that group had been caught at anything, and the sentence was mutilation. An admin who leaves a line like that standing has told every member what the room tolerates. Delete it and remove the poster. The quiet beginner reading along gets hurt, and so does every reasonable critic of AI who gets lumped in with the knife. Argue about AI as hard as you want, but a threat ends the argument for everyone in the room.
The punishment has to fit the crime
Go back to that courtroom. The charge was using a piece of software. No evidence was offered, and in half the room there was nothing to find. The sentence requested was mutilation, and the gallery let it stand without a murmur.
Any court that ran that way would be shut down by the end of the week. Writing groups don’t have anyone to shut them down, so the job falls to the admins and the members, and to anyone who still remembers that the people on the other side of the screen are human beings.
The AI and Writing Hub collects my articles on how writers use these tools and how the arguments around them play out, including the rest of the AI Derangement Syndrome series. If you’d rather work on your book with someone who’ll judge the pages and leave your tools alone, my book coaching service starts with a conversation.
A loaf of bread doesn’t earn the gallows. A grammar checker doesn’t earn the knife.
Sins of the Internet is my book on what the platforms did to us. It covers how anger got amplified over empathy until a line like this one passes for discussion.
