Post an AI image and unfriend me. That’s the ultimatum, and an ultimatum is a strange way to keep friends.
A post came through my feed last week, and I read it twice. The man who wrote it told his friends that if they ever posted an image they’d made with AI, they should unfriend him first. He added that anyone thinking about commenting with an argument should skip it and go straight to the unfriending. Then he asked if everybody was clear.
I understood the feeling behind it. I didn’t like the delivery, and I’ve been thinking about why ever since.
The sign on the diner door
You’ve seen the sign. It’s taped to the door of a little diner, hand-lettered, a bit grease-stained: “No substitutions. Don’t even ask.”
The cook probably has good reasons. The kitchen is tiny, the recipes are his, and he’s sick of people treating his menu like a suggestion. His regulars know all that. They read the sign, grin, and order the usual.
Everybody else reads it as an accusation. Before they’ve said a word, the sign has decided they’re the kind of customer who’d cause trouble. Some of them walk in anyway, wanting to test it. Some of them walk away. Nobody who reads that sign learns why the food is worth eating the way the cook makes it. That post was the sign on the door. It isn’t AI Derangement Syndrome in its full form. He never called anybody a thief. It runs on the same reflex, though, where one tool decides who you are before you’ve said a word.
Watch: The Real Reason Why Everyone is Hating AI, DM Quil
What he meant was reasonable
Down in the comments, a friend asked whether he’d rather people post Photoshopped pictures instead, since Photoshop is full of AI these days.
His answer was thoughtful. He said the problem is that AI images all look the same, they take no skill to make, and they stop people from learning a craft or hiring someone who has one. If you still have to build something yourself, he said, he’s all for it, AI features included.
Later, someone pointed out that we had a sameness problem long before AI: templates, swipe files and formulas that everybody copied because they seemed to work. He agreed, and put it well. Now, he said, we’ve handed the template to a machine and asked it to fill in the blanks too. Use AI to support your skill, and keep the thinking behind it yours.
I agree with nearly every word of that.
I use AI every day as a digital assistant in my own writing, for transcripts, summaries and first-pass outlines, and I still write the book. When the tool starts doing the thinking, the work goes flat. You can see it in the feed every day, in images that all share the same glossy skin, the same impossible light, the same vacant stare.
If he’d posted that second comment instead of the first post, I’d have hit the share button.
Why do ultimatums about AI on social media start fights?
Because an ultimatum turns an opinion into a line, and a line invites someone to step over it.
You could watch it happen in the replies.
One friend said he’d post so many AI images and tag him in all of them. The author answered that he’d block him, like the others. Another asked, deadpan, whether he could teach him how to make an AI picture so he could send it over. Most of it was ribbing between friends, and it was funny. It also showed exactly what the post invited. Nobody in that thread talked about craft. They talked about the dare.
I’ve seen the same thing on LinkedIn, in writing groups and in the comments under news stories. The words change.
The move doesn’t. Someone plants a flag, a few friends salute it, and a few others show up to knock it over, and every one of them is performing for an audience that isn’t the person they’re talking to.
That’s the small version. The large version is worse. When enough people publicly declare which side they’re on, every later post about AI arrives at a crowd that has already picked a team and promised to enforce it. I wrote this week about a mother who asked a simple question about publishing her daughter’s bedtime story with AI illustrations and got called a thief. The people who piled on her weren’t monsters. Most of them had made posts like this one somewhere along the way. They’d drawn their line in public, so when she stood near it, they had to prove they meant it.
That’s the pattern I keep writing about in the Dogpile series. A pile-on rarely starts with the person at the bottom of it. It starts with a lot of people who already decided how they’d react before anyone gave them a reason.
And a loyalty test over a software tool is a ridiculous thing to demand of your friends. It’s a tool. Use it or don’t. Nobody ends a friendship over which brand of hammer somebody bought, and a picture generator doesn’t deserve more drama than a hammer.
The loyalty test doesn’t stop at friends. An advice column on a big literary site answered a reader whose mother kept suggesting AI tools by telling the reader to argue with her, tell her she’s wrong, and treat the fight as important. That’s a columnist advising someone to pick fights with their own mother over software.
The mother’s offense was taking an interest in her kid’s writing. If you don’t want the tool, say you like doing it yourself and change the subject. Your mom is on your side, and no hammer is worth ruining Thanksgiving over.
Nobody ends a friendship over which brand of hammer somebody bought, and a picture generator doesn’t deserve more drama than a hammer.
The site promoted the column on Facebook with the line “Don’t talk to us if you use AI in your writing,” over an illustration that plenty of commenters took for AI art. The thread went the way you’d expect. People called each other assholes, hypocrites and thieves, and one reader who didn’t like AI either asked why a literary site would throw gasoline on that fire. Good question. Outrage gets clicks, and a column that tells you to fight your mother gets a lot of them.
Is it wrong to post AI-generated images on social media?
Not by itself. It depends on what you’re passing off and where you’re posting it.
If you generate a silly birthday image for your brother, that’s not a moral failing. If you pass off a generated picture as a photograph you took, or as art you drew, that’s dishonest, and people are right to be angry. If you post AI art into a community of artisans who’ve asked, over and over, not to see it there, you’re breaking a rule you knew about, and I wrote about exactly that in Dogpiled Over AI Art at the Renaissance Faire.
The real rule is simple. Say what it is, respect the rooms that have asked you not to, and don’t let the tool do the part that’s supposed to be you. That covers almost every case. It also leaves room for friends to disagree without anyone being told to leave.
My take
Ultimatums turn a fair point into a dare. His complaint that AI makes everything look the same had merit, and nobody heard it because it arrived as a loyalty test. Make the argument and drop the threat. Label anything generated, and keep it out of groups that ban it. Write the personal parts yourself. Software isn’t worth losing a friend over.
Take the sign down
The point underneath that post deserves an audience. AI is flattening the look of everything, and it’s tempting people to skip the years of learning that make work worth looking at. That’s worth saying to everyone you know, including the people who use these tools every day.
An ultimatum can’t say it. It only tells your friends which ones are in trouble.
If you care about this, post the second comment.
Tell people what you value and why. If you want the longer version of how I think about these tools, the AI & Writing Hub lays out where AI helps a writer and where it breaks. And if your own work is caught between the two, that’s a conversation I have with clients every week.
The cook’s food is still worth eating. Take the sign down, and put up a note that tells people why he makes it by hand.
Ultimatums like this feed the machine that rewards outrage. Sins of the Internet looks at how the platforms chose anger over empathy and why furious users became the profitable ones.
