Last week I wrote a piece that went viral.
Rainbow colored AI turds were re-shared, re-stacked, and re-posted hundreds of times across the internet:
Even some of the reposts went viral:
So when it came time to write this week’s issue of The Digital Contrarian, I did what every good writer does the moment they gain momentum from writing something that gets a response like this.
I decided now would be the perfect time to…. take a break from writing!
I told myself it was fitting that [Issue #104] went viral. Because writing one hundred and four issues meant I’d written a piece every single week for exactly two years straight, without missing a week.
Naturally, I deserved at least a week off. (Right?)
On top of that, “It’s Father’s Day!” The voice of the Resistance whispered in my ear, “If there’s ever a day to take a week off, this is it, buddy.”
(Happy Father’s Day to all the other dads out there, BTW…)
So, feeling the pressure to live up to last week’s issue, instead of writing, I procrastinated by spending hours and hours reading every human comment across every reshare, restack, and repost.
Over a thousand of them, in fact.
And somewhere along the way, in the process, I stumbled on something…. unexpected.
1 | Why is using AI in your writing making more and more people angry ??
A few weeks ago, The Wall Street Journal ran a story that captured the strangeness of this dynamic perfectly. It’s gotten to the point where writers are going to all sorts of crazy extremes just to prove they aren’t using AI in their writing.
For example, some writers are starting to leave strategic typos in on purpose. Some are now using “aggressively casual” language to prove it’s them, not AI. Others are swapping em-dashes with two smaller dashes just to make it look like it was “hand-crafted.”
One writer calls the whole climate right now “the new McCarthyism.” People are demanding proof of something that can’t be proven.
But the question is this:
What is it about using AI to write that’s suddenly making people so angry?
I was curious, so I did the only reasonable thing a 43 year old father would do on Father’s Day eve:
I pulled an all-nighter to find out.
2 | Message from Your Market. Not Your Mind.
The whole foundation of the ASK Method, the thing I taught for nearly two decades, comes down to a single idea:
Your message should come from your market, not from your mind.
In other words, you don’t sit behind your computer and try to guess what people are feeling (or worse, let AI try to tell you…)
No. You actually go out and gather their exact words, in the wild, and hunt for the patterns hiding inside them.
The unfiltered, unsanitized, natural consumer language your market is using right now is the gold you’re looking for. And it requires digging.
You can get this language with a survey, or with interviews, or in some cases…
With the comment section on a post that unexpectedly goes viral…
So this is exactly what I did.
In the last 24 hours, I read over a thousand comments spread across dozens of posts and reposts across LinkedIn, Substack, Facebook, and Instagram, because I wanted to see what is it about AI writing that’s making people so angry right now.
I wanted to see if there was a common set of triggers. If there were certain reoccurring themes.
And what I found was this:
There are ten (10) themes that seem to keep surfacing over and over again.
(And BTW yes, I used AI to help me do some of the grunt work on this project. If that triggers you, you’re not gonna like what comes next…)
3 | Ten Things They Hate About You, AI.
“I know you can be overwhelmed, and you can be underwhelmed, but can you ever just be whelmed?”
– Gabrielle Union as Chastity Church in 10 Things I Hate About You
Theme #1 – “I can feel it’s fake, even when I can’t prove it”
The most common reaction that showed up again and again in the comments: People trust an instinct they can’t fully articulate. But AI writing leaves them feeling (for lack of a better term)… whelmed. Here’s a sample of comments:
- Will Soprano: “I read a couple of articles yesterday by a legitimate expert and within the first three sentences I could tell it wasn’t his voice, it was AI…”
- Jenni Salmela: “The text is always a little thick and monotone. Often repetitive. Like have you ever met someone who keeps repeating what they’re saying? They’re just filling the space with noise.”
- Sam Eitzen: “If you know writing, you just FEEL the AI in it when it’s there.”
The trigger: AI writing just feels off. And almost nobody can tell you exactly why.
Theme #2 – “Who am I even talking to?”
The story that resonated most for many in my piece that went viral, was the “message slop” experience I wrote about getting passed off to a chatbot mid-conversation by someone I had interacted with many times before. Readers brought up their own versions. A sample of comments:
- Ed Ervin “My wife & I received an AI sympathy card about 2 years ago. All the words were there but read detached from its intent.”
- Bryant Duhon: “While I despise the “things were better back in the day” thinking of so many folks when they hit 50ish, yeah, if using AI in a messaging app to have a “conversation” makes me “old”; I’ll happily wear that badge.”
The trigger: in a personal exchange, AI reads as a lack of respect. (This is the strongest argument for the disclosure rule we’ll get to later.)
Theme #3 – “You couldn’t even be bothered”
People can sense when you didn’t do the work. The tool isn’t so much the problem as is the perceived absence of care. On top of that, the writers who spent years honing their craft feel robbed with a sense of “stolen valor” from those who claim AI writing they publish as their own. A sample of comments:
- KC Procter: “Outsource the menial work to AI all you want, but don’t outsource your thinking.”
- Meghna N.: “AI is not the problem, it’s the sheer lack of effort to make it even slightly human.”
The trigger: low effort reads as low awareness. Slop is evidence that you didn’t care enough to do the thinking.
Theme #4 – “I can feel it changing how I write.”
Some people are worried about their writing turning into the very thing they hate. There is fear that if you read enough slop, slop will seep into your own writing. A sample of comments:
- Naomi Veak: “Humans trained AI, and now AI is training US.”
- Michael Baun: “It may not take words away from you, but it may give you words you never intended to use. It will corrupt a writer’s agency.”
The trigger: people can feel themselves not just losing their voice, but their identity and in some cases, even their agency as a human being.
Theme #5 – “Everything sounds the same now.”
Over a decade ago, an art project blended hundreds of real human faces, country by country, into a single “average” face for each one. The average face feels incredibly generic.
The same thing is true with writing. It’s the imperfections that make writing feel alive. A sample of comments:
- Joe Callender: “The mean of human output. The most averaged location across everything humans ever wrote.”
- Mark Roderick: “Almost every post sounds alike. You don’t hear the voice of people, just the voice of ChatGPT.”
The trigger: when everything sounds the same, everything blends together, and nothing stands out. Language becomes a boring sea of gray.
Theme #6 – “Maybe I’m the old man now.”
One of the most quoted lines from the piece I wrote was: “Maybe I’m the only man now?” There is a sense of grief for the old internet that many feel we’ve lost. One that was built by and for humans, not bots. A sample of comments:
- Dan Wardle: “The internet needs a place curated by humans. Was 2002 to 2012 ish a golden era for the information age?”
- Marq Mervin: “Will there be a point that people slow down, disconnect from this digital world, and reconnect with themselves?”
- Madalina Buzdugan: “I think I might be the old man too.”
The trigger: nostalgia and loss. “Old man” is the joke people quoted again and again from “The Shape of Enshittification” to express their mourning for something we’ve lost: an internet that once felt human.
Theme #7 – “AI is useless without the human created input it has unfairly stolen.”
Last year a trend went around where people asked ChatGPT to recreate an image exactly, “don’t change a thing,” then fed each new version back in and ran it again. A hundred times. Someone did it with a photo of Dwayne “The Rock” Johnson. The first few copies look close. By the end, The Rock has melted into a smear of unrecognizable abstract shapes and colors.
That is what AI trained on AI looks like. A copy of a copy of a copy, drifting further from the real thing with every pass. Researchers call it model collapse. As AI fills the internet, the well of real human writing runs dry, and AI is left recreating its own recreations. A sample of comments:
- Janelle Leonard: “AI does not ‘create’ anything on its own, it steals from human sources. If there is no source material it is useless.”
- Cathy Brown: “The risk of model collapse is real. Future models increasingly learning from synthetic rather than human generated data.”
The trigger: When AI is trained on AI, it begins devolving into something unrecognizable. AI is replacing humans who write without compensating them for their work, yet without the human-written content it has stolen, AI is nothing.
Theme #8 — “Wait, did you use AI to write this?”
There’s a LinkedIn writing style that some people call Broetry.
One sentence per line. Lots of white space. Short punchy fragments stacked down the screen for momentum (ahem).
It is built to be scrolled, and somewhere along the way it became a tell, shorthand for engagement bait and, lately, for AI.
And for some folks, that one-sentence-per-line format is very triggering.
I’ve written emails this way for the last two decades, way before it was cool, and way before AI.
In fact, here’s an egregious example from way back in 2014:
Several readers from the Broetry Brigrade expressed their impassioned disapproval of this stylistic choice.
They accused the guy writing a piece about detecting AI (written in the one-sentence-per-line style that many readers code as the AI tell) as having written it using AI. A sample of comments:
- Karla Starr: “I wondered why all of your sentences were on their own line. That reads as AI to me.”
- Stephen Gillie: “This feels like it was written by AI… Have to scroll forever just to read a paragraph. Which is so ironic that it’s sad.”
I have to admit, like the writers in the Wall Street Journal article, I felt the need to go to extremes to prove that I actually wrote the piece myself.
(No really! The short paragraphs are designed to be more readable, especially on a mobile device where big blocks of text can suppress readership because of how intimidating and un-friendly they look etc. etc.)
But alas, I didn’t have the time to respond to each and every comment and include timelapse footage of me in my underwear sitting in various places, in various poses, typing the piece myself over a 14-hour stretch to prove that it was “human written.”
(Next time I’ll remember to turn the camera on…)
The trigger: perceived hypocrisy. Trigger happy readers are looking for every reason to dismiss your work as AI.
Theme #9 — “It’s just a tool. Get off your high horse.”
I’ll start with a confession. I fall into this one.
Some people can’t cook, in part because they have no taste. Some can’t sing, in part because they’re tone deaf. And a LOT of people in this world simply can’t write.
They can’t get a clear thought onto the page. For them, AI is their auto-tune. It takes someone who could never string together a set of coherent sentences and makes their writing sound… passable.
Here’s the point that the dissenters raise. When you sneer at that, are you defending good writing, or just being a snob?
We have car snobs… we have wine snobs… and we have writing snobs.
Fair enough.
The love of wordcraft is real. If you’re a fellow Word Nerd and want to feel it, read Mark Forsyth‘s The Elements of Eloquence or The Etymologicon, IMHO two of the best books ever written about language.
The shaming, the dissenters argue, is a privilege.
For non-native speakers, for those with disabilities, for someone who never developed the ability to communicate clearly, AI can be a tool that helps overcome what may otherwise be a massive limitation in their life. A sample of comments:
- Richard H.: “nobody complains when digital photographers use light meters, autofocus, or Photoshop… I’m not going to judge anyone for using the tools available… unless they are using an LLM to tidy up their attack on those who use LLMs.”
- Shobha Garg: “crafting the right words under pressure is a skill that isn’t equally distributed. That’s the part most of the commentary on this topic misses entirely.”
The trigger: the moral framing that “AI users are lazy” reads as classist and ableist. To the ones on the receiving end, being called “AI slop” feels like being told they don’t belong.
Theme #10 — “Wait, this might be good news”
When “passable” writing is easy to produce, passable writing becomes worthless. Differentiated writing suddenly turns scarce. Scarce things become valuable.
The “Shape of Enshittification” piece is proof.
On LinkedIn alone, I grew my newsletter subscription base by over 800+ subscribers (a whopping 21,000%+) in a 48-hour period, all from a single human-written article. No paid promotion. Purely organic reach.
And the readers who read the piece all the way through?
Many landed somewhere hopeful. They realize that the flood of slop might be the best thing that ever happened to good writing.
- Glenn Marczewski: “more appreciation for artistic expression that can’t be faked… a counter-swing of the pendulum away from the middle of the road sameness.”
- April Jiménez: “I see more of a need than ever for real, human writing that’s interesting and engaging. And there are signs that clients are realizing this too.”
The trigger: Real is rare. And if you’re willing to do what’s rare, it’s almost always going to give you an advantage. This is the entire Return to Real thesis in a single sentence.
Which brings us to….
4 | Three Rules for Writing in the AI Era...
(AI Absolutists block your ears and close your eyes…)
It there are ten AI triggers that set people off, the obvious next question is:
How should we be using (and NOT using) AI in our writing?
In other words…
Are there RULES we should consider following based on all these insights?
I can see at least three.
First, this list of rules is NOT exhaustive. Consider it a starting point.
Second, let me clear out the AI absolutists. Some people will tell you never to touch AI, for anything, ever. I’m not one of them.
(Only a Sith deals in absolutes.)
Here are the three rules I see worth considering, based on what I’m taking away from this all-nighter diving deep into the research and analysis.
Rule #1: For information, yes. For point of view, no.
For pure information transfer, feel free to use AI freely. SOPs, recipes, meeting summaries, research summaries, etc. However, the moment AI starts ghostwriting your POV, your opinion, your argument dressed up as your own words, you’ve slipped into slop. You’re better off writing a clumsy paragraph that you wrote yourself honestly than a polished one AI came up with for you, that you try passing off as your own.
Rule #2: When it’s one human to another, disclose.
If you’re using AI to talk to a real person, one to one in a personal message, just say so. “Hey, English is my second language, so I’m running this through AI.” or “Hey, I’m super slammed today, so I’m using AI” or “Hey, I’m having a hard time saying what I’m trying to say, so I’ve had Claude help me.” It costs next to nothing, and it keeps the trust intact. Honesty and transparency are always winning strategies.
Rule #3: Is it helping you write, or helping you avoid writing?
Credit for this one goes to Brendan Hufford. If AI is helping you work through the process of writing to make it better, good. If AI is doing the work for you so you don’t have to, that’s exactly the kind of slop everyone can smell. Don’t sling slop downhill because it’s easier for you to ship. It’s almost always harder for your reader to swallow.
* * *
Those are three based on the research and analysis.
But I’m curious…
What else do you think is missing from this list??
SHOOT ME AN EMAIL to let me know…
* * *
Okay, I’ll leave you with that for now…
I really do need to get a few hours of sleep.
I am literally recording the audiobook version of Return to Real… tomorrow.
Yes. Like an insane person.
Have a great rest of your day…
And if you’re lucky enough to have your dad still around?
Tell him you love him.
Even if he was less than perfect and screwed up a ton when you were younger.
He probably did the best he could with what he had…
Maybe hearing those words from you will make his day…
Or his year.
Okay, that’s all I’ve got.
Remember to hug the ones you love.
And until next week,
Ryan :-)
