The AI Slop Survival Guide: AI Didn't Invent Bad Writing. It Made It Scalable.
LinkedIn now has a way to flag content that "Seems like AI slop."
That's quite a moment.
We've reached the point where there's enough low-effort AI content
online that platforms are giving people a way to complain about it. 😂
But AI didn't invent bad writing.
It made bad writing scalable.
Before AI, writing 20 bad posts took time. Now you can do it before
lunch.
What is AI slop?
AI slop is low-effort content created or heavily generated with AI, with
little human judgment behind it.
It can look polished. The grammar is correct. There are headings and
probably a neat framework.
But you finish reading and think: Did this say anything?
Since 30 July 2026, LinkedIn has let people flag any post as "Seems like
AI slop" from the post's three-dot menu. Flagging hides the post from
that reader immediately and quietly notifies the original poster in
their analytics, and LinkedIn's Chief Product Officer Hari Srinivasan
has called the feedback loop a top priority for improving its detection
models
(TechCrunch).
LinkedIn has also scaled back its own AI writing assist, replacing
"Enhance your post" with a more conservative proofreading tool that
keeps more of the original voice.
We know the signs
None of these proves AI wrote something, but people are getting tired of
the patterns:
- A dramatic hook that leads to an obvious point
- One-sentence paragraphs all the way down
- "It's not X. It's Y." over and over
- Words like leverage, unlock and elevate where nobody would say
them aloud - Five sections explaining an idea that needed two paragraphs
- A perfectly structured post with no story, example or opinion you
can attribute to the writer
And yes, apparently the em dash is guilty now too.
Poor em dash. 😂
The issue isn't punctuation. It's sameness.
Give AI a generic prompt and it's not surprising when generic writing
comes back.
AI is a tool
When the internet arrived, it made publishing easier. That gave us
incredible things and a lot of crap.
We didn't stop using the internet. We learned how to use it.
AI feels similar.
I use AI to... What I won't outsource
Edit grammar and typos What I believe
Structure a messy ramble My thinking
Research and find gaps Checking the facts
Challenge an argument My position
Accelerate my thinking Direction and judgment
I'm still directing.
AI has accelerated my thinking. I can give it a premise, ask it to
research, challenge what comes back, disagree and find a better angle —
that's the same instinct behind picking the best AI for content
writing
in the first place: the tool matters less than what you feed it.
But my thinking and foresight are part of what people pay me for.
English is my second language at home
This is why I'm not interested in shaming people for using AI to write.
English is my second language at home.
I've always had ideas and premises for things I want to write about.
I've also had people slide into my DMs to correct my typos and grammar.
😂
Now AI can be my editor assistant.
I give it what I'm trying to say and ask it to clean up the grammar,
question a sentence or tell me when something doesn't make sense.
The idea is still mine. AI helps me communicate it.
There's an irony here. For years, people corrected typos because
imperfect writing looked less professional. Now perfectly polished
writing can make people suspicious that it's AI.
Maybe the typo is becoming proof of life. 😂
I'm not suggesting fake mistakes. The goal isn't imperfect writing.
It's recognisable writing.
So "Did you use AI?" isn't the question I care about.
I'd rather ask: Did you have something worth saying?
AI didn't create the thinking problem
Ask AI:
"Write me a LinkedIn post about why personal branding is important."
What do you expect?
AI doesn't know what you've seen across 50 clients. It wasn't in your
workshop yesterday. It doesn't know which industry advice annoys you or
what you've changed your mind about.
Unless you tell it. That's also why sharing your expertise without
sounding like a
know-it-all
still comes down to what only you know, not what any model can generate
on its own.
People who don't know what makes writing specific, interesting or useful
can now produce more mediocre writing, faster.
AI didn't invent bad writing. It made bad writing scalable.
The AI Slop Survival Guide
1. Start with a premise
Don't start with "Give me 20 content ideas." Give it something you
think — the same starting point as writing a highly opinionated LinkedIn
post.
2. Give it material
Feed AI things it couldn't know: observations, voice notes, client
questions, stories, examples, experiments and opinions.
3. Direct it
Don't accept Draft #1 because it exists.
Tell it:
"That's generic."
"I don't agree with this."
"Cut the adjectives."
"You're being verbose."
"Why are all the sentences so short?"
Yes, I've said all of these while editing AI writing. 😂
4. Put the human details back
For this article, that's English being my second language at home,
people correcting my grammar in DMs, and how I use AI as an editor.
5. Delete
AI can make a point, explain it, then summarise the explanation. You
don't need all three.
6. Read it
Would you say this? Would your audience understand it? If you think, "I
would never say that," change it.
Potential graphic: PREMISE → MATERIAL → DIRECT → HUMAN DETAILS →
DELETE → READ
The button won't fix bad writing
A button can't teach someone what good writing is.
I use AI every day. It edits my grammar, challenges my thinking and
helps me turn a messy idea into something other people can understand.
But I still need to bring the premise.
Because generating words has become cheap.
Having something worth saying hasn't.
Frequently Asked Questions
What is AI slop?
AI slop is low-effort content created or heavily generated with AI, with little human judgment behind it. It can look polished, with correct grammar and neat headings, but it says nothing specific and could have been written by anyone.
What is LinkedIn's "Seems like AI slop" feature?
Since 30 July 2026, LinkedIn has let people flag any post as "Seems like AI slop" from its three-dot menu. Flagging hides the post from that reader and notifies the original poster in their analytics, and LinkedIn has said the feedback trains its detection models. LinkedIn has also scaled back its own AI writing assist in favour of a more conservative proofreading tool.
Does using AI to write make my content AI slop?
No. Using AI doesn't automatically make something slop. Slop happens when content becomes generic and interchangeable, regardless of whether AI or a human wrote it. The problem is sameness, not the tool.
How do I avoid writing AI slop on LinkedIn?
Start with a real premise instead of asking AI for generic ideas, feed it material only you know (stories, client questions, observations), direct and push back on its first draft, add specific human details, cut anything repetitive, and read it back to check you'd actually say it out loud.
Want a second pair of eyes on whether your content actually sounds like you?
Join The Trusted Voice community → for real feedback, not a slop detector. 🍗
Comments ()