Sure, you can open ChatGPT and ask it to write a report for you. But what you’ll get is bland, generic, and uninformative—what’s become known as “AI slop.”
It’s like those cheese balls that come in the plastic barrel: artificial, mostly air, and devoid of nutrition.
You know it when you see it. It’s everywhere. But what is AI slop, exactly?
AI slop is writing that fills space without delivering much else. The words are there, but they’re disconnected from meaning or insight. This is the kind of text you get when you ask AI to write something without giving it the necessary context and don’t edit the output you get back.
It’s disrespectful to readers and wastes their time.
Let’s check out an example:
“It’s important to remember that AI tools can offer clear, concise, effective, and impactful support for a wide range of tasks, but on the one hand they promise simple and intuitive solutions, and on the other hand they may introduce challenges that require thoughtful planning. When used correctly, AI may help ensure smoother workflows, provide actionable insights at scale, and drive better outcomes—although this may suggest that organizations need to stay flexible, open, and prepared. Many teams analyzed the data, reviewed the results, and presented the findings in ways that felt clear and actionable, but it may be helpful to maintain a balanced approach that feels simple, straightforward, and easy to use. If you’re looking for a simple solution… this might be it.”
Ughhhhh. I’ll bet you started skimming after the first sentence and then skipped to the end.
On its face, it looks like a paragraph…it has grammatical sentences with recognizable words. But beyond that, it has little in common with actual human writing. It uses a bunch of words to say nothing.
What makes this AI slop so terrible? Let’s break down the problems in that paragraph, so you can learn to avoid them.
It has words and phrases that appear over and over in ChatGPT-generated text:
- Filler phrases like “It’s important to remember that…”
- Hedge phrases like “This may suggest that…” and “It may be helpful to…”
- Middle-of-the-road, safe language that doesn’t commit to a side, like “On the one hand… on the other hand…”
- Lists of three or four general adjectives: “clear, concise, effective, and impactful” Or “simple, straightforward, and easy to use”
- Vague action verbs like “helps ensure” or “drives”
- Corporate-speak like “at scale” and “actionable insights”
- Adjective twins, like “clear and actionable” and “simple and intuitive”
It also has some of the punctuation quirks ChatGPT often employs:
- It uses em dashes when the pause isn’t needed or when a comma would do just fine
- It adds performative ellipses: “If you’re looking for a simple solution… this might be it.”
- It uses three evenly spaced clauses separated with commas: “The team analyzed the data, reviewed the results, and presented the findings.”
I guarantee that avoiding these features will help your writing feel more original and human, even if you don’t make any other changes.
But let’s also look at characteristics that you WON’T see in AI slop—things you’ll want to include in your own writing to make it worth reading:
- A hook to draw readers in
- Details like specific examples, dates, and names
- Surprises or twists
- Personality
- Narrative arc
- Strong opinions
- Humor or wit
- Original insight
I don’t mean to suggest AI can’t do any of these things. It can, with your guidance. It can help you tighten up an idea, say something more concisely, or suggest an engaging intro. But it can’t bring your unique personality, perspective, or experience. It can’t bring you to your writing.
Your readers aren’t just looking for words, they’re looking for you.