The Dash Debate?
Is your content secretly shouting "written by AI"? From the overuse of the long dash (—) to phrases like “Let’s delve into…” and “It’s not X, it’s Y,” certain writing quirks are dead giveaways.
Why the Long Dash Might Be a Dead Giveaway You’re Reading AI-Generated Text
You’re reading an article—maybe a LinkedIn post, a product description, or even an “authentic” founder story—and suddenly, there it is:
The long dash. — Like this.
You pause. You squint. You wonder: Did a human really write this?
Congratulations, you’ve stumbled into The Dash Debate —an increasingly online conspiracy theory (with some truth to it) that the em dash has become the digital fingerprint of AI writing.
So, What’s the Deal With the Long Dash?
The long dash (also known as the em dash) is a grammatical tool often used to indicate a pause. It’s stylish. It’s dramatic. It’s effective. It’s also everywhere in AI-generated content.
Tools like ChatGPT, Claude, and Gemini love the long dash. Ask them to write a product announcement, a thought leadership post, or a marketing blurb, and odds are, an em dash will appear faster than you can say “prompt injection.”
Why AI Loves the em Dash
There are a few reasons:
- It’s safe. The em dash is a reliable way to separate ideas without confusing grammar rules. No need to worry about comma splices or semicolon misuse.
- It mimics human style. In many cases, the dash gives writing a casual, conversational tone—exactly what generative AI is trained to replicate.
- It adds emphasis. Want to make a sentence pop? The dash adds some emphasis.
But here’s the thing: AI uses it too much. In trying to sound natural, it overplays its hand. And now, savvy readers are starting to notice.
Ok, so you’re thinking: “I’ll just use an AI detector to figure it out.” The problem? They don’t work.
Can You Spot the AI?
Some readers have turned the em dash into their version of a lie detector test.
- Too many dashes in a row?
- Used instead of other punctuation that would make more sense?
- Every paragraph has one?
It’s starting to become a tell—like how certain phrases ("Let’s unpack this" or "In today’s fast-paced world") scream “Ai wrote me.”
Other Common Phrases That Scream "Written by AI"
- “Let’s delve into…”
AI loves to delve. It doesn’t explore, peek, or dive in—it delves. Every. Single. Time. - “It’s not [X], it’s [Y].”
This structure tries to be clever, but ends up robotic:
“It’s not just automation—it’s transformation.”
“It’s not about replacing humans—it’s about enhancing them.”
- “In today’s fast-paced world…”
A classic AI opener. It’s trying to set the stage, but it sounds like a line from an HR onboarding video. - “Harnessing the power of…”
Another favorite. Nobody says this in real life, especially not in Slack threads. - “Unleashing potential”
Whether it’s talent, data, or productivity—AI wants to unleash something. - “Paving the way for the future of…”
It’s like a futuristic Buzz Lightyear wrote your LinkedIn post. - “Revolutionizing the way we…”
Sprinkle in some vague tech jargon, and you’ve got peak prompt prose. - “At the intersection of [X] and [Y]”
Example: “We’re building at the intersection of AI and human creativity.” - “Unlock new possibilities.”
Feels inspiring until you realize it's been used 4,000 times this week alone. - “Let that sink in.”
Spoiler: we didn’t need time to “let it sink in.” The statement wasn’t that deep.
Want to sound human? Ditch the polished platitudes. Keep the quirks. Use weird analogies. Be punchy. Be real.
And for the love of all things human. Skip the “delve.” and reduce the amount of em dashes
