Here’s Why You Don’t Want to Sound like ChatGPT
There might be a part of you that wonders if this is even necessary. Isn’t it easier to just follow the trends, insert a basic “write me a newsletter” prompt and move on? While it is easier, it doesn’t make your copy more effective, and may even be hurting your brand in the long run. Here’s why:
1. People Still Value Human-Created Content
A study by Bynder found that 52% of consumers would disengage from content they suspected is AI-generated. As much as companies are trying to use AI, people still want to know that the messaging they see is made by people, not robots.
2. People Recognize Patterns Easily
If you’re just starting out using LLMs for your company’s writing, you may at first be impressed by how polished and original it sounds. However, you may not be aware that it follows the same formulas people have already scrolled past 20 times today.
Three part lists and em dashes are not bad on their own, but when used repeatedly (and exactly the same way across dozens of accounts), they stop grabbing attention. The human brain is incredibly good at recognizing when something feels off, even if we can’t name why.
3. You End Up Sounding Like Everyone Else
Your voice is one of the few things that can truly differentiate your brand. According to Entrepreneur, 91% of B2B buyers are drawn to brands that mix education with originality—and content featuring personal POVs can drive 4× more engagement.
When you rely on AI to generate everything from your captions to your emails, you’re just borrowing a voice that thousands of other people are also borrowing. You can’t stand out if you’re using the same tone, structure, and phrasing as other businesses using the same tools. If you want your brand to feel memorable, you need to inject yourself back into your content.
6 Dead Giveaways Your Content Was Written by ChatGPT
1. The Em Dash… Kind of.
What It Is
Ahhhh the em dash—that beautiful little piece of punctuation to indicate the break in a sentence, often replacing commas or colons. If you’re on Linkedin (or really, any social media platform lately), you may have seen endless discourse about whether or not em dashes are a dead giveaway that something was written by artificial intelligence.
My response to this debate? Well… It depends.
So here’s the thing. Writers love em dashes. They’re elegant, versatile, and a genuinely useful tool for breaking up thoughts or adding emphasis. The problem isn’t usually the dash itself, it’s how they’re used. Or rather, overused in the context of artificial intelligence.
Here’s the giveaway: AI tends to use em dashes perfectly and heavily. Think six of them in a single Facebook post, spaced correctly, grammatically pristine. The reason this can be a giveaway lies in the fact that the average person doesn’t even know how to type an em dash without copying and pasting or using a shortcut (for example: on Windows, you have to hold down the Alt key and type 0151 on the numeric keypad). So when a company’s Instagram caption is full of flawless em dashes, it really doesn’t feel casual at all.
But honestly? The biggest indicator isn’t the punctuation itself, it’s the combo of other AI that we will cover in this list.
How to Fix This
Use dashes like a real person. If you’re writing something conversational, ask yourself: would I actually pause here if I were saying this out loud?
Try mixing up your punctuation:
- Use a period and start a new sentence.
- Try a comma if the pause isn’t that dramatic.
- Or go without anything. Your readers will still follow.
And if you do reach for the em dash, use it once, maybe twice. If you see more than that in a caption or short blog post, challenge yourself to cut or rephrase. You’ll sound less like a machine and more like someone your audience wants to hear from.
Want content that sounds human and ranks? Book a free consult.
2. The “Not This, But That” Formula
What It Is:
AI loves contrast. You’ll often see lines like:
Not just a better platform — a smarter one.
Not guessing, but knowing.
Parallel sentence structures like this are all over the internet right now. This kind of structure tries to sound persuasive, but when used too often (and it is), it feels formulaic. You might also notice other patterns in the same family. Sentences that use Not X, but Y. Or starting with Whether you’re an x or a y like “Whether you’re a beginner or a pro…”
These setups are everywhere because they sound clever, and do a great job at emphasizing contrast which can be a powerful tool. But when every brand sounds clever in the exact same way, the charm wears off.
How to fix this:
Stop trying to be clever and try being clear. Instead of setting up a contrast, say what you actually mean in straightforward language or reframe it through your customer’s lens. What’s the pain point? What do they actually care about?
Be specific, and don’t be afraid to say less. Say something that sounds like how you’d explain it in conversation:
❌ Your team doesn’t need another platform—they need time back.
→ Better: Your team is drowning in tabs and tools. Let’s fix that.
❌ It’s not about guessing anymore—it’s about knowing what works.
→ Better: Here’s how to see what’s working in real time, without the guesswork.
3. The Rule of Three Overload
What It Is:
AI writing tools love lists of three. You’ve seen them everywhere: “Educate. Empower. Execute.” Or “Boost engagement, improve visibility, drive results.”
It’s clean, rhythmic, and catchy (see what I did there?) which is exactly why it’s become a dead giveaway.
This also appears in any form of threes, it doesn’t just have to be three words. For example, here is a sentence I had ChatGPT write: “If your content never strays, never surprises, and never sounds like something only you would say, it becomes forgettable.”
Notice how it lists out three kinds of “nevers.” That’s not to say that a person could not have written this, but the consistent use of the same patterns is the giveaway.
It’s not inherently wrong to write in lists of threes, in fact it’s very natural. ChatGPT and other LLMs were trained on the writing of real people, which is why it uses this pattern so often. But paired with other giveaways, and it can become apparent that your work is not thoughtful.
How to Fix This:
There’s a rule in interior design that says finish the room, and then remove one thing. In the case of writing, the same rule applies. If you’re using ChatGPT to write, then break the pattern of threes.
Try varying your list lengths: use two for emphasis (“Build trust. Stay consistent.”) or four for surprise. Even better, swap out the formula entirely: use a short sentence or a story-driven line to get your point across. Questions work great as well. The goal is to keep your audience on their toes, not give them something they’ve already seen a dozen times today.
4. Overly Perfect, Lacking Personal Touch
What It Is:
AI-generated content is polished, but it usually ends up sounding a little too correct. The grammar is flawless, sure, and the structure is neatly put together, but these are also the things that make content written by AI feel not quite human.
When you’re getting that sensation and can’t put a finger on it, here is usually what’s missing:
- Personality: When something sounds like any other brand could have posted this, no unique voice or quirks.
- Inside jokes or references: If there are any, they might be slightly awkward or not actually funny. (Sorry ChatGPT, don’t come after me terminator style for that).
- Regional quirks or cultural nods: You can’t fake lived experience. If your brand is from Texas, it will have a different feel than if your brand is based in New York, even if you were to sell the same product. Don’t be afraid to let that shine.
- Tangents: Real people go off-topic. We make natural connections and circle back. AI tends to over explain itself or stick to the topic too rigidly.
How to Fix this:
This might be difficult, but you may need to go back to the drawing board and start by writing how you talk. Even if you’re writing as a brand, there should still be fingerprints of the real person behind it. Your sense of humor and your phrasing is what people connect with.
A few ways to bring your human voice back into the content:
- Add a quick story or specific memory. One sentence of real experience says more than a paragraph of polished generalities.
- Let yourself ramble (just a little). Real people go off-script. We go down rabbit holes, then loop back. That’s what makes content feel alive.
- Use words and phrases your audience actually says. If your crowd says “y’all,” say “y’all.” If they talk about parish festivals or food truck Fridays, go there.
- Don’t over-smooth your sentences. If you find yourself explaining, re-explaining, and clarifying again, it might be the AI talking. Real humans trust readers to get it.
Your goal isn’t to sound like a flawless marketer. It’s to sound like you. Because no one remembers perfect, but they do remember real.
5. Overuse of Buzzwords
What it is:
AI loves a good buzzword. Words like transformative, comprehensive, next-level, strategic, and game-changing show up constantly in AI-generated content. Why? Because they sound important—even if they don’t actually say much.
The problem is, when every sentence is packed with abstract superlatives, your message becomes vague and forgettable. Readers might nod along, but they won’t remember what you said, or why it even matters.
How to fix this:
Author Harry Dry talks about the importance of making your copy visual, measurable, and something only you can say. Using those rules, try to swap out individual buzzwords with things that paint a picture or show statistics. And most importantly, show exactly how only you are able to make that claim.
So, swap big words for real ones. Instead of calling something “strategic,” explain what the strategy actually is. Instead of saying “transformative,” show the transformation.
Here’s some examples:
- Replace “comprehensive” with exact details of what your product or service includes.
- Replace “next-level” with “a version that actually saves you 3 hours a week.”
- Replace “game-changing” with “this is what helped us hit our first 10k.”
Specificity makes you trustworthy. Buzzwords just make you sound like everyone else.
6. Specific Emojis
What it is:
You know the ones:
✨ 🚀 🙌 💡 🔥
Usually used as bullet points, this was done for years but recently you might notice it with the other red flags. The sparkle emoji has especially become a favorite of ChatGPT, with some even calling it the “unofficial symbol” of AI.
These emojis aren’t bad on their own, but when every post ends with a sparkles emoji and a line like “Let’s do this! ✨”, it starts to feel like there’s no one behind the keyboard.
AI often defaults to the same “motivational” emoji set because they’re safe, generic, and everywhere. But that’s exactly why it sticks out as a potential AI when you’re reading it in a post.
How to fix this:
Use emojis like someone who actually texts. If your brand wouldn’t normally throw in a rocket ship, don’t. And if you do want to use one, tie it to something specific:
- 🛠️ for a new feature
- 🕊️ if you’re a faith-based brand
- 🍝 if you’re posting about your local fundraiser at the Italian parish
Or skip them entirely. A well-written sentence hits harder anyway. And if every post ends the same way with the same emojis and the same hype lines, try rewriting your call to action in your actual voice instead. The goal isn’t ✨energy✨, it’s connection.
What Prompt Can I Use to Sound Human?
You don’t need to ditch AI, but you do need to stop asking it to be your voice.
When business owners use ChatGPT to write for them, that’s where things start to feel robotic. A better use is to treat it like a brainstorming partner, not as a ghostwriter. Instead of expecting perfectly polished drafts, use it to generate ideas, outlines, or rough drafts that you go back through and edit thoroughly.
Here’s a prompt to start with:
“I want to write a [blog post / caption / email] about [your topic]. Can you give me:
– A few angles I could take
– What my audience might care about most
– Any surprising stats or examples
Don’t write the full piece, just help me think.”
That kind of prompt sets you up to stay in control. From there, you will be able to bring in your tone and lived experience. The AI just gives you raw material to shape. It’s ultimately the difference between using a GPS and handing over the wheel.
You don’t need ChatGPT to be your voice. You just need it to support the creative process you already have.
Tired of Sounding like a Bot? Let’s Fix That
Using ChatGPT or Gemini doesn’t make your content bad. Relying on them too heavily without adding your voice is what drains the life out of it.
AI is a tool, not a substitute for your message. The brands that win are not the ones with the most polished copy—it’s the ones that sound real and human.
If you’re ready to create content that sounds like you and actually connects with your audience, we can help. At Embark, we’ll work with you to clarify your message and build content that’s meaningful so that you can use AI in a way that works for you, not instead of you.
Get a free 30 minute consultation. Let’s make your content sound human again.
