Your AI "voice" sounds nothing like you. Here’s why.

If you use AI to write anything, trust me - it’s more obvious than you think.

Take it from a journalist, copywriter and guy who spends 40 hours a week doomscrolling LinkedIn. After thousands of hours of skimming success stories, mini lectures and overly-sentimental confessions, I’ve developed a strong nose for what’s written by you and what’s written by… not-you.

I’ve made a mental list of the many telltale signs of AI writing, and if your post has at least 3 of them, it’s almost always a dead giveaway. You might think you’re slick, but really you look mad suspicious if you use:

→ Waaay too many emojis
→ Tons of em dashes (—), bold text, italics text and #hashtags
→ Phrases like “And honestly?”, or “Here’s the thing”, or “It’s not X. It’s Y.”, or “Not X. Not Y. Just Z.”
→ Use arrow symbols for bullet point lists (like I just have).

‍

Okay, but… who actually cares?

Fair question. After all, what matters most is that your posts get clear intended message across. If you use AI to achieve that, and it does it well, what’s the issue?

I’m not saying you absolutely shouldn’t use AI in writing. It certainly has its uses. I personally never use it in my own work. Everything you see with my name on it – LinkedIn posts, concert reviews, news stories and articles like this one – is written by hand on an old-timey Bluetooth keyboard.

That said, I still incorporate AI into my professional work; I’ll often use it to make “sh*tty first drafts”, as my favourite English teacher Dr. Campbell would call them. Before anything gets published or sent away for review, I always add some human touches by reading over everything, replacing words, switching sentences around and freshening up the grammar to perfectly meet every brief.

I’ve found AI useful when I need to quickly consolidate information and form rough drafts, especially during time crunches. Everyone’s got their justifications for using it too. But if your reliance on it gets too heavy, it’s easy to neglect a key trait of the writing process that maintains your humanness and uniqueness: your tone of voice.

‍

The illusion of “sounding like me”.

There’s a trap I see friends, colleagues and connections fall into all the time: You hop on ChatGPT or Claude, ask for a 170-word LinkedIn-friendly spiel, quickly scan over what gets spat out, then decide it sounds like you.

But does it really sound like you? Or does it just sound “good”?

Regardless of how well-trained generative AI is, it’s algorithmically designed to respond to every prompt by taking the safest possible route towards the outcome it thinks you’ll prefer the most. That means when you ask Claude to write you a LinkedIn post, it’ll use whatever writing style it’s already been trained on. Then it fills the gaps with “good” words, phrases, structure and grammar.

‍

It’s this “safe” filler - like the examples I listed above - that leads to AI posts sticking out like sore thumbs.

But there’s an irony to it: while many people now view these writing features as “AI slop”, algorithms love them for how popular they actually are. From an algorithm’s perspective, if the em dash is commonplace, it means it’s a safe choice. It’s easily understood, and thus qualifies as good writing.

Any writer will tell you that before they developed a tone of voice, they emulated anything they read solely because they liked how it sounded. There’s brief confidence in using smart-sounding language that, in all reality, doesn’t truly reflect who you are. If you’re still developing a tone of voice, it’s easy to look at anything AI writes and think “this sounds like me!”, when really it sounds good, but nothing like you.

‍

So what should you do?

AI writing is a gift and a curse. It’s not my place to tell you whether or not you should use it, but from my perspective there’s 3 paths you can take:

1.      Ridin’ solo. “Old school”, if you will. Writing takes longer and is harder work without AI. But you’re rewarded with an improved vocabulary and more confidence in your tone of voice (and you can certainly use AI tools to train those writing muscles!)

2.      All in on AI. The flipside to AI’s mediocre wordsmithing is that it’s safe and reliable language. By making your posts more accessible, you could see potentially higher engagement. If you’re not too fussed about tone of voice and just want to get a clear message across, you may decide this is the way to go. But be warned – public sentiment on AI isn’t great, so unless you’re always training your bot, it could reflect poorly on you in the not-too-distant future.

3.      A healthy mix. This is where I reside. I use AI where it makes sense in the writing process, but I stick to my own words when it’s more personal. Sometimes I just need to get a simple message across. But when I’m writing from a place of passion and vulnerability, I want readers to really feel what I’m saying.

‍

There’s no wrong answer here. Not everyone writes for a living, so choosing the path that makes the most sense to you is what matters.

But before you decide, please do me a favour: write a post or two without AI, just off the top of your head. Doesn’t matter if the grammar’s bad or reads like a 12-year-old wrote it. The goal is to get a sense of how you truly sound in writing and how it makes you feel.

Give it a whirl, have fun and let yourself suck at it – even if you’re not some version of “good”, you’ll never sound more yourself!

-Devin Pike, Copywriter

‍