There has been an invasion going on for the last year or more.
Not one on the southern border that required wankers in sunglasses to arrest random, bown-skinned people, but an invasion of crap content and em dashes.
The em dash is slightly longer than a normal dash and beloved by writers for creating dramatic pauses or setting off explanations.
Before AI it was something of a rarity requiring (at least on a Mac) a three-key combination to lure the little fella out of hiding.
However, it has experienced a massive surge in popularity at the exact same time as many previously mediocre writers began posting content that would have earned them the official badge of marketing guru before the arrival of ChatGPT.
As somebody who considers writing and marketing to be two of a tiny number of things I’m adept at, it was frustrating to see these people pull not just the wool, but the entire sheep over people’s eyes.
Try as I might to educate the world that these people were, for the most part, copying and pasting content straight from ChatGPT, and the em dashes proved it, the trend continued to gather pace.
I felt a bit like the first person warning that there were a lot more grey squirrels knocking about than usual, and we needed to be careful if we were to save the red variety from impending doom.
Oh, shut up, you miserable old bastard. They’re cute,” everybody would protest, blissfully unaware that the tree-hugging rat bastards were already murdering their genuinely cute and smaller red cousins in their beds.
But as more people noticed that a grammatical oddity they had never heard of six months ago was now more common than a change in tariffs, and that content was starting to feel all the same, the pushback began.
The em dash pirates were being called out in public, and their expertise was being questioned.
Phew.
Sadly, innocent copywriters who had lovingly, correctly and sparingly used the em dash for decades were now being accused of using ChatGPT for their writing.
But with change comes adaptation.
And so it came to pass that the cunning bastards just took their em dashes out.
Fuck.
Now what do we do?
How can we identify the individuals who are knowledgeable in their chosen field and thus avoid hiring those who rely solely on copying and pasting?
Worry not, I’ve got your back because ChatGPT and even the much better Claude, for that matter, are still a long way from nailing it.
They still give off many tell-tale signs that the author isn’t sharing entirely original work.
Before I tell you what they are, though, let me say this.
You should be using AI
I’m in no way averse to coaches using AI to help them write, especially if they’re far better coaches than writers.
In fact, I’m in favour of it.
I barely use AI for my writing, but that’s mainly because I’ve been writing articles and posts for almost 20 years, and I don’t need to.
I would have used it back in 2006 and 2007 when I was cranking out a lot of content every single day.
I now only write a couple of articles per week and prefer to handcraft them, just as I have with this one, which is 95% written by me.
However, I’m most definitely against people using AI to position themselves as experts on topics they are not intimately familiar with.
That isn’t very honest, and I hope you would never do such a thing.
Which is why I have a cantankerous habit of calling such rapscallions out, especially when it’s apparent people reading the content think they are bona fide experts.
Although you may be more Zen about life than I am and not feel the need to be outspoken (gobby to us Brits), you may still want to know how to spot AI-generated writing so you’re better informed.
Plus, you’ll know what to edit out to make sure that if you do use AI, it isn’t screaming fake to everybody.
As such, here are the biggest giveaways I see on a day-to-day basis that you may well be reading AI-generated content..
One or two of these doesn’t necessarily mean you are definitely reading bot-speak, but if you see more than two or three, it’s unlikely the author has a brain, a heart, or courage.
1. Lack of humour
I want to kick this off because it’s something close to my heart.
AI is fucking horrible at humour.
Or rather, any humour that isn’t highly formulaic.
ChatGPT does occasionally make me laugh, but only with inside jokes that it’s effectively feeding back to me.
Things like ‘Fuck me Tim you’re right, I’m such a douche for even suggesting that‘
But because it’s trained to predict the most obvious next word, it will never give you anything original.
AI could not have written Monty Python’s ‘The Dead Parrot Sketch,’ George Carlin’s ‘7 words you can’t say on TV,’ or anything by Richard Pryor, Mitch Hedberg or Robin Williams.
If something is making you laugh, you either have a dad joke sense of humour, or you’re not reading AI.
2. A lack of personal stories
AI won’t tell you about the time it sent an email to a client asking if she had considered intermittent fisting when it meant fasting, like I did.
Nor will it be able to share your stories about your life.
Or your clients’ remarkable results and success stories.
If you’re reading personal stories, you’re probably not reading AI.
Or at least, not entirely AI
What you can do: Get more personal and get more vulnerable.
And personal stories don’t just differentiate you from AI, they also build trust.
3. The Rule of Three
As many people are now editing out em dashes, this may have replaced them as the easiest way to identify AI-generated content.
They’re silly. They’re lazy. They’re unnecessary.
Often, it’s on one line, as per the above.
But the (slightly) more cunning users may spread it over three lines to avoid suspicion.
However, I’d never use AI in this way.
Never did.
Never have.
Never will.
The reality is, we don’t speak like this; it’s a copywriting trick used for emphasis and often used in twee homepage subheadings like:
Laugh. Love. Live.
Simple. Reliable. Effective.
An alternative to these, but still using the rule of three, is Not this, not that, but the other.
Sign up to work with Tim Brownson and you’ll get no fluff, no bullshit, just solid advice.
Now that you know about the rule of three, you will see it used a lot!
What you can do: Break the pattern. Use four things. Or two. Or just use it occasionally.
4. Rhetorical questions
These can be really powerful if used sparingly.
But sadly, AI has no comprehension of sparingly and will drop them in every fucking article or post it gets its hands on.
Think I’m joking? Well, I’m not.
You probably never even noticed that I just did it.
Why? I have no idea.
Look, I did it again.
It would have been easier for me to have typed I’m not joking, and made the exact same point.
Similarly, I could have said I had no idea why I did it again.
What you can do: Use them sparingly. One per article, maximum.
5. Correlative conjunctions
I said I haven’t used AI to write this post, but we both know I’d no idea what the fuck a correlative conjunction was 5 minutes ago and AI bailed me out.
I, and I suspect you, know these as ‘not only, but also‘ statements.
We use these a lot, but AI presumably bathes in them.
Not only does it bathe in them, but it also showers in them.
Not only does hiring Tim help you get clients, but it also allows you to contribute to his retirement fund.
Besides being annoying, they are also red flags.
What you can do: Nobody will notice or be bothered by the occasional use of these, but they are classic sales page copy and thus may unconsciously feel manipulative, so they should be avoided.
6. Emojis as bullet points
AI loves bullet points, but not just bullet points as God created them, but fucking emoji bullet points.
They are rampant, especially across social media.
Green ticks not so subtly telling you this line of things is good.
Red crosses warning you that the things on this line are to be avoided.
🧨 And bombs about to go off to signal, who the fuck knows what?
Hire Tim, and you will get…
✅ His dashing good looks ✅ Lots of swearing ✅ Some great marketing advice
Hire Tim, and you won’t get…
❌ Help with your diplomacy ❌ A man who can control his dogs ❌ A boring session
Bullet points can be brilliant at breaking copy up, but other than ad people running shit Facebook PPC campaigns, who even knew emoji bullet points existed two or three years ago?
What you can do: Use proper bullet points like a grown-up. If you must use emojis, use them in the actual content, not as formatting.
7. Common AI Phrases
Here’s a hot take: AI uses hot takes more often than I eat hot takeaways. That’s a lot, by the way.
It also tends to use a lot of the following:
“Here’s the kicker”
“Here’s the deal”
“Here’s the truth”
“Let’s talk about”
“Uncomfortable truth”
“Listen up”
“Secret sauce”
“The key takeaway here is…”
“Ultimately, it comes down to…”
“One fundamental principle is…”
“Let’s break it down”
None of these scream AI in isolation, but here’s the deal: if you’re seeing the same generic phrases in content over and over, then the uncomfortable truth is you’re probably reading AI and not a hot take.
What you can do: Write like you actually talk. Use your own expressions, not these generic AI favourites that everyone’s sick of seeing.
Better yet, have a Word doc that includes all the words that you want excluded and use it in every prompt you use when you write something that you care about.
Over to Claude for the (boring) conclusion
Identifying AI-generated content has become increasingly important in today’s digital landscape. By understanding these seven key indicators, you can make more informed decisions about the content you consume and the professionals you choose to work with.
The rise of artificial intelligence in content creation presents both opportunities and challenges. While these tools can be valuable resources when used appropriately, it’s essential to maintain authenticity and transparency in all communications.
As we move forward in this evolving digital age, the ability to distinguish between human and AI-generated content will become an invaluable skill. By staying vigilant and applying these detection methods, you can ensure that you’re engaging with genuine expertise rather than automated responses.
Remember, the goal isn’t to eliminate AI tools entirely, but rather to use them responsibly while maintaining the human connection that makes coaching truly effective. Take action today by implementing these strategies in your content evaluation process.
