Ok. For this one I’m going to need you to strap in, as I might go a little over the top at times, but hear me out! Over the past few weeks and months, I’ve noticed a strange, worrying trend online. It’s a specific flavour of anxiety bubbling up in writer circles(and bloggers if you consider that a separate sub-species): people are being told to surgical-strike the personality out of their work.
The catalyst for this particular outburst? A collision of opposing philosophies that I experienced in quick succession.
First, I saw Sam Altman posting on X a few weeks ago, effectively celebrating the fact that we can now lobotomise ChatGPT to avoid em dashes— (like this, used here for dramatic pause effect by a human, me, to improve this paragraph) treating a glorious stylistic flourish as if it were a software bug to be patched out. It felt cold, clinical, and frankly, a bit sad.
Then, almost immediately after, I found myself sat in the audience at FOLSEA 2025, utterly captivated by Paul Hamilton’s keynote. (I will expand and share more lyrically about that later, but spoiler alert: it was the antidote I didn’t know I needed).
The contrast was jarring. On one side, tech leaders cheering for sterile uniformity; on the other, an educator deeply entrenched in the EdTech sphere championing the messy, beautiful human elements of tech.
And amidst all this, writers are being advised to delete their em dashes, tone down their adjectives, and flatten their voice into a beige, bureaucratic paste.
Isnt that what happened to Starbucks, McDonalds, and so many more or the colourful building of times past. No more funky shapes and brightly coloured arches, as it made the building unsellable. So now were left with mature, but boring looking boxes that feed us. Great!
And why are we being told all this in regards to writing?
Because “AI detectors might flag it.”
Let’s be absolutely clear about this: The em dash is not the enemy.
It is not a “tell” that you are a robot. It is a tell that you understand rhythm. Writers have used it for generations—brilliant writers, iconic writers, writers who shaped entire genres long before the microchip was even a remote fantasy in a scientist’s dream.
The idea that we should abandon our stylistic flourishes now because a clunky, paranoid algorithm thinks it looks “AI-ish” is, frankly, ridiculous.
The Detectors Are Marketing at its best!
We need to address the elephant in the server room: AI detectors are fundamentally broken.
They are guessing. And often, they are guessing badly.
I’ve seen entirely AI-written paragraphs sail through as “90% Human,” while my own original writing has been flagged as “80% AI-generated.” So what is happening…
The pièce de résistance of this absurdity?(Yes, I do speak a fair but of French. One of the few benefits of having been raised in Belgium) I recently tested a deliberately stitched-together paragraph inspired by the chaotic, yet expertly written, gothic brilliance of Mary Shelley’s Frankenstein. You would think an AI detector would recognise a monster of carefully crafted sentences written in the 1800s when it sees one. Apparently not.
The detector’s verdict: Frankenstein was, with 79% certainty, written by Artificial Intelligence. As you can see form the image below, very clearly written by Gemini, or perhaps it was Claude, I can’t quite tell…
Poor Mary Shelley. I guess creating the entire science-fiction genre wasn’t enough; she’s now being accused of being a chatbot by a piece of software that can’t tell the difference between Victorian angst and a ChatGPT prompt.
The “Cheating” Fallacy
So, here is the real question:
- If using a spellcheck doesn’t diminish your humanity…
- If having a friend edit your paragraph isn’t “cheating”…
- If Grammarly doesn’t make your ideas less yours…
Why are we suddenly judging people for using modern tools?
This isn’t about AI taking over. This is about fear distorting our understanding of what authentic writing looks like. We are in danger of writing worse—writing with less flair, less risk, and less soul—just to appease a broken machine. ( See what I did there. More dramatic pauses—and another!)
If we strip away our em dashes, our sentence fragments, and our weird metaphors to “pass” a detector, we aren’t proving we are human. We are voluntarily turning ourselves into robots.
The Silence of Suspicion
This culture of fear goes beyond just bad writing habits; it is corroding trust.
Jason Gulya, a Professor of English and Communications at Berkeley College who thinks deeply about these things, recently nailed this dynamic on LinkedIn. He pointed out that many educators are now silently building a “mental portfolio of misconduct” against students. They read a submission, suspect it sounds “too perfect,” but are terrified to say anything.
The result? A standoff where fear and distrust control the relationship.
Gulya’s solution is simple but profound: We need to normalise the question, “How did you build this?”
Not as an accusation. Not as a “gotcha” moment. But as a genuine inquiry into process. We need to move away from policing the output and start supporting the learning journey. If we are too scared to discuss how we write; whether we used a thesaurus, a friend, or a chatbot, we aren’t teaching. We are just playing a game of cat and mouse where nobody wins.
Where AI Fits (and Where it Absolutely Doesn’t)
So, am I suggesting that AI should write everything for us? No. Should we keep the human in the loop? Absolutely. Does AI have value for drafting, checking, brainstorming, or improving clarity? Yes, 100%.
It is about balance, not panic.
This brings me back to Paul Hamilton’s keynote at FOLSEA 2025. He shared something powerful that resonated deeply: During the process of publishing his picture book he was confronted with a choice. Human or AI. He shared about how he hired an illustrator—a real human, at real cost—because that illustrator brought history, context, and emotional resonance to the story and eventually to the piece, the illustration, the artwork. Could he have used AI-generated images? Sure. It would have been cheaper(Trust me…a lot cheaper!) and faster. But for meaningful work, the human touch mattered. And the audience could feel it.
At the same time, he’ll happily and proudly admit he still uses AI. Often for low-stakes tasks—mockups, quick ideas, exploratory sketches, coding projects, or project that would have otherwise remained in what he coined the ‘What if…’ stage. That isn’t cheating; that’s just efficient workflow. It’s the difference between using a microwave to reheat coffee you forgot about and hiring a barista to brew you a slow-roasted blend of hand-picked beans from Panama with Pacamara from El Salvadorto. Both have their place.
The Jekyll and Hyde Approach: Knowing Your Mode
Here is the secret that the “delete your em dashes” crowd doesn’t understand: Writing isn’t a one-size-fits-all jumper. It is a wardrobe.
As teachers, trainers, and in this case writers in EdTech, we need to be able to code-switch. I personally tend to operate in two distinct modes, and knowing when to switch between them is the difference between a confused reader and a delighted one.
Mode 1: The Storyteller (The “Full Thesaurus” Mode)
This is the mode you are reading right now. This is for thought leadership, opinion pieces, and manifestos about the future of learning. It’s where I let loose. Where i get to write down what otherwise would turn into a verbal rant with few interruptions. It’s what happens when I slow down my speech to a crawl and put it down onto paper, read and re-read, to eventually change what I wrote multiple times.
In this mode, we embrace the texture of language. We use metaphors that meander; we use pauses for dramatic effect; we use adjectives that we had to dust off from the back of the shelf because, for goodness sake, I bookmarked a thesaurus and I intend to use it!
This mode is about connection. It is about making the reader feel the rhythm of your argument. It is the place for the em dash, the rhetorical question, and the occasional literary flourish.
Mode 2: The Technician (The “Get It Done” Mode)
Then, there are the other posts. The instructional ones. The “How-Tos.”
When I am explaining how to set up a digital portfolio or sync a classroom set of iPads, nobody needs a metaphor about a wilting flower. They need to know which button to click, when.
In this mode:
- Clarity is King.
- Sentences become shorter.
- Bullet points replace paragraphs.
- We trade “voice” for “velocity.”
We don’t need Mary Shelley references here. We need precision.
With One Major Exception… There is one technical subject where the Frankenstein analogy remains perfectly valid: Student Information Systems (SIS).
If you have ever looked under the hood of a legacy SIS, you know that they are, quite literally, monsters. They are terrifying, stitched-together abominations of 1990s code, decomposing UI, and modern patches that shouldn’t work but somehow lurch forward, groaning, terrorising the villagers (or in this case, the teachers).
In that specific instance, feel free to use the monster metaphor. In fact, it might be the only way to accurately describe the horror of the user interface.
The Devil’s Advocate: Is “Plain” Always Better?
Before the social media keyboard heroes come knocking at my door with a pitchfork made of simplified syntax, looking for Em dashes, let’s acknowledge the nuance.
There is a valid argument for simplicity. We have all read writing that tries too hard—sentences so stuffed with adjectives and obscure vocabulary that you need a machete to hack your way through the jungle of meaning. That is not “voice”; that is waffle. And the average LinkedIn post likes to waffle a lot!
But clarity does not require sterility. Our writing does not need to treat patients, or grow penicillin-resistant bacteria.
There is a massive difference between “clear writing” and “writing that sounds like a sat-nav.”
- Clear writing respects the reader’s time but still engages their brain.
- Fear-driven writing (the kind stripped of all punctuation just to please an algorithm) respects nothing but the AI-detectors.
When we obsess over making our writing “undetectable” or “perfectly optimised,” we risk optimising the humanity right out of it.
Here’s the Heart of It
We don’t have to fear the tools. We don’t have to flatten our style into something unrecognisable. And we certainly don’t need to join the social-media outrage, pointing fingers and accusing people of being “fake” because their syntax is tidy, or dare we say, sounds too ambitious ‘for them’!.
We should continue writing the way we’ve always written—with voice, with quirks, and with our beloved em dashes. We should embrace AI as a partner in our workflow, not as a threat to our authenticity.
It isn’t AI or Humans. It is AI and Humans. Thoughtfully. Purposefully. With balance.
And if someone tries to take my em dashes away to satisfy an algorithm? They will have to pry them from my cold, thesaurus using, pause enjoying hands. And yes, in case you were wondering I do teach my year 2 students who are capable about interesting adjectives, the rule of three, and the difference between a comma, em-dash, and hyphen!
A Dash of Dashes – Our Quick Refresher
(Because if we’re going to use them, let’s use them with conviction)
The Hyphen (-) The tiny one. The worker bee. Used for connecting words to create new meanings: well-being, twenty-one, sugar-free. Think of it as the “matchmaker” of punctuation, forcing two words to get along.
You can find it on any Keyboard and type it quickly, with confidence.
The En dash (–) The medium one. The sensible middle child, but forgotten by most. Used for ranges: 10–12, Monday–Friday, ages 3–5. It basically says, “from here… to here.” It’s functional, but it’s not the life of the party.
On Mac, you hold down Option, and press the hyphen key.
The Em dash (—) The long, dramatic one. The diva. Used for emphasis, interruptions, and pure, unadulterated voice— like this. It represents a pause longer than a comma but less final than a full stop. It is punctuation’s equivalent of a confident hair toss or a dramatic gasp. Think of that presenter pausing for just that second too long. THAT’S the Em dash!
On Mac, hold down both SHIFT and Option, then press the hyphen key.
Now go forth and sprinkle your dashes boldly— mix them, match them, stretch them out— and don’t let a glitchy AI detector bully you out of your own writing style. Or if you want to really push it create patterns by alternating them!
— ––– ——— ––– — (Ok, maybe don’t do this…)