We’re back with Part 2 of “On Writing Really Very Goodly and Stuff, Vol. 1: Making AI Sound Not Like AI” for those who were waiting for the conclusion (and wondering about em dashes).
Don’t worry, we are going to recap to refresh your memory before we dive into Part 2!
In Part 1, we talked about the rise of “AI slop,” content that technically looks fine but feels mass-produced and oddly soulless, like the writing equivalent of factory-processed cheese slices. The key takeaway: AI works best as a tool (think GPS), helping with things like brainstorming topics, gathering research, and checking logic, rather than as the one actually driving the car. And, as we began to see at the end of Part 1, there is also a subtle tell that can make even human-written text look suspiciously AI-ish… the em dash.
And here we go with Part 2.
(In case you missed it, you can check out the full version of Part 1 here.)
I’m going to be completely honest, I’m sure I’ve encountered em dashes at some point in the wild; I don’t think I ever paid them very much attention, but they seem to be relatively popular with professional writers and authors given their relative flexibility as punctuation.
And I’m sure you, too, dear reader, have encountered them at some point or another, but very likely how sparingly they were used kept you from noticing them. Who, realistically, could have been bothered to notice the lowly em dash when the period and comma were taking center stage, clearly evidenced to be far more popular through their frequency of use?
In doing the research for this blog, I found that em dashes are called “em” dashes because they are roughly the width of an “M.”
Here is a size comparison for reference:
Em dash: —
Em letter: M
Worth noting is that there is also something called an “en dash,” called that because it is roughly the width of an “N.” In spite of what I have spent my entire life thinking, these are not in fact hyphens, which is something I found slightly earth-shattering.
But alas, we are not here to talk about en dashes and hyphens, nor are we here to pick up the shattered pieces of my once en dash-free world; we are here to talk about em dashes and the problem they present, even for — especially for — text that isn’t AI generated.
In Part 1, I mentioned using AI to list a bunch of ideas to potentially use for this blog, and I included the very first sentence that the AI gave me in reply, which itself included an em dash. The rest of the list included a smattering of em dashes, far more than any normal person would ever think to use, even the professional writers who seem so fond of them.
And this is the problem. Em dashes are anywhere and everywhere now: forum posts, press releases, e-mails, twitter posts, you name it; they’ve gone from being something that one might happen upon in some rare, chance encounter — perhaps reading a book or a newspaper article or something — to being ubiquitous.
And this ubiquity is in spite of their not being very easy to natively use; Word, for example, requires that you either know how to manipulate auto-correct into generating one or that you know a relatively simple yet still obscure keyboard shortcut — it’s insane.
This isn’t some punctuation that the average person is ever going to bother using — unless by some accident or typo — and yet it now seems inescapable.
And the reason for that — the reason for all of this punctuation madness — is that AI LOVES the em dash.
The em dash had AI at “Hello” and these two are now inseparable.
The frequency with which AI uses them is bad enough at this point that even if you are one of those literary types who enjoys using all sorts of fancy punctuation in your writing, you’re better off ignoring the em dash for the time being — if for no other reason than because it is going to signal to a growing population of your potential readers that there is a reasonably high probability that the prose that you’ve lovingly and painstakingly crafted — that you’ve poured your blood, sweat and tears into — may in fact in actuality be the work of AI.
To bring this full circle: if you’re using AI to generate portions of text for editing and expounding in your own writing, part of that editing should be ensuring that you remove any and all em dashes, replacing them with their respective punctuation — failing to remove the em dashes is tantamount to placing a bright, red blinking sign on whatever you’ve created that it was actually generated with AI.
While AI can be incredibly effective at streamlining your writing workflow, there are a number of pitfalls that can be easy to fall into if you’re not careful.
Best practices at this point involve using it as a tool rather than as a one-touch article generator; using it for brainstorming, summarization and research, and putting your own personal touches on anything that you do generate, and — of course — making sure that you remove any and all em dashes you might see — just to be safe.
Just say no — to em dashes.
Or you could just get into full-on arguments with AI, COMING IN VOLUME TWO.
With XMAT or our other platforms, you can utilize machine translation and generative AI in a secure environment without worrying about the transparency of the cost. At Kawamura International, we do more than just provide machine translation or generative AI services. We offer a range of support to help make your translation operations run more smoothly. If you are facing translation challenges at the moment or have any questions about our language services, please feel free to contact us.