We’ve all heard em dashes are a sign of AI writing. Now, we’re hearing the same about key rhetorical techniques.
One such technique is antithesis, which is when you say, “it’s not X, it’s Y”. You see examples of this in almost every AI-written social media post (just three seconds on Facebook allowed me to find this poor specimen: “he wasn’t making a film about war; he was making a film against it”).
Then there’s the tricolon, which is when you group items in threes. For example: “he was thoughtful, principled and committed to positive change.”
Parallelism is when you repeat a word, phrase or sentence, such as “what you see is what you get” and “easy come, easy go”. Those are good, classic pre-AI examples. Parallelism can also take the form of repeating a type of word – and this is what AI is especially fond of. A bad AI example is: “Each one was a map of her emotional landscape – the heartbreak, the love, the surgeries, the resilience.”
That last example is made even worse by the metaphor (“a map of her emotional landscape”).
Metaphors are tough for even experienced communicators to use, as they require a high degree of subtlety and judgement. Ideally, a metaphor should change the way people view something and ought to be original: but that requires genuine insight.
This is why they should be used sparingly – yet AI slop throws as many cliched metaphors as it can into writing, making everything sound like dialogue from a bad soap opera. To give a melodramatic example from an AI post about a Hollywood production: “the story had already been written in whispers and side glances on the set.”
What does that even mean?
Ditto with this doozy: “The crew set up lantern light, the kind that feels like memory.”
Then there’s using multiple short sentences to build power and pace, which is usually a mark of good writing except when AI overuses it – as in: “He was a thinker. A leader. A pioneer among adventurers.” (This, btw, is also an example of the tricolon.)
So if people are now associating these techniques with AI writing, should good writers stop using them?
No … and yes. But mainly no.
You see, the problem isn’t with the techniques: it’s with how AI uses them. (And yes, after writing that sentence, I realised I used antithesis and then wondered whether I was sounding like AI myself.)
Abraham Lincoln’s Gettysburg Address used these techniques and it’s still as powerful as ever: yet an AI-generated post that uses the same techniques is obnoxious.
The difference lies in how and why the techniques are used. When AI uses antithesis, for example, it’ll do it again and again and again, often inappropriately or in a needlessly melodramatic manner. It’s like learning how to hit a beautifully high note and then just singing it again, and again, and again, until people want to scream.
Using rhetoric requires sense and human judgement: it’s not something that can be automated.
Rhetorical techniques have been around since Ancient Greece: it doesn’t make sense for us to suddenly abandon them now. They’re used because they’re effective in moderation. And moderation is the key word, because whereas in the past we might simply have looked heavy handed if we overused these techniques, now we’ll look like we used AI.
Great writing, as with great speaking, primarily involves using plain English. If you use rhetorical techniques, use them sparingly and with good reason.
