If you’ve never seen Jim Carrey’s 2007 psychological thriller The Number 23, then congratulations. It is a film about a man who sees the number 23 so many times that he ends up going bonkers. I used to think this film was stupid. However, now I appear to be living it.

My own personal number 23 is a rhetorical device: “It’s not X, it’s Y.” Everywhere I look, there it is. Whenever I hate myself enough to scroll through Facebook’s wilderness of algorithmically suggested posts, I find myself being smacked in the face with sentences such as: “Self-improvement isn’t a trend, it’s a lifestyle shift,” and “The small wins aren’t just moments, they’re the majority of your life.” Once you notice it, it becomes impossible to ignore. This weekend during a Peloton class (I know, shut up), I heard an instructor bark a variation of “this isn’t X, it’s Y”. Yesterday, a character did the same during a TV show I was reviewing, and I dropped a star from its score in retaliation.

You know where this is coming from, don’t you? “It’s not X, it’s Y” is an AI mainstay. It’s one of ChatGPT’s most insidious tells. No matter how innocuous a prompt you enter, AI will always find a way to sneak it into its response. Ask it if you should put more ham in your pasta, and it will tell you: “Ham doesn’t just taste good – it makes everything else taste better.” Ask it if you should chase a bee around your garden and it will say: “Bees aren’t stupid – they’re hyper-specialised”.

“It’s not X, it’s Y” has become such a shorthand for lazy AI slop that, as soon as I see or hear someone telling me that something isn’t something because it’s actually something else, I automatically tense up on the assumption that I’m not dealing with a human, I’m dealing with a datacentre. That might not necessarily be the case – there is a possibility every example is completely organic – but it’s a sign of the times that we can’t just relax and assume the things we see and hear were made by people.

Although “it’s not X, it’s Y” predates ChatGPT, I cannot hear it without assuming that AI made it. A few weeks ago, I was rewatching the Mad Men episode where Don Draper pitches a watch. “It’s not a timepiece,” he says. “It’s a conversation piece.” A decade ago, I was amazed by Draper’s elegant turn of phrase. But now I can’t see it without thinking that a chatbot vomited it out between daytime scotches.

There are plenty of other linguistic gimmicks that appear to come direct from ChatGPT. Vague, soft intensifiers are one: if you ever see anything described as “quietly powerful” or “deeply transformative”, then that should set your spidey-senses tingling. ChatGPT is also known for being a bit too liberal with em-dashes. So am I, but the robots can rip them from my cold dead hands. Nevertheless, nothing haunts me quite as much as “it’s not X, it’s Y”.

This is my life now. I’ve become so hypervigilant to the construction that it has seeped into my subconscious thoughts. This isn’t a cup of tea, I say out loud to myself, it’s a precious respite. That isn’t a window, it’s a portal to a new way of thinking. This isn’t food poisoning, it’s a quietly powerful reminder not to eat raw chicken off the kitchen floor.

So now, whenever I sit down at my desk, I waste all my energy trying not to write any variation of “it’s not X, it’s Y”, because I don’t want you to think I use AI. It’s much harder than it looks. I literally used it four paragraphs ago, with the datacentre thing. It has made me even more determined to prove that I’m a human. What do I need to do? Come to your house and free-associate a column at you? Send out vials of my saliva? I’ll do whatever it takes.

Hopefully this won’t be for ever. AI evolves so quickly that “it’s not X, it’s Y” will soon become a thing of the past. It will probably be replaced by a new stylistic quirk, no less sinister but harder to detect. And if that doesn’t happen, you have my full permission to lock me away for my own safety. “This isn’t incarceration,” you can tell me as you slam the door. “It’s a quiet reset.”