I read recently that students and companies alike are beginning to work errors into their work. Whether it’s spelling mistakes in essays for students, or grammatical errors in blogs, whitepapers, or other content the reason is the same. We’re aiming for the new imperfection to make it look like it’s been written by a human. Not AI.

It’s an interesting gambit. For, as long as marketing has been pumping out copy, we’ve all, like the Borg, been aiming for perfection in our writing. There was a view, and rightly so still is, that imperfections in written copy have the potential to advertise imperfections in the product, service, or solution a company is providing.

So Where Does That Leave Us?

The problem we face now, with Claude, CoPilot, ChatGPT and Gemini among others, is they are so good, and becoming so widely used, that we’re now all wondering what has actually been written by a human and what by a Large Language Model (LLM).

So, how do we distinguish ourselves against them? What’s happening now is copywriters, and students, are starting to put mistakes into their work to show that it’s been written by an actual human being, not an LLM. What was once seen as an imperfection and subpar piece of work is now starting to be viewed as authentic and real. The irony isn’t lost on me.

A Game You Can’t Win

The problem is, it’s only a matter of time before LLMs pick up on this and start copying it. So, you may start to find the odd spelling mistake in the work they provide for you. They are imitators, not creators, after all. What happens then? Do we then strive for perfection again and start off this vicious cycle of perfections vs imperfection?

The human-error strategy will always be a losing game, because the real issue isn’t about spelling mistakes. It’s about trust. A business that outsources its voice wholesale to an LLM isn’t just producing polished copy, it’s removing the relationship.  Gone is the human element, whether it’s original thought, humour, tone of voice. That’s what they’re actually responding to.

The Advantage We Still Have

And that brings me to where I think the genuine advantage lies. LLMs are not creative, not yet. Even Data, as advanced as he was, struggled to get real creativity into his music and paintings. Let’s not even mention his sleep-inducing poetry recitals! We still have the monopoly on original. So let’s use it. Seek out the pieces that talk from the heart, or discuss an angle that offers a genuinely fresh perspective. Don’t just produce content. Have a point of view.

I’ll hold my hands up. I use LLMs in my own work, to proof, to refine, to pressure-test an argument. But the thinking, the angle, the voice? That has to come from somewhere real. That comes from me. That’s the standard worth protecting.

That is, at least, until LLM’s have mastered the art of creativity and start getting their works hung in the Louvre.