The fast and the forgettable
Imagine an autonomous car, powered by a rocket engine. It can cut your office commute from one hour to a minute. Incredible.
Except it doesn’t take you to your office. Instead it takes you somewhere… close-ish. And you need to walk from there.
This is where AI copywriting is at the moment.
The speed is astonishing. And the output often looks good at first glance. But it doesn’t get you all the way.
Not, at least, if you need subtle, nuanced writing. The kind of writing that builds subtle, nuanced brands.
Expect the… expected
AI definitely has its uses in copywriting. For instance, it is good at generating plausible language, almost implausibly quickly. And for stretched teams, that can really make a difference. Having a first draft, fast is a great place to start.
But it is only a start.
Because, as you’ve probably noticed, AI has a habit of drifting towards the familiar. The classic category phrase, the obvious benefit, the polished line that looks right because we’ve heard the cadence a hundred times before.
But looking like the thing is not the same as being the thing. Real impact needs to come from somewhere real. Somewhere human.
Where effectiveness really comes from
AI has learned from an extraordinary library of recorded human expressions. These give it range, fluency and a pretty convincing impression of confidence.
But human source material goes far beyond the archive. Personal memory. Emotion. Taste. Frustration, even. The odd private connection between two things that have no obvious reason to meet.
That doesn’t make human writing ‘magical’ – it’s just working from different inputs. And, ultimately, different inputs create different outputs. Those unexpected references, bolder risks, truer truths.
The stuff humans draw upon is messy and internal, but it’s often where the true heart of effective writing beats.
For brands, that is not decoration. It is how a message becomes valuable: by making the offer feel relevant, the promise feel credible, and the product feel worth choosing.
The cost of meaning is time
There is an obvious response to this: brief the AI better. Or edit harder.
And yes, both help. But this is where the human-in-the-loop matters – feeding the machine the customer insight, the brand nuance, the unspoken truth. Then sharpening the draft it produces until it does the job it needs to do.
Which is exactly the point: the more effective you want the output to be, the more human work you have to put in.
Not as a final glance before something goes live, but as active judgment at both ends of the process. What goes in. What comes out. What gets changed before anyone else sees it.
The speed trap
AI may be able to help us produce more copy, faster. But the efficiency claim is less simple.
Used well, AI can help us test, structure and sharpen. But if copy is only measured by how quickly it reaches the page, we risk mistaking speed for progress. And efficiency for success.
If you want copy that has impact, the slower work still matters: letting the unrecorded world feed in, knowing what to leave out, and understanding when the words have earned their place.
AI has solved the efficiency problem – the effectiveness problem is still ours.
(For now, anyway 🤖.)
Using AI to move faster? We can help make sure the words still land. Just get in touch.