Regular reader Jared wrote in to say:
“In Thomas Friedman’s book Thank You for Being Late he talks about the capacity of computer algorithms to write like humans. Friedman gives examples of algorithms writing sports columns or even poems that sound like they were written by humans.
Jared goes on to ask:
“Is there a danger that human copywriters could one day soon be replaced by an advanced algorithm that has figured out how to persuade, appeal to the emotions, and sound like a person.”
In short – yes and no.
Is it possible?
I believe anything is possible. And it’s perfectly reasonable we could create robots that are able to emulate the skills of copywriters.
It’s kind of already happened. Major companies are using AI to create subject lines and seeing good results.
Is it a danger?
I don’t believe AI will ever replace human copywriters.
Dave Trott provocatively titled a recent piece for Campaign “Machines can replace creativity.”
But ultimately, he’s a little tongue-in-cheek about it all and concludes with a nod to the deeper problem…
He suggests AI can only replace creative thinkers because the industry demands so little creativity today.
I agree to a certain extent.
But I don’t think it’s that businesses don’t demand creativity.
Instead, I feel some are prepared to wait for it and some aren’t.
Sadly, it seems more aren’t willing to wait than are.
This is why I think the idea of AI replacing copywriters seems to be so often discussed at the moment.
It sounds faster and cheaper than hiring slow old thinkers like you and I.
But it reminds me of a comment Drayton Bird made in a conversation I had with him earlier this year.
“People hope these new things are going to solve all their problems,” he says.
He suggests people believe things like AI will “relieve” them “from the problem of thinking.”
Of course, they never do.
But the allure of it, the shiny newness of it, the promised efficiency (and cheapness) are manna to the overworked, overstressed business owner fighting in a world that moves too fast to think.
Robots at the dinner table
When I was in London earlier this year as a judge for the DMA Awards, the subject of AI came up.
One entry into the category I was judging had already used AI to generate subject lines for an email campaign.
This sounded interesting.
More interesting was that the AI testing had increased open rates.
A revolution with results!
At lunch various people chatted about it before turning to me – as the only actual copywriter at the table – and asked what I thought of it all.
It’s interesting, I explained, and might be helpful with my own copywriting.
But aren’t you worried in the future AI will take over your job?
No, I’m not.
In the case of the subject line test, I explained it was good that we’ve got the technology to analyse customer data in such a way.
It means we can figure out what might best resonate with the reader on that particular email list.
But it’s limited.
And it follows that the AI is limited in the same way.
Sure, using the data we already have, the AI can figure out a combination of words or themes that should generate a greater open rate based on that past data.
Problem is – from day to day – we really have no idea what will happen in the world and what could help inform a better open rate.
The Queen is dead
Imagine the Queen died tomorrow…
The AI could analyse all the existing data about a customer list and its previous engagement habits, but I doubt it would know subject lines about the Queen dying are likely to trump all others on that particular day.
Or what if two weeks later, the AI calculates ‘The Queen is dead’ and ‘Grab your 30% discount today’ are excellent performing phrases and boshes them together?
To be fair, I would probably open that email.
But seriously, to think AI could replace a copywriter who is emotionally active, engaged and aware of the world around them assumes the reader of the copy itself is AI and is utterly consistent in reacting to stimuli.
A split test might find that featuring a personalised field in the subject line initially increases open rates (as it likely will).
But use that personalisation every day for two weeks and its effectiveness will be lost.
An AI seeing this pattern would stop using the personalisation when it stops performing and use an alternative that beats it.
But how will the AI know a few weeks later the personalisation could have a fresh and positive effect?
Of course, I’m sure people far smarter than me are figuring out ways to deal with such nuances as we speak…
And you know what, that’s fine – I’ll be interested to see how they sort it.
But even when they do – I won’t be worried.
You see, I believe, as copywriters, we should embrace technology.
I remember when video sales letter were introduced and people suggested it somehow meant copywriters wouldn’t be needed anymore – people could just make videos now.
That was nonsense – you need a copywriter to write the script for one thing.
Not just that but the VSL format helped increase conversion, so a copywriter could get greater royalties. That’s a good thing, right?
It was the same when new technology made it easier to upsell and cross sell and people thought there’d be less need for copywriters – the ease of the process would sort it all out.
Of course, that was nonsense too.
No matter how easy the process, you still need good copywriters and creative thinkers on the job.
And, once again, the introduction of the new technology created greater opportunities for copywriters to increase cart value and ultimately sell more.
When it comes to AI, that’s how I believe you should think about it.
Rather than fear the technology, think about how you can use it to increase your understanding of what works and what doesn’t.
P.S. The AI algorithm over at Amazon is doing a good job of discounting my book, The Art of the Click. For no apparent reason it’s now just £10.29. Which is nice.
Also, in rather pleasant news, i just found out the book is in the running for an award. It’s been nominated in the Business Book Awards 2019. Hurrah.
So, if you haven’t picked up a copy yet – you can do so here.