What is conversational AI, and can it replace a human copywriter?

by Elliott Fudge

image overlay

AI technology has come a long way in a short space of time. While many of us first experienced the technology through tools like Google translate, AI is now poised, we’re told, to take over the world of copywriting.

From machine learning to natural language processing (NLP), the science might be complex but the reality is simple: every day machines are getting better at not just understanding us, but predicting our behaviour as well.

Even though conversational AI is increasingly capable of writing content at pace, is it capable of writing content with quality? We’re not worried.

What is conversational AI?

Conversational AI combines two main concepts in computer science: machine learning and NLP. The former allows machines to organically learn and respond to patterns of data, the latter is a machine’s ability to understand spoken and written language far more accurately than ever before. Conversational AI then, allows software programmes to properly understand and learn from language to such an extent that it can more accurately mimic realistic conversations, or even copy.

Warnings about the future automation of copywriting have been circling for some time. Tech giants Google and Facebook have already ploughed vast sums of money into conversational AI. Considering under the right circumstances a conversational AI can even write short essays, it’s easy to see why. When conversational AI works well, it can produce short, snappy, keyword-heavy marketing copy in a heartbeat. In the tone of a famous podcaster or celebrity? No problem!

That’s not to say AI is perfect – it’s far from it. But when it comes to producing short copy in quick times to fixed guidelines, there’s no doubt it’s impressive.

What about copywriting?

As it turns out, quite a lot. This is a technology in its infancy, and though it can produce keyword-dense copy in double quick time, it usually takes a few tries. As it stands, AI copywriting seems better suited to quantity than it is quality, and where quality matters – regulated reports, tone of voice consistency and expert insights from experienced writers – AI is nowhere near the standard required.

At a time when writing is increasingly proving capable of moving stock markets, share prices and confidence, a human copywriter still holds all the cards:

  • precision: only an experienced writer can judge how small differences can drive big impact
  • judgement: while an AI can string lots of correct words together, it can’t tell if they are the right words at all
  • creativity: machines can only learn from what they have already ‘seen’. When it comes to producing unique and nuanced writing, creativity is still a very human faculty
  • empathy: AI is still a long way from being able to process moving, impactful and driven content

A means to an end

AI copywriting will prove useful, but is it last orders for writers? Probably not. Conversational AI may well mean that writer’s roles pivot towards more specialised, long-form writing, but that’s far from the automation promised.

Copywriters can rest easy in their jobs for now. For the foreseeable future, AI is at worst only likely to rebalance writers’ priorities from writing to editing, and for those writers with specialist niches and skills, AI might even prove useful for getting through a lot of your other, shorter work. A replacement for copywriters? Unlikely. A means to an end? Let’s see.

From quality to quantity?

When the stakes are high, choosing the right words matters. It can’t be left up to a machine and the endless inputting and refreshing of a word generator. Data is no substitute for awareness, expertise and judgement, so when it comes to winning with words, there’s no match for a copywriter.