UX writing and AI — the future is written
This isn’t a blog bashing AI or saying it’s going to take over the world and we’ll all be mindless servants to its indomitable will. I for one welcome our new overlords…
But the reality is AI will likely become another tool. A powerful and potentially dangerous tool — like the car, air travel and the internet. When used well it will help a lot of people achieve a lot more than they were able to before. When used recklessly, it will doubtless pose a significant risk. Just how much risk is at the heart of numerous AI discussions right now.
But what of UX writing?
As arguably the newest writing discipline with short instructional copy at its heart, AI may seem like a logical tool with which to take over a vast majority of UX writing work. Especially if it can automatically understand the best, tried and tested use of copy in set locations — such as error messages or within the checkout flow.
So does the death knell strike for UX copy? No, I don’t think it does. And the reason can be summed up quite neatly: tone of voice.
The longer I’ve gone on in my writing career (20+ years and counting!), the more tone of voice matters. At first, I thought it was simply a basis for good writing: write as you would speak to a friend, use everyday words, keep sentences short and easy to understand… but there is more to it. Despite the fact you’ll see those descriptions of tone of voice in just about every brand style guide you’ll ever read!
Tone of voice is vital
Tone is an area where I feel AI will struggle. Good UX writing, where understanding of human nature and behaviour is required to deliver the most user-friendly outcome for a specific customer base, is not something easily recreated using AI.
Fundamentally AI will use commonly used phrases and proven high click call to actions. But how should it choose between welcome options of ‘Welcome’, ‘Hello’, ‘Hi’, ‘Hi there’ or ‘What’s up? One of these options may ‘win’ under certain criteria, but do you want to end up with your brand welcoming customers with ‘What’s up, dawg!?’ across the whole estate?
Tone is vital to a brand. People can be made to feel privileged customers of an expensive brand (‘Reassuringly expensive’) and happy to pay more or they can be made to feel they’re getting a good deal from a tried and trusted favourite (‘Mum’s gone to Iceland’). Tone of voice helps brands make an emotional connection to their customers, which is difficult to measure and is easily lost by poor communication.
Standing out from the crowd
If you want to stand out from the crowd — and not use cliches like ‘stand out from the crowd’ (!), it requires imagination and creativity. In fact, very human qualities. AI can mimic human qualities very convincingly and the way great copy reads, but it can’t mimic originality.
As a writer you’re told not to repeat words and to use different structures to engage an audience. Like this. A short, punchy sentence will break up a long paragraph and catch the reader’s attention. Why? The human eye is trained to look for patterns. If common words are used repetitively it helps a reader skim a page to find the standard phrases they expect. This can be useful when you want to help a customer navigate a webpage quickly and skim read. But — and it’s an elephant sized ‘but’ — if you want your copy to grab the reader’s attention by the ears, AI is not going to be able to do that because it’s not being programmed to do things differently — it can only do what’s been done before. Having tried out a few AI tools, you’re very often left with a rehash of relevant themes and ideas in response to your brief — a bit like a TV remake. There’s nothing really new.
There are currently AI tools that claim they can replicate tone of voice but if originality, wit and human empathy are at the heart of your tone of voice that’s very difficult to replicate.
AIs innate bias
One fundamental concern over the use of AI is its bias to what it has been programmed to regard as ‘successful’. It becomes worrying when you consider how AI could use the actions of many to make assumptions about individuals. Those in minorities due to race, income or accessibility requirements may find themselves excluded: when shopping online they may only be shown items aimed at a more mainstream user. You could argue that humans suffer from unconscious bias so what’s the difference. The issue arises in that by surrendering control to automated systems we lose the human perspective and make it harder to reverse system errors and inequalities.
Location, location, location…
A crucial part of UX writing is understanding where to put a message to have the most impact. Often you have to have a designer’s eye for a web page to understand how its structure will determine what the user will do — and what you want them to do. Very often a message may need to move to have its desired outcome. The words might be fine but the location is not right and they are not having the desired outcome. An awareness of this potential pitfall is crucial — and probably why many UX writers are now called Content Designers!
So no good?
Let’s be clear, AI does have a place. By providing useful ideas, words and phrases it can help a UX writer do more faster. And it’s already being used by many businesses in customer services to handle customer queries through the use of chatbots and email requests. Where they cannot offer the solution immediately then can certainly point a customer in the right direction.
It will also certainly help less experienced writers put together more meaningful long-form copy and help with things like structure. For more experienced writers it will be a good sounding board for different ideas that are commercially or proven to be successful. But currently it certainly requires a human ‘sense’ check.
‘Good writing will always remain relevant.’
As some fears over ‘big brother’ watching you grow, it’s worth finishing this blog with the writer whose book ‘1984’ coined the phrase, George Orwell. He wrote six rules for writing and here’s the first:
- Never use a metaphor, simile, or other figure of speech which you are used to seeing in print
Originality is at the heart of what he requires from good writing. Not mimicry or copying overused phrases. Until AI is capable of independent creative thought then good writing will always remain relevant. I think UX writing will become even more important as brands look to set themselves apart and develop a unique conversational tone to which customers can feel more loyalty towards.
As I began this article by saying, AI will not take over and the reality will likely be more straightforward. AI will become a tool — and like all good tools, in the hands of the right practitioner it will become an asset.
It requires someone with experience to input the right data to provide relevant copy — and a writer with a good eye to cut through the wealth of ideas they could be inundated with.
If you rely on AI too heavily and follow it blindly, you move towards middle-ground mediocrity (‘vanilla’) and may well find yourself alienating minority customers. Inevitably you will lose that human connection — which is what good writing is really all about.
