Will Robots Replace Copywriters?

Article by Daniel Smith | August 29, 2018

Copywriters used to laugh in the face of robots. Not any more. As AI descends upon a vast range of content, generating everything from headlines to football reports, human copywriters are looking on in disbelief at how quickly their field is succumbing to automation. In fact, when it comes to creating optimized text and maximizing engagement with your audience, there’s an argument to suggest that AI can already do it better than humans.

Not to be upstaged at their own craft, discussions about AI in copywriting have been preceded by a wave of doom-laden headlines, with one writer after another predicting the death of the industry at the hands of a merciless machine. However, once the initial fear subsides, a host of questions remain. Many copywriters are coming to grips with machine learning technology for the first time, and are wondering how AI is already able to do creative work.

To answer these questions and make more grounded predictions about the evolution of copywriting around this new development, it’s important to understand how AI platforms are able to write in the first place.


What is Natural Language Generation?

Our introduction to Natural Language Processing (NLP) gives a sense of the wide variety of mechanisms that play a part in AI understanding and producing language. When it comes to writing, the most important of these processes is Natural Language Generation (NLG).

Let’s imagine that you’ve asked Siri to show you the best pizza places in New York. As part of Natural Language Processing, Siri has translated your speech into instructions and used them to compile a data set of the most relevant places. Natural Language Generation is the final part of the process, where Siri says “Here’s what I found” and presents the list to you in generated text. In short, NLG is the process by which a machine turns structured data into text that humans can understand.

As simple as it may initially seem, NLG goes far beyond just uploading information into a template and displaying it on your phone screen. Instead, teams are developing AI that can identify the focus and intent of a piece of content, taking everything that’s relevant and constructing it into an easy to understand, trustworthy piece of text. Essentially, NLG tools are being designed to write content in the same way as humans.


How can AI write copy?

While NLG technology was originally designed to perfect the transformation of structured data into analytical reports, it has found a wide variety of uses in copywriting. This is largely because writing short copy, like landing pages, is far less creative than it may initially seem. As Rory Stobo explains:

Marketing copy is only very rarely about saying something new. People have trouble processing ‘new,’ and instead respond much better to simple emotional button-pushing.

As a result, the approach to short-form copy is generally pretty formulaic. This means that NLG software is able to analyze which words or phrases produce the most engagement in any given consumer base and create optimized content for that audience.  For example, companies like Phrasee have had success in generating headlines and email subject lines that fit seamlessly with a wide variety of brand voices. Persado plan to take it one step further, using their algorithms to create personalized copy for each customer based on individual engagement profiles.


So, will robots replace copywriters?

AI seems poised to play a greater role in copywriting in the near future, but humans don’t need to lose sleep over it just yet. NLG tools excel at writing content with clear templates, simple aims and set criteria. However, the process encounters significant challenges when it comes to creating longer content with more ambiguous or complex goals. The native-level mastery of a language needed to bring a text to life through humor, nuance and subtle changes of tone is well beyond AI’s current abilities. As a result, it will take several more stages of development before machine learning can craft the uniquely compelling stories that support successful marketing strategies.

Despite the doomsday predictions, human copywriters will have the edge for the foreseeable future. Instead, the signs point to cooperation rather than destruction. By shouldering the burden of more predictable content, AI will enable copywriters to do more of what they excel at: developing unforgettable brand voices.

If you’re searching for someone to write the content that will take your business to the next level, Lionbridge AI has all the tools and experience you need. With 500,000+ certified language specialists in 300+ languages, our crowdsourcing platform is built to optimize your content for global markets – or to help your AI learn to optimize itself! We can fine-tune your data through the creation of custom datasets, tags or analysis. Any AI is only as strong as its training. Lionbridge takes pride in making your platform better.

The Author
Daniel Smith

Daniel writes a variety of content for Lionbridge’s website as part of the marketing team. Born and raised in the UK, he first came to Japan by chance in 2013 and is continually surprised that no one has thrown him out yet. Outside of Lionbridge, he loves to travel, take photos and listen to music that his neighbors really, really hate.


    Sign up to our newsletter for fresh developments from the world of training data. Lionbridge brings you interviews with industry experts, dataset collections and more.