Back to Blog

The copy bots are in our heads

Tom Lucente, US Creative Director

Artificial intelligence changes everything. Even well-established professions untouched by other technologies have seen AI land at their doorsteps. My own is no exception. But, do I really think AI will transform copywriting as we know it?

Simply stated, I think it already has.

Case in point: the emergence of content as the bright, shiny object in the marketer's toolbox. Writers and other creators are employing content in ways they never imagined using copy. Then again, the latter has long been a product of style over substance, persuasion over communication.

Not so with modern content. It's often deeper and, if the data's right, more relevant to the desires of today's audiences. With ABM and other digital technologies, content is downright personalized—often to a one-to-one level. All of which is why content creation is less about headlines and more about headspace. That makes content ripe for automation and, yes, AI.

When I was recently asked to consider the potential benefits of adopting AI-enabled tools at April Six, I welcomed the question. Putting aside the emotional (and, for some, apocalyptical) reaction to the thought of machine-learning copy bots replacing living, breathing copywriters, it wasn't an unreasonable request. We've clearly entered a new era in marketing, and some integration between human and artificial intelligence is inevitable. The risk of disruption is too great to think otherwise.

Besides, no decent copywriter is getting replaced anytime soon, right? (Actually, hold that thought.)

AI comes in peace (mostly)

I've learned, perhaps reluctantly, that copywriting AI is not the enemy. In my experience, the vast majority of today's copywriting tools were designed not to replace humans but to make us better, more efficient writers and editors. The current crop of offerings leverage core AI capabilities to automate the mundane, the elemental, the repeatable. That said, the promise of the technology has always felt more, well, promising than that.

It's not that we haven't seen significant signs of change. In 2016, Goldman Sachs' landmark investment in what was then called, "The world's first automated copywriting startup," ignited the era of AI-based content creation—or so it seemed. That same startup (Persado) now says it's, "Reinventing marketing creative by applying mathematical certainty to words." Vague, yes. Interesting, perhaps. But hardly a foot-stomping, flag-waving declaration of copywriting independence.

Maybe it's premature to conclude that we've entered a glorious, new era of content creators enabled by artificial intelligence. According to Econsultancy.com blogger Rebecca Sentance, “Despite fears of an AI takeover, at present only a handful of companies in the marketing and advertising space offer tools for automated copywriting and language optimization.”

The fact is, no one has yet cracked the code that would enable an AI-based system to "write" creatively conceptual, natural-sounding copy at the push of a button. Perhaps software developers have realized that writing for human consumption is difficult—even when it's done by humans.

It's not for a lack of trying. The range of AI-enabled writing solutions is already a wide one. Ever use the proofreading editor embedded in Microsoft Word? That's copywriting AI. Not enough AI for you? How about Phrasee's tool that claims, without hesitation, "To write better subject lines, Facebook ads, and push messages than humans"?

Likewise, the AI-enabled grammar-checkers—Grammarly, Ginger, WhiteSmoke, ProWritingAid and others—have developed their own approaches and attracted their own followings. With more than 6.9 million daily users, Grammarly's system combines machine learning and natural language processing. According to the company's website, "Natural language processing is a branch of AI that involves teaching machines to understand and process human language (English, for instance) and perform useful tasks, such as machine translation, sentiment analysis, essay scoring, and, in our case, writing enhancement."

And, finally, there's AI Copywriter from the digital-marketing division of China’s Alibaba Group (the largest retailer on the planet). The proprietary tool utilizes “deep learning and natural language processing amassed from millions of top-quality existing sales" to generate text on Alibaba-owned websites. Gone is the pesky task of writing copy. Instead, users choose from a set of machine-generated options. Within these closed, retail environments, AI is being utilized not so much to assist professional copywriters, but to help everyday sellers and business owners market their products on China's most popular social platforms.

New discussion, old debate

It’s one thing for an AI-based copywriting tool to pass the Turing Test (exhibiting intelligent behavior equivalent to, or indistinguishable from, that of a human). It's quite another to elevate copywriting to something resembling art. After all, isn't that what great copywriters and campaigns do? Shouldn't the bots be held to an equally high standard?

Volkswagen's “Think small,” Nike's “Just do it,” and Dove's “Real Beauty” are the copy written products not of data and schema, but of creativity and cultural literacy. As Goodby, Silverstein & Partners Chairman Jeff Goodby puts it, "The pursuit of artistic merit and success for the brand are connected. Art gets things noticed and re-evaluated.”

Our 21st century argument is reminiscent of a famous 20th century debate in which opposing schools-of-thought wrestled for marketing mindshare. On one side were the "neo-traditionalists"—Rosser Reeves and David Ogilvy included—who saw advertising as a science to be articulated, reiterated, and measured. On the other side were the "creative revolutionaries"—most notably, DDB's legendary Bill Bernbach—who were dedicated to driving memorable engagement through unexpected, unconventional creative.

That helps explain why, as I develop work for clients, and direct the work of copywriters, I feel as though I'm part of a diminishing breed. I still think ideas matter more than data—although I can't deny that the evidence to the contrary is growing.

Wherefore art thou copy bots?

Words mean as much as ever, but what meaning and whose words? Lawrence B. Solum, Professor of Law at Georgetown University Law Center, has defined artificial meaning as that which is, "Produced by entities that are not natural persons. One example of artificial meaning is provided by communications that are generated by AIs.

"Today, such meanings are produced by the complex algorithms that produce the various texts generated by Amazon.com or Siri. Even today, these AIs produce meanings that cannot be reduced to the meaning generated by the programmers (and others) who have input into the texts," he believes. "These AIs say new things because their algorithms combine words into sentences that have never been uttered before."

Combining words into sentences that have never been uttered before? Sounds like the work of a great copywriter, no? Apparently, it's also within the wheelhouse of a transformative technology. And that brings us back to the idea of decent copywriters losing their jobs. I'm afraid I have no credible theories on that. Fortunately, my April Six colleague, Director of Strategy Trent Talbert, has an optimistic thought for all concerned.

"In a world where two plus two equals four (repetitive or formulaic content, for example) a computer can create faster and better," he says. "But, in a world where two plus two equals five (say, the creation of brilliant lines like 'Impossible is nothing' or 'Got milk?') it takes a human to make the leap. That’s where AI falls short today: the big leap. Creating art." 

However it all shakes out, there's no question that the copy bots are in our heads. It remains to be seen whether we take them into our hearts.