Why Marketers Need to Keep Their AI Content Human

Source: This article originally appeared in Medium

In a recent interview with Forbes, John McNeil, the CEO of a highly regarded brand design studio in Berkeley, Calif., explains as well as anyone I’ve seen to date the emerging challenge for marketing pros to use AI tools in a way that retains the essential human element in the content they produce.

“As much as I’m critical of the way in which folks are starting to deploy AI, I think that the reality is that you have to use AI, and AI is actually a phenomenal tool for creative people,” says McNeil, whose company John McNeil Studio creates copy, designs, photo, films and music for major business clients.

“There are tons of things that have been created by AI, or assisted with AI. I’ve shot commercials where we do character replacement, and we do automated masking and things like that using AI,” he adds.

Having established that he’s no AI hater, McNeil goes on to make a strong argument about the need to deploy GenAI in creative processes that maintain a quality of human authenticity. Otherwise, you risk losing the trust and interest of your customers and target audiences.

“It’s not that AI as a tool is bad,” he says. “It’s this idea of relinquishing all creative standards and all craftsmanship to this novel thing that’s going to make it for you quick and fast.”

Keeping It Real

At Presspool.ai, the company where I currently serve as VP of content, we often lean on GenAI for certain time-consuming tasks that might otherwise consume a lot of resources, but aren’t determinative to the quality of any final work we produce.

That said, given some well-formulated prompts (an essential skill for creators to get the most value from GenAI), ChatGPT in its current version can serve as a fairly clever but un-creative research assistant that scours the web, producing a raw brief of publicly available research and analysis related to the copy you’re developing.

How you treat that raw material produced by GenAI then becomes a question of ethics and, in some cases, copyright. But beyond these important issues, there remains the ever-critical challenge as to whether or not your content will resonate on a human level with your target audiences and customers you seek to engage.

Some recent research findings about how different demographic age groups today perceive content that’s generated using AI provide relevant insights here. GenZers in particular are better able than older audiences to discern content that relied upon AI, and they want brands not to attempt to conceal the fact.

It isn’t that GenZers are fundamentally against AI-generated content — they just demand transparency and authenticity in the content they consume. Otherwise, they’re liable to turn away from any brand they perceive as trying to pull a fast one using AI.

“Be aware of your brand when using AI and use it in a way that makes sense with existing personality and values,” advises Subha Sivakumar, a student at New York University, in a recent PR Week article. “Or,” she adds, “stand out by highlighting the humans behind your campaigns, emphasizing how their artwork and contributions are important to your brand, supplemented with AI features.”

To Label or Not to Label

Which brings up the whole question of whether or not — or even how — to label content that was produced to some degree using AI. After running through the various key considerations in this regard, Julia McCoy, president at Content at Scale, offers some sound advice.

“If you’re publishing AI content with quality in mind and include your human touch in the production process — AI labeling isn’t necessary,” she says. “However, there may be some instances in which you may want to consider it depending on the context, topic, and content you’re sharing. It all comes down to good judgment.”

A simply stated principle, and one that’s hard to dispute. Relying on common sense and keeping up with changing attitudes about AI content labeling will remain paramount to success for marketers going forward.

In a recent study conducted by David Rand at MIT Sloan School of Business, he points out that the line between when you should label content if AI played some role in its development, and when not to label, will likely shift as public perceptions about the tech continue evolving.

“Under the principle of implied authenticity, the more content that gets labeled, the more that content without a label is assumed to be real,” he points out.

Avoiding Downward Spiral

Ultimately, what GenAI content represents is a rehash of pre-existing material, and for any copywriter who has spent much time using the technology to get work done this truth becomes painfully obvious. Not that a re-hash can’t be useful. Again, it’s a way to get up to speed quickly on a given topic, but it doesn’t take it to the next level.

Without the constant influence of human creativity adding to the collective body of new ideas and innovations in language and art, the source material for GenAI to scrape becomes finite, and everything it produces increasingly predictable and plainly derivative.

As McNeil puts it brilliantly, “It’s a real spiral that I think is going to change the way in which we view art, frankly, not just commerce. Some of it is that we’ve lost this idea of this human dialogue, this continuum, where we’re building on art history, we’re building on the definitions of what we’ve come to understand a creative idea to be, or what we’ve come to understand marketing and communication to be. And we’ve passed off a lot of the memorable aspects of that for the desire for something that could be made quickly, much cheaper, and on the fly.”

As long as we human content creators don’t get lazy and rely on AI to do the most imaginative parts of our jobs for us, then there should be nothing to really worry about. It will just be another (powerful) tool in the kit that enables us to produce work that audiences will recognize as fundamentally original, authentic, and human.