The Case for (and Against) AI in Content Marketing

It seems like you can’t go to marketing Twitter without stumbling into AI discourse. As the founder and CEO of Planable, the content collaboration tool, I’ve been following the debate closely.

So thought I’d share my own thoughts on the actual benefits and drawbacks of using AI in marketing, more specifically, in the creative stage. Is there cause for concern that AI will replace creative content marketers, or should we embrace it as a powerful tool?

It’s not the first of its kind

ELIZA, the first chatterbot, was created by Joseph Weizenbaum in 1966. Compared to current AIs, it was pretty rudimentary. Essentially, ELIZA simulated a conversation by repeating users’ input in the form of questions.

Source

But there was a catch: the AI was designed to be a parody of a psychotherapist of the Rogerian school, which involves reflecting the patients’ words back to the patient.

In essence, it used the methodology of person-centered therapy as a template to create a rudimentary language model replicating the idea of sentience.

Joseph Weizenbaum was shocked to see people anthropomorphizing the language-based model. That was not the goal of the technology, and it created a massive ethical debate at the time.

There’s an anecdotal story about Weizenbaum’s secretary asking to be left alone to have some private time with the computer.

Flash forward six decades later. We’re having the same conversations about whether AI is sentient (of course it’s not) or if it will make thousands of jobs obsolete (it might, but not in the way we think).

Unfortunately, I don’t have an answer. But I do have some thoughts.

The Case for AI

Content marketing involves a lot of grunt work, and any marketer can attest to that.

Take research, for example.

First, content marketers haunt social media and message boards for insights. Second, they interview customers and experts. Then, they piece those insights together into marketing material that serves a specific purpose (conversion, generating leads) and offers value to the audience.

Research obviously takes a lot of time. Marketers spend 3.55 hours/week on research alone. That may not seem like a lot, but it adds up.

Then there’s the issue of framing and angling — again, how do you marry business objectives with providing value?

This is where AI can help. Not with collecting data. ChatGPT is notoriously bad at providing accurate data, oftentimes outright making stuff up. For that reason, I don’t see any AI replacing the metaphorical footwork sitting at the core of good marketing. At least not for now.

This take is not unique. Rand Fishkin of SparkTorro pointed out that ChatGPT’s training data cuts off in 2021.

It’s safe to say that taking the information that ChatGPT generates at face value without doing your due diligence is a recipe for disaster (and possibly ridicule).

ChatGPT doesn’t possess “institutional memory”; its compressed version of the internet will never be a substitute for the know-how accumulated over decades of marketing, and it’s hardly conducive to the kind of cerebral thinking that leads to great marketing.

But AI tools can be helpful with framing, structure, and the minutiae of optimizing each sentence to fit those parameters.

Think back to all the brain juice you spent thinking about “angles.”

You might not realize this on a conscious level, but every time you obsess over sentences and flow, you’re actually thinking about framing. Is this sentence tonally consistent with the rest of the article? Will I lose the reader if I subtly inject this paragraph with a small dose of marketese?

Since AI effectively scans the internet for things actual human beings have written, it should give you a rough estimate of how these kinds of articles should look, provided it’s fed the right data and context. Of course, human editorial input will always be needed.

There are two more cases for AI that are worth touching on.

1. Using it as a starting point

That’s the biggest struggle when creating content, filling up that blank page with the first words. It’s most often because your ideas are not yet clear to you. Your opinion has not yet crystallized. But once you have an initial output, you start realizing what you agree and disagree with. It becomes clear what you like and dislike.

2. Using it for infinite iterations

Content has so many nuances. You see words, and you get that vague feeling that “something feels off”.

That, again, frequently leads to a point where you’re stuck. You know something’s off, but you can’t quite put your finger on why. Giving AI feedback and seeing dozens of iterations in seconds can help you get a lot faster to what you envisioned the end result to be.

The case against AI

In the long run, AI can become a barrier for new blood entering the industry.

AI-driven marketing tools, while they might be useful in certain scenarios, can also act as a huge barrier for newcomers.

It’s all too easy to rely on the same old titles and topics over and over again. If there is an AI-backed tool that can churn out content for you quickly and efficiently, why would you bother training someone from scratch?

The second consequence is that it robs writers of valuable experience.

Having a computer do all the difficult work takes away from what makes someone a good writer—experience with various styles and tones of writing. Sure, you could feed data sets into an AI system to train it to write like any given author or style guide, but this would still be far removed from actually writing and honing your craft.

If viewed from a purely business standpoint, then yes, the hours spent on arranging sentences and looking up synonyms, and moving paragraphs around is time wasted. But isn’t that process of iteration exactly what makes good marketing, well, good?

The New Yorker’s Ted Chiang put it best:

Obviously, no one can speak for all writers, but let me make the argument that starting with a blurry copy of unoriginal work isn’t a good way to create original work. If you’re a writer, you will write a lot of unoriginal work before you write something original.

And the time and effort expended on that unoriginal work isn’t wasted; on the contrary, I would suggest that it is precisely what enables you to eventually create something original. The hours spent choosing the right word and rearranging sentences to better follow one another are what teach you how meaning is conveyed by prose.

As someone who enjoys writing, this quote struck a chord with me. Bad writing eventually leads to good writing.

Now, I may be an idealist, but I believe in the power the human brain has for original and unique ideas.

And in “ideas,” I include angles, points of view, metaphors, and comparisons. The thousands of ways humans can spin a topic and a situation. The chase to get content on the first page of Google has already robbed us of much of that uniqueness. Abusing AI could rob us of that entirely.

Finally, AI is a self-destruction machine. At the moment, AI is trained on existing content. Thoughts, opinions, and knowledge shared by billions of people around the world throughout time.

If AI would take over search engines, people wouldn’t create content anymore. So one question inevitably pops up: on what content will AI be trained in the future?

Final thoughts

While AI might have a place in content creation, it should not replace the humans involved at any cost.

There are certain areas of the process where human input is indispensable. And if these aren’t taken into consideration, then there is no way to ensure that high-quality content will be produced. And high-quality content is what eventually brings you loyal and relevant audiences. It’s what builds brand love. The key takeaway should be to use AI as a tool rather than as a substitute for human creativity and experience.