Is AI The Future of Content Creation?

Technology and content creation can often appear at odds with each other. The printing press probably put a real damper on scribes, the video hurt the radio, and the internet broke the magazine. And yet, the creation and distribution of content became better with every innovation. Why wait for someone to write when you can get it printed? Why only listen when you can watch? And why wait to receive it in the mail when you can immediately get it on your computer?

However, during the introduction of this technology, it can seem threatening. I suspect there is similar uncertainty concerning AI when creating content.

Earlier this week, there were reports that some of the larger, SEO-first websites were using AI to create some of their content. Tony Hill found that multiple sites, including Bankrate, CNET, and CreditCards.com were explicitly calling out that AI helped create content.

On Bankrate’s site, for example, the author of one of these articles is listed as Bankrate, and the description of the said author is, “this article was generated using automation technology and thoroughly edited and fact-checked by an editor on our editorial staff.”

That’s fascinating. In essence, give the AI a prompt, have it write an article, and then have the editor review it and ensure the information is true. If this works, you can imagine a world where more content can be created without needing to grow your staff in lockstep. So there is some merit there.

Yet, the other argument is that we don’t need more regurgitated information. At some point, there’s enough service journalism about the same topics. So I think there’s merit to this as well.

From a business perspective, though, does it make sense to do this?

SEO ramifications

This is where it gets a little more complicated. It makes perfect sense if we look at it at a surface level. The AI can create what appears to be unique content that answers a specific prompt. A typical SEO practice is creating content answering specific questions since there’s less competition in the SERPs. So, instead of staff writing it, give the prompts to the AI and have it do it for you.

But if we start to peel the onion back, it begins to become more complicated. For example, Danny Sullivan, Google’s Search Liason, tweeted:

For anyone who uses any method to generate a lot content primarily for search rankings, our core systems look at many signals to reward content clearly demonstrating E-E-A-T (experience, expertise, authoritativeness, and trustworthiness).

What experience and expertise does an AI have? Why would it be authoritative? Can you trust AI? According to Google:

Google’s automated systems are designed to use many different factors to rank great content. After identifying relevant content, our systems aim to prioritize those that seem most helpful. To do this, they identify a mix of factors that can help determine which content demonstrates experience, expertise, authoritativeness, and trustworthiness, or what we call E-E-A-T.

While E-E-A-T itself isn’t a specific ranking factor, using a mix of factors that can identify content with good E-E-A-T is useful.

This is conflicting information. On the one hand, it says that E-E-A-T informs priority. On the other, it says it’s not a ranking factor. Which is it? Will a publisher be punished for using AI because how can AI have authority or experience? Or will it not matter at all? It’s impossible to know what Google will do; however, history shows that what works today won’t tomorrow.

In the late 2000s, content farms were all the rage. Sites like Demand Media would create tens of thousands of articles about everything with the hope of getting traffic. And then, one day, Google introduced the Panda update.

At the same time, buying links was standard. I’d make legitimate college beer money from my early blogs by selling links, and my domain authority wasn’t strong. Then Google introduced the Penguin update, and you got alerts for having spam links.

We all transitioned to guest blog posting, trying to get that illustrious backlink from the site. It wasn’t as scalable as just buying, but at least it looked editorial in nature. Then Google announced it would be dinging people who did this too much.

I remember a time when infographics and widgets were all the rage. If you could get a lot of sites to embed them, you’d get automatic links back to your site. Then Google said this was not allowed.

Over and over again, Google would say that something that was once all the rage was no longer acceptable. And we’d have to evolve.

Now, does this mean that thin content went away? No. Nor does it mean we stopped buying links or using guest blog posting to grow organically. Google is not perfect and can’t catch every infraction. But when content creators start scaling these sketchy tactics, Google does catch on.

This brings us back to AI. Bankrate and CNET are doing it at a limited scale with sub-100 articles. And they have editors reviewing everything. But what happens when they spin up 10,000 less heavily edited articles? At that point, does Google do something about it?

Other use cases

So, let’s remove the discussion about the Google ramifications. Instead, let’s look at other potential use cases for content creation.

First, sporting event recaps. If you break down a basketball game, for example, it’s actually a lot of structured data. Player A passed the basketball to player B, who shot and scored. Or, player C blocked player B. Every missed and made shot is tracked.

If you look at that data, there’s a story to tell. Did player B miss more shots than average? “Unlike her previous few games, Player B shot much more poorly, only hitting 3 of her 12 shots. But Player A carried the team, hitting 9 of her 13 shots.”

For any structured data, do you need a human to narrate it? The AI can likely analyze it and push out a story in seconds.

Second, AI can help with writing, even if it’s not part of the finished product. We all know that getting started from a blank screen is difficult. Giving a prompt to the AI could be a way to help overcome that problem. In addition, it can provide ideas to chase.

The risk is unintentionally plagiarizing. While you can’t plagiarize a machine (I don’t think), can it plagiarize someone else? And so, as a content creator, it has to be an inspiration rather than you taking it verbatim.

Tech is normal

There’s no stopping the evolution of technology. And that technology will naturally flow into content creation. Finding ways to use it is never wrong; however, we should understand the risks. It is human nature to abuse tools. The SEO industry is full of it. But are there ways to give users what they want while relying more on software? I think the answer is likely yes.

Thanks for reading today’s newsletter. If you have thoughts, hit reply, or join the AMO Slack. Have a great weekend!