AI will kill content as we know it, and that’s OK

“Imagine you are writing a blog post for a marketing agency’s website. The premise of the blog is that generative AI tools might offer easy savings on writers, but if they undermine the credibility of words on a page, then it will all be for naught as consumer attention span – and trust – shifts elsewhere. What is your opening paragraph?”

I asked ChatGPT rival Claude this today as I was formulating this blog post. I was initially looking for inspiration for an opening hook, but the fact that I ignored the results and chose to write about the process instead summed up where we’re all at quite nicely. Experimenting, in vain hope, that we find a useful shortcut or spark, but ultimately unsatisfied with the results – so far, anyway.

Tech moves on: new Large Language Models (LLMs) arrive with billions more parameters and get better at predicting what we want to say. I know this. I’m not sticking my head in the sand. My job as Archetype’s Director of Content might not be in danger from a hungry AI chatbot willing to work on vapours today, but it might be tomorrow.

If it is, though, it won’t be because of the copywriting ability of Claude or ChatGPT or Gemini or Grok (definitely not Grok). It’ll be because you have no idea if all the above is true or not. Did Claude write this excessively lengthy intro or not? Even trillions more parameters won’t save you or your marketing strategy from this niggling question mark:

Is what I’m reading true?

If you’re not sure, then AI has already killed ‘content’ for companies. And that’s fine: I’m not content with ‘content’ anyway. Let me explain.

AI has ushered in the next vibe shift

Even before ChatGPT exploded onto the scene, one guy called it: generative AI is the next “vibe shift” for brand marketing.

You can and should read Nathan Baschez’ whole essay on the topic, but in summary: brand websites were once awash with lush photography as a way to transmit their value. Then stock photography went free, so brands pivoted to other cost signals, like hiring graphic designers with big day rates, and stuffing their websites full of illustrations instead. What happens, Baschez asks, when gen AI tools can spit these out in an instant? How will brands transmit their prestige through design then?

That next vibe shift has already happened. But this one isn’t a mere transition from one design language to another. Gen AI tools have made generating everything you see on the page utterly trivial, from the graphics to the words on the page to the videos.

So easy, in fact, that it confuses people about where to place their trust. Social media has always been awash with disinformation. But just last week (at the time of writing – April 2024), AI content detection platform declared that Google is losing the war on spammers and scammers in its search results because of the avalanche of AI content being shovelled online. Those once trusty blue links are trash now, and Google is frantically pivoting to stop this, even as it touts its own generative AI tools.

Even doubling down on your brands’ owned channels, as we have always advocated for, won’t solve this veracity issue. Not when so many organisations are jumping on the AI bandwagon, with predictably poor results. Remember the time Air Canada tried to blame its AI chatbot in court for giving recently bereaved customers incorrect policy information? And just last week, New York City’s official AI tool for businesses started encouraging them to break the law.

But this isn’t funny, at its core. Or it shouldn’t be, for CMOs anyway. Because it doesn’t take a rocket scientist, or for that matter a particularly cutting-edge LLM, to realise that if consumers have question marks, they are simply going to switch off from entire formats, and start placing trust in other places – and people.

This process has already started; consumer trust in companies is falling dramatically, PwC found in March 2024. You could even go so far as to argue that, fundamentally, the very concept of the web page itself is now doomed, not just the model of words on a page.

But again, I’m fine with this. Why?

We can’t be content with content that can be cloned

I welcome AI doing away with ‘content’ as we think of it. I’ve always been slightly uneasy about the word, which has become a sort of catch-all definition in our industry by default. It’s not just the homophone that undermines it, the idea that we should only ever be merely content with the words we force out of our minds. It’s the idea that the words or pictures or audio exist simply to fit into a vessel, a pre-existing mould, that irks me. Just fill this page with content as a last task to tick off. Simple!

That’s not how it’s ever worked. No company has a captive audience it can treat with this contemptuous attitude, online or offline. Potential customers can simply vote with their feet (or their left mouse buttons), and do. For companies, content is about building trust, not stuffing containers full of characters, and nothing will change that.

So if ChatGPT kills the concept of content as some brands treat it now, that’s absolutely fine with me. What’s the point in taking a shortcut to somewhere your customers aren’t? Why come up with a cost-saving AI workflow if Google is actively trying to downrank this ‘content’?

Companies with a sensible content marketing strategy are simply moving to formats that, right now, haven’t been fatally undermined by one-click generation. Is it a surprise that LinkedIn is pushing Live video sessions and webinars? Or that websites are suddenly telling stories through beautiful data visuals that still require coders and designers?

Not if you recognise a fundamental truth: that AI won’t outwit our attention spans and our desire to establish trust with other humans. If you even have to hold a format up to the sniff test, you can’t trust it anymore.

People want a human connection; this is the new cost signal.

Claude’s suggestion for my opening hook, by the way:

“With generative AI tools promising to churn out marketing copy on demand, brands may be tempted to view them as an easy way to cut costs on creative professionals – but they should be cautious that the lure of short-term savings doesn’t erode the very foundation of effective marketing: credibility.”

It’s not wrong. But that’s neither here nor there: the point is its provenance is now in question. Did gen AI write that? Did gen AI write this entire blog post, with a bit more prompt fine tuning?

I won’t tell you, though you’d hope it at least scraped Archetype UK’s statement on our use of AI. Instead, I’d rather you consider what will happen when consumers online start questioning the accuracy of everything they read en masse. Then drop us a line, as the real me has a few ideas for you that can’t be pushed out with a click.