All your free AI ‘content’ isn’t doing what you think it is

The inferior footwear is all the free AI-generated marketing content that two years ago was a mere trickle, but now has become a tsunami.

So, as it’s evident that we'd like to persuade you to think more deeply about your use of AI-generated content, perhaps you've already stopped reading--feeling that this is simply another plea for human, expert-based marketing, communication, technical writing, content generation, and editing. In other words, something you have to pay for.

If you have stopped reading, we can't get you back… but if you are still with us, we suspect it's because you may have a slight suspicion that things aren’t quite as rosy with all this amazing free material as they might seem.

Many people draw a parallel between the rise of AI-generated business and B2B content and the early days of machine translation. When machine translation started, it was genuinely terrible—full of errors, risible in its accuracy, and many professional translators were highly sceptical that it could gain any traction.

They were discomfited, because clients and the public rather liked the idea of free machine translation even if the quality was abysmal, as it was, well, free. And so its use rapidly climbed, so too did the corpus of text available to machine translation engines, which led to rapid improvement. Allied with lot of back-end engineering, machine translation (Google Translate and DEEPL being two great examples) is today of a very high standard.

Along the way, however, many professional translators—whether bilingual by family circumstances, university studies, or professional training through organisations like the UN or EU—found themselves in an employment “nuclear winter.”

If you’ve met a translator recently it’s very likely they worked part-time, and almost certainly as an AI system’s trainer, not as a freelance independent anymore.

Is your free AI generated marketing ‘content’ adding to your brand?

Very possibly, the same thing could happen to professional copywriters. That's not, however, the main point of our discussion. What we’re genuinely concerned about is that. much like the early days of machine translation, many clients are simply turning on the AI tap and assuming what comes out is “good enough” to use without any fact-checking, editing, shaping, or improvement.

In our experience so far, this is particularly true with clients whose first language is not English and who therefore prefer ‘International English,’ and may overlook some of the infelicities of what AI can do to the written word.

It must immediately be said that AI content is 99.99% grammatical and correctly spelled—though often in American or international English. That’s not a problem for most marketing purposes; however, it’s not so much the issues of hallucination (making up references or sources simply to please us), or its “timelessness” (serving up the global consensus thinking on subjects like networking or fibre engineering).

No; what concerns us is that there is never a through-line, story, or shaping narrative in this stuff. There is therefore no motivation for why your specific target customer or ideal buyer persona should be reading this content from you.

For example, recently (and obviously, we're not naming names of prospects or individuals) we received what, on “paper” (pun intended), looked like a completely respectable, high-quality "analyst-style" report. The machine learning engine had sourced, with great detail and accuracy, a set of data about a specific vendor in the market; it formatted and presented this as an authoritative discussion about the company.

But there was no reason for anyone to read it—no introductory framing, no guidance explaining whether Company X is a good fit for a long-term technology investment. It was simply a collection of facts and data points about publicly available investment information—some reasonably recent, some cited from the company’s activities. Here’s your problem: Is that enough for a potential enterprise customer to commit budget to your solution?

Surely, there is an interpretive layer missing here: explaining context, competitors, market position? Is this company ahead of its time, or simply in the middle of the adoption curve and thus reliable but unremarkable?

Simply getting a photocopy from the internet of its financials and recent headlines is not such analysis. It’s not helpful for a decision-maker or your target reader as a budget holder looking to take action.

What free AI generated marketing ‘content’ just can’t do

Facts alone, data alone, are not an argument. They do not form a real proposition. They cannot be a sales message. As professional communicators, stakeholders, and frequent participants in industry debates, we feel we are asking the machine to do something it simply cannot currently do.

This isn’t a discussion about AGI (artificial general intelligence) and if it is or isn’t coming, it’s about what AI is capable of at this moment. What ChatGPT-style content generation gives you is facts without context, statements without direction, and no shaping narrative.

You may well disagree. Perhaps you asked ChatGPT to write you an article yesterday about fibre optics, and what you got back was, at first glance, great. What you probably received, though, was a general-purpose, non-specific text—undifferentiated to you, your client base, your position in the market or quadrant, or whatever metric your enterprise customers use for evaluation.

It’s just content; it’s ballast. It’s not going to help you. There’s also the question—which we’ll probably return to here in our blogs—of what the internet and search engines actually look for in your content now.

If what you publish is the same as everyone else’s and is not credible, original thought leadership, the algorithms will simply ignore you. There will be no impact, nothing will change. That is a waste of your money and opportunity.

Do you want to be Potemkin or the Empress?

Our analogy for this situation is the Potemkin Village. Potemkin Villages were fake settlements allegedly constructed by Russian minister Grigory Potemkin in 1787 to impress Empress Catherine II during her tour of Crimea, creating a false appearance of prosperity and development. Since then, it’s become a term for any facade or deceptive construct designed to make a situation look better than it actually is.

Potemkin fooled people into thinking something had been constructed and real action taken. But in reality, nothing had happened. No villages were built, no peasants better housed, no economic opportunities created. If you simply rely on unedited, unfiltered, un-engineered AI content, all you’re doing is creating empty marketing and educational “villages” that are one pixel deep.

Is that what you want to be looking back on a year from now, knowing that genuine human opportunities may have passed you by? We think it’s something that should be debated in your company—and perhaps with us—right now.

Yes, we’ll try and sell you a new pair of content “shoes”. But maybe you do really need some.