I’ve Been Thinking About AI, SEO and Why Most Content Is Becoming Worthless

It’s been a while.
Not because I’ve run out of things to say. those who know me know i never run out of things to say. The problem is there’s so much being said that finding anything worth listening to has become increasingly difficult.

Over the last few months I’ve watched an endless parade of AI experts, SEO gurus and LinkedIn philosophers confidently explain the future of search. Apparently all I needed was a subscription to someone’s prompt library, a 97-page PDF and a framework with an acronym that sounds like a government department.

Funny enough my bullshit radar was going into overdrive.
So instead of jumping on every bandwagon, I’ve been doing something considerably less glamorous. Reading. Testing. Breaking things. Changing my mind. Trying to work out what actually survives when everyone has access to the same AI tools.

Because here’s the uncomfortable bit.

The Internet Doesn’t Need Another Average Article

If ChatGPT, Gemini or Claude can produce something that looks remarkably similar to your article in thirty seconds, what exactly have you contributed?
That’s the question I’ve been wrestling with.

I don’t think I’ve found the answer. Anyone claiming they have is probably trying to sell you something.
But I do think I’ve found a principle that’s becoming increasingly difficult to ignore and it seems so obvious its kind of stupid i have not thought about it before.

The businesses that will win won’t be the ones producing the most content. They’ll be the ones producing the most original knowledge.

One of the biggest shifts AI has created isn’t that content is easier to produce. That’s obvious. The more interesting shift is that average content has become almost worthless.

---

Information Gain Isn’t an SEO Tactic

A few years ago, writing a well researched article gave you an edge. Today, that same article can be generated in seconds. It might not be perfect, but it’ll be coherent, well structured and probably cover the same points as the first twenty results in Google. The uncomfortable truth is that if your article can be recreated by someone with access to a llm, it probably wasn’t contributing much in the first place.
I don’t say that to dismiss AI. I use it every day. It has genuinely changed how I work. But somewhere along the way we’ve started confusing the ability to generate content with the ability to generate knowledge, and they’re two very different things.

An AI model can explain almost anything that’s already been explained. It can compare ideas, summarise research and connect concepts in ways that would have taken hours only a few years ago. What it can’t do is tell the world something that wasn’t already there.

It doesn’t know what your customers argued about on yesterday’s sales calls. It doesn’t know why eighty per cent of schools chose one planner layout over another. It doesn’t know which packaging finish causes the most production delays or what patterns you’ve noticed after quoting hundreds of projects. None of that exists in its training data because it only came into existence when your business experienced it.

That’s the distinction I think a lot of people are missing. AI is exceptional at reorganising knowledge. Businesses still have a monopoly on creating it.

Eventually you’ll come across the phrase information gain. It’s become one of those terms the SEO industry loves to overanalyse. There are endless discussions about patents, algorithms and whether it’s a direct ranking signal. If I’m honest, I think a lot of people are missing the point.

Forget Google for a minute.

Imagine you’ve just finished reading an article. Close the tab and ask yourself one simple question.

Do I know anything now that I didn’t know ten minutes ago?

If the answer is no, then what exactly was the point?

That’s information gain.

Not because Google has a patent for it, but because that’s how people have always judged whether something was worth reading. We don’t remember articles because they were well optimised. We remember them because they taught us something, challenged something we believed or showed us evidence we’d never seen before.

The problem is that most businesses are trying to create better content when they should be creating better information.

Those aren’t the same thing.

Better content is often just better presentation. Better writing. Better graphics. Better structure. All worthwhile improvements, but none of them change the underlying knowledge.

Better information is different. It leaves the reader with something they couldn’t have discovered simply by reading the ten articles that came before yours.

That’s becoming incredibly difficult for AI to replicate, because AI doesn’t have a sales team, a customer service department or twenty years of conversations with real clients. It doesn’t run surveys, manufacture products or sit in meetings listening to customers explain why they bought one solution instead of another.

Your business does.

Most businesses treat those conversations as operational noise. I think they’re one of the most valuable content assets they’ll ever own.

Your Business Already Knows More Than the Internet

One of the things I’ve realised is that most businesses don’t have a content problem. They have a curiosity problem.

Every day they’re surrounded by information that never makes it beyond a Teams call, a CRM note or a conversation in the warehouse. A salesperson notices that customers keep asking the same question before buying. Customer support spots a pattern in returns. Someone in production works out why one material consistently performs better than another. Marketing hears the same objection for the fiftieth time.

The conversation ends and everyone gets on with their day.

Six months later the marketing team asks ChatGPT what they should write about next.

That strikes me as completely backwards.

The answers businesses are looking for rarely live inside an AI model. More often than not they’re sitting inside the business itself, quietly accumulating every single day.

I’ve seen this with clients over and over again. A sales team will casually mention something they’ve observed for years, completely unaware that nobody outside the company knows it. To them it’s common knowledge. To everyone else it’s genuinely useful.

That’s the trap. We stop seeing our own knowledge because we’re surrounded by it every day.

The internet doesn’t need another article explaining what a school planner is. It might benefit from understanding which planner pages teachers quietly stop using after the first term, or why students engage with one layout but ignore another. That information isn’t sitting in a training dataset waiting to be summarised. It only exists because someone bothered to ask the question.

The same is true whether you’re manufacturing packaging, selling software or fitting kitchens. Every business has access to knowledge that the rest of the world doesn’t. The companies that recognise it will have an advantage that’s becoming increasingly difficult to copy.

---

Stop Publishing Content. Start Publishing Evidence.

This is where I think most content strategies go wrong.

The question isn’t, “What should we write about next?”

It’s, “What do we know that nobody else does?”

Those are very different conversations.

The first usually ends with someone opening ChatGPT, searching Google or looking at what competitors have written. Before long, everyone arrives at roughly the same conclusion and publishes another version of an article that already exists a hundred times over.

The second forces you to look inwards.

• What questions does your sales team answer every single week?
• What patterns have emerged after thousands of customer enquiries?
• What assumptions have you proven to be wrong?
• What surprised you when you looked at your analytics?
• What did your last customer survey reveal that you weren’t expecting?

Those aren’t content ideas.

They’re evidence.

That’s the difference.

An opinion can be copied. A survey can’t.

A generic article can be rewritten. Five years of customer conversations can’t.

Anyone can tell me that a particular product is popular. I’m far more interested in why it’s popular, what customers nearly bought instead and what eventually changed their minds. That’s information with weight behind it.

The businesses that understand this won’t just create better content. They’ll slowly build a library of proprietary knowledge that nobody else can replicate because nobody else has lived the same experiences, spoken to the same customers or collected the same data.

Ironically, I think AI makes this more valuable, not less.

The easier it becomes to generate competent content, the more valuable original evidence becomes.

Because that’s the one thing AI can’t simply conjure out of thin air.

Could Your Competitor Write This Tomorrow?

It’s a question I’ve started asking whenever I’m reviewing content.

• Not whether it’s well written.
• Not whether it’s comprehensive.
• Not whether it ticks every SEO box.

Simply this.

Could a competitor produce something almost identical tomorrow?

If the answer is yes, there’s a good chance you’ve created content rather than an asset.

That’s a difficult distinction because we’re conditioned to think that publishing is the goal. It’s not. Publishing is just the mechanism.

The goal is to create something that becomes increasingly difficult to replace.

Think about the articles you actually remember reading. They probably weren’t memorable because they were 3,000 words long or because they had a clever title. They were memorable because they contained a statistic you’d never seen, a perspective you’d never considered or a piece of research that changed how you thought about a problem.

That’s what made them valuable.

Now compare that with the average AI-generated article. It might be factually correct. It might even be enjoyable to read. But once you’ve finished it, how much have you actually learned?

That’s the standard I’ve started applying to my own writing.

If everything I’ve written could have been produced from publicly available information and a decent prompt, I haven’t added much to the conversation.

That doesn’t mean every article needs a nationwide survey or months of research. Sometimes information gain comes from analysing your own Search Console data. Sometimes it’s interviewing five customers. Sometimes it’s sharing a mistake your business made and what you learned from it.

The scale isn’t what matters.

The originality is.

---

The Question I’m Asking Myself

I said at the beginning that I don’t think I’ve found the answer.

I still don’t.

The pace at which AI is improving means anything I write today will probably need rewriting in a year’s time. That’s just the reality of the world we’re working in.

What I am becoming increasingly convinced of is this.

The businesses that thrive won’t necessarily be the ones with the biggest marketing teams or the largest content budgets. They’ll be the ones that become obsessed with understanding their customers better than anyone else.

Because every conversation, every survey, every sales call, every support ticket and every experiment creates something AI doesn’t have.

Knowledge.

Not recycled knowledge.

Not summarised knowledge.

New knowledge.

Maybe that’s what information gain has always been about. Not gaming an algorithm or chasing another ranking signal, but contributing something to the internet that genuinely wasn’t there before.

Ironically, I think AI is pushing us back towards something we’ve always known.

The most valuable content has never been the easiest to produce.

It’s been the hardest to fake.

So the next time you’re about to ask ChatGPT what you should write, try asking yourself something else instead.

What does my business know that the internet doesn’t?

If you can answer that honestly, you’ve probably found your next article.

Start Asking Better Questions

I don’t think the biggest change businesses need to make is investing in better AI tools.

I think they need to become more curious.

Every business has blind spots. Not because the information doesn’t exist, but because nobody has thought to ask the question.

Instead of asking your marketing team what content to write next, ask your sales team what objections they hear most often.

Instead of asking AI what customers care about, ask your customers what nearly stopped them buying.

Instead of asking Google what people are searching for, ask your support team what people still don’t understand after they’ve become customers.

Those conversations are where information gain starts.

The irony is that most businesses already have access to the raw material. They just don’t see it because it’s buried in day-to-day operations.

I think that’s going to become one of the defining characteristics of good marketing over the next few years. Less time asking AI for answers and more time asking people better questions.

Because every genuinely interesting article starts with curiosity.

Not with a prompt.