Run that number forward. If you line up four articles in your industry newsletter or LinkedIn feed right now, three of them were shaped, at least in part, by the same handful of language models trained on largely the same internet. That’s not a content strategy problem. It’s a sameness problem, and sameness is exactly what a flooded market can’t sell.

Why agreement stopped being proof of anything

There’s a habit worth breaking here, and it isn’t unique to machines. Ask three different AI models the same strategic question and they’ll often converge on the same answer. It can look, for a moment, like confirmation. It isn’t. Three models trained on overlapping data arriving at the same conclusion is not three independent minds checking each other’s work — it’s one set of assumptions echoed three times.

Humans do this too. Ask five people in the same industry the same question and they’ll often converge, not because anyone checked independently, but because they read the same three books last year, or sat through the same conference circuit. Consensus has always been a weaker signal than it sounds. AI just made the pattern easier to see, and faster to produce at scale.

This matters directly for anyone building a personal brand or a content platform. If your content agrees with the consensus, it’s forgettable by design, no matter how well it’s written. The value was never in publishing an answer. It’s in noticing something before the consensus caught up to it.

What the market is actually paying for now

This isn’t a philosophical preference. It shows up in how B2B buyers behave, and the numbers are specific enough to build a strategy around.

Separately, Momentum ITSMA’s Value of Thought Leadership 2025 study of 600 senior decision-makers found 99% now consider thought leadership important or critical to how they make decisions, and 66% said they won’t work with a provider whose thought leadership is weak. That’s not a marketing nicety. That’s a disqualifying filter, sitting upstream of every RFP and sales call you’ll never get invited to.

Put the two findings together and the shift is clear. Competence — the ability to produce fluent, correct, well-formatted content — is now available to almost anyone for almost nothing. What can’t be bought anywhere is a person who already sees the pattern the rest of the market is about to hit. A polished quarterly report or a well-optimised blog post used to signal that a company knew what it was doing. It doesn’t anymore, because AI can produce something that reads exactly as confident whether or not it’s actually right.

Content creation and thought leadership are no longer the same job

The flood has forced these two things apart, whether or not any individual brand chose that split.

• Content creation says: here’s something useful. AI is genuinely good at this, and I use it for exactly that most days — briefs, outlines, first drafts, research summaries.
• Thought leadership says: here’s something I noticed, and it might change how you see this. That still requires a person who spots a pattern before anyone else confirms it, says so under their own name, and is willing to be wrong about it in public.

That distinction is exactly why rewriting old, generic posts into research-backed, opinionated pieces tends to outperform simply publishing more of them. Volume was a moat when volume was expensive. It stopped being one the day every competitor got the same tool you did.

A rented opinion collapses the moment someone asks a real follow-up question. A tested one gets stronger.

What a machine still can’t manufacture for you

Photography didn’t end painting. It ended the job of painting things that simply looked like the thing. AI is doing the same to expertise, only faster. It can write the report, build the deck, summarise a decade of research, and do all of it competently before lunch.

So the useful question is no longer “can you produce the work.” It’s “can you tell which of six fluent, well-written options is actually right for this client, this moment, this level of risk” — and that’s a completely different skill. It looks like judgment. It looks like taste. It looks like a reputation built from the specific, uncopyable shape of the decisions you’ve already made and been right or wrong about, in public, under your own name.

None of that shows up as a bullet point on a services page. It shows up as the pitch you didn’t take, the trend you didn’t chase, the draft you deleted because it was clever but not true. This is the same territory covered in how creators are monetising trust rather than reach — the audience that pays is the one that already believes the person behind the words has actually done the thinking.

A simple loop, run in public

The practical version of all this isn’t a content calendar. It’s a loop: observe, think, test, publish, listen, refine, repeat.

1. Notice something specific, from client work, a live project, or a thread you can’t leave alone.
2. Ask what it might actually mean before you’re fully sure.
3. Hand it to AI and make it argue the opposing case — let it try to find the evidence that proves you wrong.
4. Publish it while it’s still a little rough, rather than waiting for a polished, forgettable version.
5. Watch what people push back on. That reaction is data, often better than any survey.
6. Sharpen the idea and run the loop again.

A funnel assumes you already know the answer and are just moving it downstream. A loop assumes you don’t know yet, which is the more honest place to actually start from — and it’s also why smaller, specific audiences with real trust in you consistently outperform large ones that only half-believe you. This is the same logic behind why generic SEO content is losing ground in an AI Overviews search landscape: search engines are getting better at summarising the competent middle, which leaves more room, not less, for the point of view that can’t be summarised away.

What this is worth, practically

None of the revenue this creates starts with more content. It starts with being known for something specific: a consulting practice, a speaking calendar, a course, a membership, a piece of software, a side project that quietly becomes the actual company. AI gives one person more reach than a small agency had a decade ago. Reach on its own is just noise at scale. Reach paired with a real, tested point of view is a business model — one honest opinion, a machine that executes at speed, and an audience that already trusts you enough to keep opening what you send them.

So the smaller, harder question isn’t “how do I publish more.” It’s: what do you already know, or notice, or believe, that’s worth becoming known for?