The war on differentiation

Collective intelligence is extraordinarily good at one thing. It finds the average with brutal efficiency. What it does not find, at least not reliably, is the extraordinary. And yet we are now handing it the keys to our brands, our language, our culture and our decisions at scale.

 

In 1906, a man with deeply troubling beliefs accidentally revealed something uncomfortably true about modern times. Sir Francis Galton believed people could be ranked by intelligence, by worth, by destiny itself. Convinced that most humans were fundamentally incompetent, he went looking for proof.

 

He found his test case at a county fair in Plymouth, England. The challenge was simple: guess the weight of an ox. Seven hundred and eighty-seven people entered: farmers, butchers, shopkeepers, curious onlookers. People with experience. People with none. The guesses ranged wildly.

 

Galton collected the slips, expecting to demonstrate collective failure. Instead, the average of all 787 guesses landed within one pound of the ox’s true weight.

 

One pound.

 

The crowd wasn’t wise in any philosophical sense. But it was right. Not because anyone understood the whole, but because mistake cancelled mistake. Noise collapsed into accuracy. That mechanism, aggregation without understanding, now quietly runs much of the modern world.

The machine-made middle

 

Optimisation engines and large language models increasingly pull everything toward a statistically “correct” centre. Language, marketing, pricing, product design, strategy. Even humour. The middle is becoming machine-perfect.

 

And that should concern every leader, every marketer, every CEO and CMO, whose growth and success ultimately depends on differentiation.

 

The real risk is not that these systems will make mistakes. The risk is that they will make the same decisions, in the same ways, with the same logic, at massive scale. The same phrasing. The same tone. The same narrative structures. The same emotional arcs, delivered through different logos. We are not moving toward artificial intelligence. We are moving toward artificial averageness.

 

You can already see the convergence. SaaS websites follow identical layouts promising identical efficiencies. Consumer brands rely on near-indistinguishable aesthetics and storytelling frameworks. Corporate mission statements read as if they could belong to any company in any industry.

 

Individually, each example seems harmless. Together, they signal a market drifting toward sameness.

 

That’s not branding. That’s statistical camouflage.

 

The limits of collective intelligence

 

Some argue that collective intelligence will save us from this flattening. But what Galton observed, and what our modern systems now replicate at scale, is not wisdom but the gravitational pull of the centre.

 

Collective intelligence excels at finding the average. It is far less reliable at producing the extraordinary. And the middle is exactly where differentiation disappears.

 

More optimisation will not fix this. Faster consensus will not fix this. Outsourcing judgment to probabilistic systems trained on past patterns will not fix this.

 

Efficiency compresses variance. Progress depends on it.

 

The case for human friction

 

Advancement comes from disagreement. From competing interpretations. From people willing to slow momentum in order to challenge what everyone else is quietly accepting.

 

This work is uncomfortable. Collaboration feels inefficient. Debate feels messy. Tension feels risky.

 

But efficiency is not the same as progress. And friction is often the price of clarity.

 

If we surrender decision-making and creative conviction to systems designed to converge on the statistical middle, we should not expect originality. We will get polished sameness at precisely the moment the world demands sharper edges.

 

Leaders cannot allow their organisations to drift toward the safest version of themselves. The future rewards those who defend the edge, nurture unconventional thinking and champion ideas that don’t yet fit.

 

We do not need smoother outputs; we need bolder ones. Not average; not optimised – unmistakable.

 

Where we stand at SH/FT

 

This belief defines how we lead at SH/FT and how we partner with clients.

 

We do not use AI to generate more content or to develop systems that blend into the market. We use it beneath the surface, inside operations, decision systems and execution models, where it sharpens performance without flattening identity.

 

Our work focuses on uncovering what is already extraordinary inside an organisation and building intelligent systems that amplify it with speed, discipline and measurable impact. The goal is not to automate sameness. It is to protect and scale what makes a company impossible to confuse with anyone else.

 

For leaders who want to go deeper, we’ve outlined this philosophy and operating model in a short paper outlining the impact of AI below the stack. It details how organisations can unlock real AI value without surrendering strategic distinctiveness.

---

The goal is not to move faster in the middle. The goal is to stand, unmistakably, apart.

---

Learn more here