Are We Actually Building Or Just Prompting?

We are currently witnessing the mass-industrialisation of the mean.

For most of history, producing good writing, good code, or good design required skill. The ability to think deeply and execute well was a distinguishing factor.

That logic no longer applies.

When AI makes output effortless, it removes the friction that once forced people to ask better questions. When you take away the need for critical thought, what’s produced is whatever sits closest to the average. The barrier to entry collapses, and the whole system reorganises itself around new incentives.

These models generate text by sampling what is statistically plausible in context. Left untouched, they lean toward the common, not the singular. So when you ask AI to write your code, design your UI, or draft your strategy without much manual intervention, you aren’t requesting originality; you’re requesting the statistical middle.

We are not building; just prompting. And when everyone prompts the same models with the same intent: growth, retention, speed, we all converge on the same output. We automate the middle of the bell curve. The models can produce outliers, but only with real guidance, and most people don’t apply any.

This is why everything is starting to look the same: the same SaaS dashboards, the same content, the same designs, the same ideas dressed in marginally different wrappers. The tools have given everyone the ability to create anything imaginable, yet most people replicate whatever already appears to work. This isn’t a failure of imagination; it’s simply how people react to abundance. Faced with infinite possibilities, people default to what looks proven. Imitation feels safer than originality. Copying feels easier than thinking.

Because the tools compress effort to near nil, generic output becomes economically rational. Markets reward speed and distribution over depth and craft. The first people capitalising on this shift aren’t the most skilled; they’re simply the fastest to package templates, playbooks, automations, and wrappers at scale. The product changes; the behaviour stays the same.

And as the marginal cost of building collapses, the logic becomes self-reinforcing.

Founders aren’t incentivised to solve deep problems; they’re rewarded for shipping the quickest version of whatever already exists. The ease of building removes the intellectual friction that once forced people to reason, question, and develop actual conviction.

Consumers reinforce the loop. They click what is familiar. They buy what resembles what they’ve already seen. Sameness survives because both builders and audiences operate in the same behavioural circuit: low friction, low thought, high imitation.

AI hasn’t reduced creativity; it has revealed how little of it there was to begin with. It has surfaced the average that was always there and stripped away the excuses that used to hide it.

The leverage isn’t in the tools. It’s in the judgment guiding them.

When the cost of creation collapses, the value shifts to what can’t be automated: judgment, clarity, and critical thought. Creation becomes infinite. Discernment does not. The model can generate a thousand strategies; only competence can tell which one survives contact with reality or whether the right answer is the one the model could never generate at all.

We are moving from an economy of builders to an economy of editors. But editing isn’t a softer form of creation; it is a harder one. The machine produces the centre of the distribution. The outliers will be the people who think critically, see clearly, and refuse to settle for the mean.