The End Of The Average

A market position defined as merely competent or adequately differentiated is now extinct. The real question is not whether AI is competition, but whether your work ever truly mattered beyond mere adequacy.

The Middle Is Gone

Consider what has actually happened. AI can now generate, in seconds, what would have taken a human hours or days: blog posts, stock photography, marketing copy, basic design work, even passable music. None of it is transcendent. All of it is adequate.

AI has industrialised the text or picture that our perception accepts. With millions of articles, tweets, and YouTube videos published every day, and a fast-paced society, the outcome counts, but not how it was produced.

For decades, the content economy operated on scarcity. Producing a thousand words of coherent prose required time, a certain level of intelligence, and access to information. These barriers created value, even when the output itself was sometimes of low quality. The professional middle class of content creators—journalists, copywriters, designers, consultants—could sustain themselves not through genius, but via reliable competence.

Take your local newspaper, how much do you see that could be AI-generated? Do you trust this source of information or the tweet of an intellectual person on X?

But Wasn’t the Middle Always a Lie?

The problem is that the “middle market” was never about quality. It was about information asymmetry and friction costs.

You could charge $5,000 for a marketing strategy deck, not because your thinking was irreplaceable, but because the client didn’t know where else to look, didn’t have time to assemble it themselves, and needed someone to blame if it failed. You were selling convenience packaged as expertise.

AI didn’t destroy genuine value; it revealed the inherent instability of the average—showing that simply being "pretty good" was never a sustainable advantage.

The disruption we’re witnessing is not new. Consider what happened to travel agents when Expedia emerged. Or to stock photography when Shutterstock proliferated. Each time, the people in the middle—those who aggregated and repackaged but did not originate—were obliterated.

AI has rapidly extended this pattern: wherever average work thrived purely due to inefficiency or friction, it now collapses. The end of the average is now universal.

So perhaps the question is not “How do I compete with AI?” but rather “Was I ever doing anything that couldn’t be replicated?”

Niche or Volume as False Binary

The typical response to this collapse is to choose between two strategies: focus intensely on a narrow niche, or scale efforts to achieve high volume. Both are practical business approaches and might offer temporary advantages, but each has critical limitations.

These two approaches—niche specialization and volume scaling—are based on a common logic: they assume success comes from either being uniquely targeted to a specific group or being everywhere at once to capture broad attention.

This is where Slow Culture offers a third position: refuse the game entirely.

Defining your niche means connecting with peer groups and contacts, and starting local, even though the internet is global. Begin in your immediate community and seek feedback: Does your idea have true market potential, or is it more of a hobby?

The niche strategy says: “I will out-specialize the machine.” But specialization, taken to the utmost, becomes performance. You are not a creator; you are a curator of your own brand. You spend more time signalling your uniqueness than actually being unique.

With a volume strategy, you create lots of high-level content and maximize your presence across many points of contact. This approach requires a significant investment of time and money. For example, if you pursue a social media campaign, you must invest heavily in advertising to ensure your content or products stand out. Without high content volume and visibility, you risk being lost in the everyday noise, both online and offline.

The volume strategy says: “I will out-produce the machine.” But outdoing an algorithm in volume is impossible and exhausting.

What Cannot Be Averaged?

Here is what AI cannot do, not because of technical limitations, but because of technological impossibility:

AI cannot have lived your specific, unrepeatable life.

It can simulate the manner of lived experience and can generate text that sounds like reflection, memoir, like hard-won wisdom. But it cannot have actually endured the years of doubt, the specific failure, the texture of a particular grief, the exact quality of light on a particular morning that changed everything.

The essay, in the sense of a Slow Culture approach, that begins “I spent three years trying to save my family business before realizing I was actually trying to save myself.” cannot be generated. Not because AI lacks the words, but because it lacks the personal experience and meaning.

What are you creating that AI never could, not because of skill, but because only you have lived it?