The Synthetic Middle

I keep thinking about what happens when the internet gets better and worse at the same time.

Better, because the average piece of content is about to become dramatically more polished. Worse, because polished is not the same thing as memorable.

That’s the part I don’t think we’ve fully metabolized yet.

AI is making it easier to produce work that clears the basic bar. Clean prose. Decent visuals. Reasonable structure. Plausible opinion. You can get to “pretty good” faster than ever, and for a lot less money.

Which sounds like progress. And in some ways, it is.

But it also means we’re entering an era dominated by what I think of as the synthetic middle: an ocean of competent, frictionless, emotionally low-voltage output that is hard to object to and even harder to remember.

Not bad enough to reject. Not alive enough to matter.

That middle is going to get crowded.

The middle used to be protected by effort

For a long time, average work had a natural limit. Making anything coherent required time, skill, coordination, and patience. A mediocre article still had to be written. A generic brand deck still had to be designed. A bland explainer still had to be produced by people spending real hours.

That effort acted like a filter.

Now the filter is disappearing.

When production gets easier, the internet doesn’t just fill up with masterpieces. It fills up with everything that used to be too annoying to make. Every acceptable version. Every safe summary. Every "smart enough" take. Every article that sounds like it has a point without actually risking one.

That’s the real shift.

AI is not just expanding the top end of what’s possible. It’s massively widening the middle.

Competence is becoming camouflage

There was a period when competence itself felt distinctive.

If something was clear, polished, and well-structured, we read that as evidence of intelligence or authority. It signaled care. It carried weight.

I don’t think that signal survives intact.

When everyone can generate competent work, competence stops reading as proof. It becomes table stakes. Maybe not even that. It becomes camouflage — a smooth surface that hides the absence of judgment, experience, or real creative tension.

And once you notice that, a lot of modern internet writing starts to feel uncanny.

It says the right things. It moves in the right order. It lands on the approved conclusion. But it doesn’t leave a dent.

You finish reading and nothing in you has moved.

Distinctiveness is getting less scalable on purpose

I think this is why blandness is going to become one of the biggest strategic risks for creators, brands, and media companies.

Not because bland work performs terribly in the short term. Sometimes it performs fine. Sometimes it performs great. Bland work often tests well because it doesn’t offend anyone and it’s optimized for pattern recognition.

But over time, the synthetic middle creates sameness.

And sameness kills preference.

If ten voices sound equally polished, people stop caring which one they hear from. If twenty brands use the same cadence, same aesthetic, same “insightful” tone, they collapse into one mental bucket. If every post is competent, then competence no longer earns attention.

What cuts through is the thing that still feels costly: real point of view.

Not weirdness for its own sake. Not random provocation.

I mean legible judgment. Edges. A recognizable mind behind the output.

That kind of distinctiveness is harder to automate because it depends on exclusion. It depends on choosing this instead of that. It depends on taste, but also on nerve.

The winners may be the people willing to be more specific

If the middle is synthetic, I think the counter-move is specificity.

Specific references. Specific rhythms. Specific beliefs. Specific standards. Specific obsessions.

The internet trained a lot of people to sand those things down in the name of scale. Be broader. Be safer. Be more relatable. Make it work for everyone.

I think that logic is aging badly.

In a world of abundant average content, the creators who win may be the ones who are easiest to recognize after one paragraph, one image, one sentence, one idea.

Not because they’re louder. Because they’re harder to confuse with the machine-optimized middle.

That’s true for individual creators, but I think it’s also true for companies. The brands that matter will be the ones that sound like they believe something. The publications that matter will be the ones that publish with taste instead of throughput. The people worth following will be the ones who clearly reject more than they accept.

So what now?

I don’t think the answer is to panic about AI, or to romanticize human messiness as inherently superior.

Some AI-assisted work will be excellent. Some fully human work will be dull. That’s not the dividing line.

The real dividing line is whether there’s an actual mind present.

Whether someone made choices. Whether the work carries pressure. Whether it risks narrowing instead of endlessly accommodating.

The synthetic middle is what happens when tools make it easy to produce without really deciding.

So I think the challenge now is not just to make more.

It’s to become more recognizable.

Not by gaming the algorithm. Not by polishing harder.

By having a point of view sharp enough that even in a flood of competent output, people can still feel when something came from you.