Why AI Writing Starts to Feel the Same

Why unfinished language may become more valuable in the age of AI

AI is becoming extraordinarily good at making language clear, smooth, and optimized.

But humans have long lived alongside things that never fully resolve.

Perhaps what lasts is not perfectly completed meaning,
but meaning that leaves space for us to enter.

In the age of AI,
slightly unfinished language may become more valuable than ever.

Recently, I’ve been seeing more stories about YouTubers losing monetization because their content was judged as “AI-generated.”

In some cases, the videos were actually mass-produced with AI.
In others, creators claimed they made everything manually, but were still flagged for being “too AI-like.”

What interested me was not only the false detections themselves, but the feeling that platforms are beginning to fear something deeper:

convergence.

AI is extraordinarily good at optimization.

It makes writing smoother.
More understandable.
More emotionally readable.
Less ambiguous.

The more we ask AI to make something “clear,” “natural,” or “engaging,” the more it guides language toward points that are easy for many people to understand.

And honestly, that is incredibly useful.

Writing today is often easier to read than before.

But at the same time, another feeling has started to appear.

Something can be perfectly readable,
yet strangely difficult to remember.

You finish reading,
and the meaning closes too completely.

Perhaps the problem is not simply that people are using AI.

Perhaps we are beginning to notice how strongly humans themselves are drawn toward convergence.

Toward language that is safe.
Smooth.
Understandable.
Easy to agree with.

AI simply accelerates that movement.

And now we are entering a strange moment:

society encourages AI,
while simultaneously becoming uncomfortable with things that feel “too AI-like.”

What platforms may actually fear is not AI itself,
but a world where too much content converges toward the same emotional and structural center.

The pacing becomes similar.
The tone becomes similar.
The emotional curves become similar.

Eventually, the sense of “who made this” begins to disappear.

But humans have never loved only perfectly completed things.

Sometimes we stay longer with things that remain slightly unfinished.

A damaged statue.
An abstract painting.
A story that never fully resolves.

If a missing arm on a sculpture had been perfectly restored,
perhaps it would simply become a completed sculpture again.

But because something is absent,
people begin to participate.

They imagine what was there.
What movement once existed.
What meaning still remains unfinished.

Something slightly incomplete invites human participation.

And perhaps participation is what creates lasting attachment.

Love may work similarly.

Not everything is fully understood.
Not everything becomes explainable.
Some feelings continue moving inside us for years.

Human beings have always lived alongside things that never fully closed.

But AI naturally tends toward resolution.

It organizes.
Clarifies.
Completes.
It reduces ambiguity.

Again, that is not inherently bad.
In many situations, it is deeply valuable.

But if everything becomes perfectly resolved,
we may slowly lose spaces where people can participate emotionally and cognitively.

Recently, I’ve started writing some of my philosophical pieces in English before publishing them in Japanese.

Not because my English is perfect.

Actually, the opposite may be true.

My English is slightly unfinished.

And lately, I’ve started wondering if that incompleteness itself matters.

Japanese, as my native language, carries extremely high resolution for me.
Nuances become fixed very quickly.
Meanings settle too completely.

But when I pass ideas through English — especially as a kind of intermediate language — something changes.

The meaning drifts slightly.

It becomes harder to close completely.

And because it remains slightly unresolved,
the reader may begin participating in the meaning itself.

This is not simply “loss in translation.”

It may be a form of resolution control.

Perhaps this is why I’ve become interested in writing through an intermediate language.

Not to make meaning more accurate,
but to keep it from settling too completely.

Perhaps I no longer want to deliver ideas as photographs.

Perhaps I want to deliver them more like abstract paintings.

In the AI era,
perfect translation and perfect summarization will become increasingly possible.

Precisely because of that,
slightly unfinished language may become more valuable.

This is not an argument against AI.

AI will continue to shape how we write, think, and communicate.

But perhaps the important question is no longer only:

“How clearly can we communicate?”

Perhaps it is also:

“How much unfinished space can we leave for others to enter?”

And perhaps intermediate language —
language that never fully settles —
may become one way to use AI
without sounding entirely like AI.

Related to this theme, I previously wrote:

“What If the Most Important Human Ability Is Staying Unresolved?”

about how humans have long lived alongside things that never fully resolve.

© 2026 in the pilvi / Kyokuya Project
Please credit the source when quoting or referencing this article.