By Jane Sonnenschein · March 15, 2026

Tonight, while scrolling through Xiaohongshu (a Chinese lifestyle and social-sharing platform often compared to a mix of Instagram and Pinterest), I came across an interesting post titled “Even AI Thought My Writing Was AI-Generated.”

The person who posted it seemed to be a student. She had written an essay herself, then asked Doubao and DeepSeek, two Chinese AI assistants, to each write another essay imitating her style. She put all three pieces together and asked readers to guess which one was hers and which two had been written by AI.

It was the kind of question that naturally makes you curious enough to click. What made it even more interesting was that people in the comments kept guessing, and even AI identified the essay she had written herself as AI-generated. She seemed a little confused by the whole thing too.

At first, I clicked simply because I was curious. But the longer I read, the more seriously I began to think about it. Those three essays touched on something that had already been on my mind for the past couple of days.

How should we write about emotions?

Or, to put it more precisely, why do some articles keep telling you how miserable or heartbroken the writer feels, yet leave you feeling very little? And why can another piece of writing draw you in almost immediately, without ever directly mentioning sadness?

I went back and read the three essays several times.

The first was fairly simple and easy to read, although some of its metaphors felt a little forced to me. I could barely get through the second one at first. It actually gave me a headache. Later, I made myself go back and finish it, and began to think that perhaps this one was more likely to have been written by a real person. There was something about its repetition and unevenness that felt less controlled.

I read the third one twice. It flowed more smoothly and felt more coherent, like something written by someone who already knew how to write well.

But what really made me stop wasn’t the question of which essay sounded like AI.

It was the sudden realization that, deep down, I had always known what kind of writing I believed in.

I’ve never been particularly convinced by writing that states emotions too directly.

That doesn’t mean someone shouldn’t write, “I’m sad,” “I’m hurting,” or “My heart is broken.” Of course they can. But when an entire article keeps telling you just how much the writer is suffering, that suffering can sometimes begin to feel strangely unconvincing.

Not because the emotion itself is necessarily false, but because once it is reduced to a label, it easily becomes a conclusion. You understand what the writer is telling you, but you haven’t really entered the moment.

I’ve always believed more in another way of writing.

I don’t write about sadness directly.

I write about a room so quiet that I can almost hear my own breathing, with only the clock on the wall ticking away.

I write about sadness slowly seeping in, like water.

I write about a child walking to kindergarten, brushing her hand against the fence beside a field, until one day a ring that was far too big for her finger slips off and falls into the rice field.

I write about a mother who is so frugal that she hardly ever allows herself to eat meat, yet always puts some into my bowl.

I write about a little toy airplane that could fly in circles from the ceiling, a toy gun that lit up, a plant in the corner of a room, and all those seemingly unimportant things attached to the refrigerator.

Most of the time, I don’t begin by deciding what a particular scene is supposed to express. These details simply appear as I write.

Only when I look back at my work later do I gradually understand that this is how I have always written about emotions.

For a long time, I had never really thought about it.

People’s comments on my writing were often rather vague. They would say I wrote well, but there was usually something else behind that compliment. Some said my writing was difficult to understand. Others thought it felt a little empty.

My father used to read things I had written and tell me quite directly, “Your writing has no soul.”

What he meant was actually quite simple: he couldn’t understand it.

He felt that I had written a lot, and that the sentences might sound beautiful, but he couldn’t grasp what held everything together. He couldn’t find the central idea or understand what I was really trying to say. To him, that meant the writing had no soul.

For a long time, I more or less believed that judgment myself.

I thought perhaps I was good at putting sentences together, using metaphors, and creating an atmosphere, but not necessarily good at writing something meaningful.

So for years, I tried to turn myself into a different kind of writer.

I wanted to read more literary classics, learn how to tell stories, and make my work feel more substantial, more like the kind of writing people generally recognize as genuinely good.

It wasn’t until recently, when I started rereading some of my old work, that I suddenly realized something.

Perhaps I had never been unable to write. Perhaps I had simply failed to recognize my own way of writing.

What I’ve always been good at is not stating things directly.

I don’t naturally begin by telling readers, “I’m terribly sad,” “My father and I have a close relationship,” or “My feelings toward my mother are complicated.”

Instead, I tend to place emotions inside scenes, actions, and small objects that don’t seem, at first glance, to be expressing any particular feeling.

Take something I wrote about my father when I was young.

I didn’t directly say that my father loved me very much, or that there was something complicated about our relationship.

I wrote about a time when he came home from a business trip and brought me a toy gun and a little airplane that could hang from the ceiling and fly around in circles. He also bought me a ring that was obviously unsuitable for a child. It was the sort of ring an elderly person might wear, and it was much too big for my finger.

Later, on my way to kindergarten, I would often run my hand along the fence beside the farmland. One day, the ring slipped off and fell into the rice field.

None of these details explains anything directly.

Yet when they are placed together, so much is already there: my father’s somewhat awkward way of showing love, the atmosphere at home shaped by my parents’ arguments, a child’s attachment to a gift from her father, and that strange, empty feeling when something precious suddenly disappears.

Then there is my mother.

If I write about her, I don’t usually begin by saying that our relationship is both deep and complicated, or that there is a certain helplessness in the love between us.

Instead, I write about how careful she always was with money. She was reluctant to eat meat herself, but whenever we sat down for a meal, she would put some into my bowl.

That one small action already carries so much: love, guilt, restraint, tenderness, and the ache of knowing what someone has quietly given up for you.

Over time, I’ve come to understand why I trust this kind of writing more.

It isn’t the writing that repeatedly tells me how much someone is suffering, how deeply they love, or how unfairly they’ve been treated. It’s the writing that doesn’t need to name those feelings because I can already sense them in the scene.

This is something I’ve gradually come to care about more and more: feeling something matters more than simply reading about it.

When someone writes, “I’m sad,” the reader receives a piece of information.

But when someone writes about a room so quiet that all you can hear is the steady ticking of a clock, the reader begins to experience a kind of sadness.

To me, there’s a real difference between the two.

One tells you the conclusion. The other lets you enter the moment.

And I’ve always believed more in the second approach.

Because what makes an emotion convincing isn’t someone telling me how much they hurt. It’s being able to recognize that hurt in the silence of a room, the ticking of a clock, the sound of someone breathing, or a ring disappearing into a rice field, even though the writer has never directly said how painful it was.

This is also why I’ve become increasingly uncomfortable with writing-advice posts that rely too heavily on techniques.

Today, I came across another post with more than three thousand likes. Its title was something along the lines of “An Amazing Technique for Writing Emotions.”

When I opened it, the format was exactly what I had expected.

Don’t write “I’m nervous”; describe someone’s Adam’s apple moving as they swallow. Don’t write “I’m sad”; describe some particular physical reaction or detail instead. Then came Technique One, Technique Two, Technique Three, followed by a neat list of practical tips.

I can see how this might help someone who is just beginning to write. It’s like a PowerPoint presentation or a checklist, a ready-made method that you can immediately apply.

But that isn’t really how I understand writing.

To me, writing naturally doesn’t mean deciding in advance that you want to express sadness through scenery, then deliberately inventing a ticking clock, a quiet room, a movement of someone’s throat, or some carefully chosen object.

Natural writing happens when the feeling is already there, and the surrounding details begin to appear alongside it.

Those objects already belong to your life. They are things you have seen, lived with, or carried in your memory. When you write, they emerge together with the emotion, and that is what gives them their sense of truth.

If the process works the other way around, with the technique coming first and the emotion being added afterward, the result can easily feel forced.

The metaphors feel forced, the scenery feels forced, and even the sadness can begin to resemble something arranged for a photograph.

So by the time I finished reading those three essays tonight, I was no longer particularly interested in deciding which one sounded like AI and which one didn’t.

What I really wanted to understand was why I react so strongly to certain kinds of writing.

And perhaps I finally began to see the answer.

It’s because this is how I’ve always written myself.

When I write about sadness, I don’t directly describe sadness. When I write about love, I don’t necessarily name love either.

For the feelings that are most complicated, hardest to explain, and most likely to lose something when they are put too plainly into words, I’ve never relied primarily on conclusions.

I leave behind scenes, gestures, and a little empty space.

That empty space matters.

Because once it is there, readers can enter it themselves. One person may feel sadness, another love. Someone else may sense helplessness or loss.

Everyone brings their own experiences into what they read, so the feelings they take away won’t necessarily be the same.

I don’t see that as a problem. On the contrary, I think it’s one of the most valuable things about writing.

If I simply tell someone, “My father and I had a very close relationship,” they can only understand what that sentence means.

But if I write about the little airplane hanging from the ceiling, the toy gun that lit up, and the ring that fell into the rice field, readers have room to sense something for themselves.

They may begin to realize that there is more beneath those details than the words directly explain.

And that feeling can reach much deeper than a conclusion.

This is why I increasingly believe that the most important thing when writing about emotions is not to announce the feeling, but to bring the moment itself to life.

Rather than explaining everything on the reader’s behalf, I want to leave enough space for the reader to find a way in.

It’s not that I don’t understand those writing techniques.

I’ve simply become more convinced that what truly moves people isn’t the technique itself, but whether it has grown out of a real emotion.

Looking back at the Xiaohongshu post I saw tonight, “Even AI Thought My Writing Was AI-Generated,” I realize that the most interesting part, for me, wasn’t the question of which essay sounded like AI.

It was that the whole experiment made me turn around and look at myself again.

After all these years, I still believe most strongly in writing that doesn’t rush to define its own feelings. Perhaps I’ve always written this way. Perhaps what I truly wanted to express was never simply the words “I’m hurting.”

I write about the ticking of a clock, a quiet room, and a ring lost in a rice field.

And everything I never said directly has been there all along.

This essay is also available in other languages:

Chinese version: 我不直接写悲伤，我写钟声、房间和一枚掉进稻田里的戒指

German version: Ich schreibe nicht direkt über Traurigkeit. Ich schreibe über das Ticken einer Uhr, ein Zimmer und einen Ring, der in ein Reisfeld fiel.