I.

At the École Normale de Musique de Paris, I spent two years. In the practice rooms of those two years, nearly every day I was repeating the same thing — long tones, from soft to strong, then back to soft. Catherine Cantin, my teacher and the principal flutist of the Paris Opera, would sit in the corner and listen.

I didn't think about what it meant at the time. I simply did.

Only later did I understand: it wasn't about the flute. The practice was teaching me something far more difficult — how to keep going through a process where the results are invisible.

II.

In the spring of 2026, I used an open-source AI model to generate a flute solo in ten seconds. The tone was clean. The fingering was accurate. The breathing landed in exactly the right places. If you placed it on an album, most listeners couldn't tell.

I didn't panic in that moment. What I felt was a strange clarity.

Because I suddenly understood something — the value of classical music has never been about "beauty" itself.

If beauty were the endpoint, AI has already won. It doesn't get tired. It doesn't get nervous. It doesn't drift sharp because it slept badly the night before. It is precise every time.

But the reason classical music is still alive is not its precision. It survives because it is proof that a human being can keep company with uncertainty over a very long span of time.

III.

I began studying flute seriously at the age of nine. From China to the Yong Siew Toh Conservatory at the National University of Singapore, to the École Normale de Musique de Paris. Twenty years.

What I have done across these twenty years is simply this: keep reopening the same questions, in more and more refined forms.

Why does the attack on this note come out hollow? Why does my breath always break in the wrong place in this phrase? Why can I do in the practice room what disappears the moment I step on stage?

These questions don't have answers. Or rather, the answers keep changing. What gets solved today returns tomorrow in another form.

The rocket scientist Ozan Varol wrote a sentence I have quoted many times since: "Progress happens in the dark room." I quote it because it describes exactly what a practice room feels like.

Studying classical music is, fundamentally, training in a single capacity: to keep investing in something when there is no external feedback, no immediate reward, no certain outcome.

This is becoming, year by year, a rare ability.

IV.

Let me put it another way.

Over the past decade, I have watched a generation grow up inside the algorithm. They are accustomed to every action producing feedback — a video posted, traffic measured within ten minutes; a message sent, the blue tick confirming it was read; every effort answered with a quantifiable return.

There is nothing wrong with this. It is the grammar of our time.

But certain things — their value lies precisely in the absence of a feedback system.

Practicing one long tone won't make you visibly better. Finishing a serious book won't make you visibly smarter. Walking alongside someone as they grow won't immediately produce a relationship.

These things share a single structure — they require you to keep believing in what you're doing, even when no return is visible.

Classical music is one of the oldest training grounds for this structure.

V.

So when parents ask me, "What's the point of my child studying flute?" my answer now is different from what it was five years ago.

Five years ago I would say: music develops the brain, trains focus, refines aesthetic sensibility. All of which is true. But this reduces something profound into a utilitarian explanation.

Now I say —

In an age when AI can do nearly everything for you, what your child really needs to learn is how to live with something that does not produce immediate results.

Classical music happens to be exactly this kind of training. It is not the only one — calligraphy, Go, long-distance running, serious writing all qualify. But it is among the few fields that still possess an intact lineage, living masters, and public standards of measurement (ABRSM, conservatories, international competitions).

Whether your child studies flute is not the point. The point is — whether, at some stage of their life, they passed through something that demanded long investment without immediate verification.

If they did, they will know something AI cannot teach them: To keep walking through uncertainty is itself a capacity.

And this capacity, over the next thirty years, will only become more valuable.

VI.

I am writing this from Chengdu. April rain outside the window.

Every morning I still practice an hour of fundamentals. The same way Catherine taught me, in that practice room in Paris.

Not because I need to get better — I'm a working performer, I teach, I curate, I collaborate with brands. Technically I don't need to practice long tones every day.

I practice because the things most worth doing show you no horizon while you are inside them.

That, perhaps, is the deepest meaning of studying classical music in this age.