The Slow Burn of "Mastery"

Why 10,000 hours is still worth it, even when a machine can do it in 10 seconds.

I was recently revisiting Thinking, Fast and Slow by the late, great Daniel Kahneman, and I stumbled upon a passage that stopped me cold. It was about chess, and the sheer, grinding volume of time it takes to become a master.

Kahneman writes that expertise isn’t a single skill, but a collection of mini-skills. He notes that to understand a complex position at a glance, a player needs at least 10,000 hours of dedicated practice. (Yeah, I know Malcolm Gladwell popularised the 10,000 hours concept in Outliers, but let’s not get distracted. This is going somewhere good.)

Anyway. 10,000 hours. That is six years of playing chess for five hours a day. Six years. Five hours a day- Just to get to the starting line of “mastery.”

I had the privilege of seeing Kahneman speak at the Beacon Theatre a few years ago, in conversation with Sam Harris. It was one of those nights where you can feel the collective IQ of the room rising just by proximity. Listening to him talk about his work with his partner Amos Tversky, I was struck by how deeply human their discoveries were. They spent decades mapping the glitches in our own operating systems…

…which brings me back to the 10,000 hours.

In the age of AI, where a machine can be trained on the collective output of every chess grandmaster in history and beat them all before its morning coffee, we have to ask ourselves: Why bother? Why spend six years learning to do something a computer can do in six seconds?

The answer, I think, is in the “doing.”

We’re currently being sold a lie that the value of creation lies solely in the output: the final image, the finished essay, the checkmate. But Kahneman’s point about “intense concentration” and becoming familiar with “thousands of configurations” suggests something else. The value isn’t just in winning the game; it’s in the neural architecture you build to understand the game.

When we cheer for a basketball player making a buzzer-beater in an All-Star game (you guys see that on Sunday?), or stand in awe of a seasoned firefighter navigating a burning building, or even appreciate a perfectly cross-hatched cartoon (Hi, Kevin KAL Kallaugher! 👋), we aren’t just applauding the result. We are applauding the time. We are witnessing the compression of 10,000 hours of struggle, failure, and learning into a single moment of- if I can use the word without grimacing… “grace?” It’s grace.

I’m committing to it. It’s grace.

If you skip the struggle, you skip the meaning.

---

Related Viewing:

---

A machine can mimic the result of expertise, but it can’t replicate the journey of expertise. It’s never felt the frustration of staring blankly at the screen like I am right now, figuring out how to say the thing I’m trying to say without being preachy or dull. (Am I being preachy and dull?) The AI has never had a breakthrough after three weeks of staring at a blank sheet of paper. It has never felt the specific, tactile joy of ink hitting paper in just the right way, and a happy accident creating a splodge that turns into something nobody anticipated.

Painter Helen Frankenthaler reminds us that everything -even the mistakes and misfortunes- is material for your next move:

“You have to know how to use your accidents.”

Source: Paraphrased from Tyler Graphics interview (July 11, 1994)

If you enjoy my work and would like to support, please take out a premium subscription (just $1 per week)

That’s why supporting human artists, writers, musicians, actors, and thinkers is more important now than ever. We aren’t just buying their work; we’re validating the years of life they spent figuring out how to make it. We’re saying that the human act of learning, mastering, and creating has value in itself, regardless of how fast a processor can run.

Emotional learning might be quick, as Kahneman says, but expertise is slow. And thank goodness for that.

The slowness is where the life happens.
…Or so my LLM told me.
(Kidding!!)

Aaaanyway.

‘til next time!
Your pal,