I Reread Peak twice — and finally understood "mental representations"

  • Deliberate practice
  • Mental models
  • AI coaching

From the “ultimate three questions” to using a smart assistant as a coach — pulling mental representations from the book into real life.

I bought the same book five or six years ago, flipped through it, and stopped. This time I picked it up again — and read straight through.

1. Why I could finish it this time

Last time I read Peak (deliberate practice), I only knew it was “about deliberate practice.” This time I followed one spine first: mental representations → how to build them → unconscious decision-making. After that, flipping the table of contents, I roughly knew what came next: the same idea applied to art, work, and life, with examples arguing that many experts are made, not born.

Reading with that prediction, it mostly matched. Once the purpose was clear, the book opened up.

Sometimes you can’t finish a book not because the book is bad — but because you haven’t found the spine yet.

2. What is a “mental representation”? In my own words

The book’s official wording is a bit dense. My take: it is a capability you only gradually gain after deliberate practice.

It is not “hear it once, own it.” It has to be trained. Early deliberate practice barely shows results; you may feel unchanged. But through repetition, intentional correction, and reinforcement, it becomes habit. Once the capability is real, facing related situations you no longer need long deliberation — responses become more automatic, even unconscious. That is the book’s “unconscious decision-making.”

3. My first mental representation: the reading “ultimate three questions”

That reminded me of how I read. After each piece, I use a mind map and deconstruct with the ultimate three questions:

  • What was said?
  • For what purpose?
  • What should be done?

After tearing apart countless pieces, at work — whether talking to people or clarifying requirements — my first reflex is these three. That is probably a beginner-level mental representation forming. I don’t force it; it jumps out on its own.

4. A recent deliberate practice: I let a smart assistant be my coach

Recently, around some personal stuck points, I ran a fairly complete deliberate practice with a smart assistant. It gave me a deeper felt sense of “mental representation.”

Step 1: A smart assistant as practice coach. I threw confusion, small feelings, even triggered flashback memories at a smart assistant. It helped unpack them, gave positive feedback, and offered concrete drills.

Step 2: Something new for me — “remember the feeling.” I practiced deliberately, then consciously recorded bodily and psychological sensations afterward. That broke my long-standing habits.

Step 3: Catch chances — practice again, record again. Find similar scenes in life, practice again, record again. Loop: practice → memory → practice.

Step 4: The key — accept the “flashback.” Facing similar situations, the old thinking or behavior really does flash back like a ghost wall. Before, I might have quit. This time, through talks with a smart assistant, I kept telling myself: “This is normal. Completely normal.” Then I consciously returned to the new method.

Reinforcing that way, a few days ago I suddenly noticed: facing that similar situation, the new way from deliberate practice had become my first reaction — a conditioned reflex. The old flashback seemed overwritten.

I felt, in my body, the shift of a mental representation from deliberate effort to unconscious behavior. And I learned: real change is not that the old habit never appears. It is that when it appears, you accept it — then firmly choose the new path.

5. An unexpected analogy: DeepSeek’s business model feels structurally similar

Oddly, after that section, something loosely related popped up: DeepSeek’s business model.

Roughly:

  • User side: a free tier lowers the barrier and pulls more people in;
  • Revenue side: real money mainly from selling APIs (heavier, commercial use cases pay) plus fundraising (growth and influence traded for resources).

Why does this rhyme with deliberate practice / mental representations? Same path:

Early on, sink heavy cost to first make the capability “stable” and settled (people: repeated deliberate practice; models: repeated R&D iteration); once stable, output results with a lower barrier and less effort (people: more automatic responses; product: free access for more users); then use resources from that output to keep rolling forward.

That “I’ve seen this shape before” feeling finally landed.

6. Closing

Rereading Peak these two weeks, the biggest gain was not a new theory — it was pulling “mental representation” from the page into my life.

From the reading “ultimate three questions,” to deliberate practice with a smart assistant, to seeing a book’s structure — these were mental representations growing quietly.

Ability is not heard. It is practiced. And the sign it has been practiced is that you no longer feel “I am deliberately doing X” — you simply do it.

By Miaoqiu