Minshen Zhang
Home Blog FLink
<- Back to all posts

The Feynman Mode: A Third Path to Excellence for the Rest of Us

2026-04-16

Rejecting the two dominant "genius narratives" — head-start and leverage — and unpacking three anti-utilitarian cognitive algorithms Feynman left behind: playfulness, fighting knowledge vanity, and the twelve-problems trick that grows real taste.

The Quiet Despair Behind Two Genius Narratives

Our age has two genius narratives cornering us.

One is Sheldon-style: reading at three, admitted to a gifted program by fifteen. At its core is a worship of efficiency — getting ahead, speeding up, winning at the starting line. It satisfies the public's fantasy of a deterministic shortcut.

The other is Altman-style: you don't need to write the hardest code yourself, but you must spot the inflection point, marshal resources, and take on risk. At its core is a worship of leverage — one well-placed bet to move the world. It satisfies the romance of the "hero who makes his moment."

Two narratives, two quiet despairs:

  • Efficiency worship condemns you to "too late" — miss the golden age, and you are out of the race.
  • Leverage worship condemns you to "not enough" — no capital, no appetite for risk, not even a seat at the table.

If you have neither a time machine to sprint nor a fortune to leverage — what's left?

Two "peaks of despair" — Efficiency and Leverage. The ordinary person is pinned between them, but an overlooked trail winds around to the far side. Hover the peaks.

The Third Path: The Feynman Mode

Richard Feynman spent his life validating a third path.

It doesn't depend on timing or circumstance — only on cognitive mechanisms you already have. It is not a shortcut, but an anti-utilitarian personal-development algorithm.

Dig into it, and you find three top-tier mechanisms most of us ignore.

Algorithm 1: Replace "Purpose" with "Play"

Both efficiency worship and leverage worship are built on utilitarianism — winning, making it, securing the resources.

The fatal flaw of utilitarianism: the moment external feedback disappears, the moment boredom or setback settles in for the long haul, your motion deforms and inner drive runs dry. You can grind for a year. You can't grind for ten.

In 1946, during his most lost years at Cornell, Feynman had nothing to publish. One day in the dining hall, someone tossed a plate into the air — it wobbled as it spun. He simply thought it looked fun, and idly started calculating: what's the ratio of the spin rate to the wobble rate?

The answer was 2:1 — the plate spun exactly twice for every wobble.

He took that purposeless little game and extended it to electron orbits; and from electron orbits to what would become his Nobel-Prize-winning work on quantum electrodynamics (QED).

He later wrote in his autobiography:

"It was effortlessly. It was easy to play with these things. It was like uncorking a bottle: everything flowed out effortlessly."

Insight: Curiosity is not a means. It is the end itself. The moment you stop learning "to be useful" and start probing because you want to solve the puzzle, you own the most durable engine in the world. Utility-driven effort wins the first leg; curiosity-driven effort crushes it on a ten-year scale.

A spinning plate treated as a fun little puzzle — extended, year by year, all the way to a Nobel Prize. The moment with zero utility turns out to be the most efficient starting point.

Algorithm 2: The Real Power of the Feynman Technique — Fighting Knowledge Vanity

Most people think the Feynman technique is just "explain it in simple language." That's only half right.

The real power of the Feynman technique is blind-spot detection.

Humans have a powerful illusion: knowing a thing's name feels like understanding the thing. Young Feynman was once in the woods with his father when another boy pointed at a bird and asked, "What's that bird called?" Feynman didn't know. The boy scoffed: "Didn't your dad teach you?" Feynman's father later told him:

"You can know the name of that bird in every language in the world, but when you're finished, you'll know absolutely nothing whatever about the bird. You'll only know about humans in different places, and what they call the bird."

That became the foundation of Feynman's lifelong method: naming is not knowing.

We love piling up jargon, acronyms, and formulas to mask ignorance — and modern research and engineering are drowning in it. "Attention," "emergent," "alignment," "scaling law." Each sounds like understanding when it leaves your mouth; but try explaining it in plain words to an eight-year-old, and you freeze.

Feynman's procedure:

  1. Learn a concept.
  2. Pretend you're explaining it to an eight-year-old. Strip out every term of art, every formula.
  3. The place you stumble is your blind spot — a logical gap.
  4. Go back and patch exactly that piece. Then try the explanation again.

Notice step three — the stumble is more informative than every formula you've memorized. It pinpoints, with surgical precision, the spot where you can fool yourself but not an eight-year-old.

Insight: Real depth is the ability to compress complexity. When you can unpack a cutting-edge idea in plain words, information has genuinely been internalized as "muscle memory" in your brain. The Feynman technique is not primarily a communication tool — it's a diagnostic one. A debugger pointed straight at your own knowledge vanity.

The Feynman technique is a loop: learn → explain in plain words → find the blind spot where you stumble → patch it → try again. The stumble (red) isn't embarrassment. It's a signal.

Algorithm 3: The Twelve Problems — How Taste Is Grown

In a public talk, Chen-Ning Yang named "academic taste" directly — he used the word taste itself, and called it decisive for someone's future work. He once asked a fifteen-year-old prodigy: "Of all these problems, which do you find the most beautiful?"

Yann LeCun talks about taste constantly too: reading papers by Hopfield and Hinton, he simply knew — "this is incredible" — even when neural networks were dismissed as a dead end by the mainstream.

Taste sounds mystical. But Feynman left behind a brutally concrete procedure:

"You have to keep a dozen of your favorite problems constantly present in your mind, although by and large they will lay in a dormant state. Every time you hear or read a new trick or a new result, test it against each of your twelve problems to see whether it helps. Every once in a while there will be a hit, and people will say, 'How did he do it? He must be a genius!'"

Three things hiding inside that quote that most people miss:

  1. The number is 12 — not 1, not 100. One is too few; you get stuck in a single problem and can't climb out. A hundred is too many; new knowledge streams in faster than you can check. Twelve is the size that fits in working memory, randomly accessible.
  2. Most of the time, the problems are dormant. You don't think about them every day. But they are always running in the background. When new input arrives, they wake up and run the match automatically.
  3. A low hit rate is normal. Most new knowledge has nothing to do with your twelve problems. But because you are scanning over long time horizons, hits do happen — and each time one does, the outside world only sees the "genius moment."

Insight: Taste is not a gift from the sky. It is a high-level pattern-recognition skill, trained on a vast cross-disciplinary knowledge base × long-term structural thinking. You know where the treasure is because you have already rehearsed it in your head a thousand times.

Each "piece of knowledge" falls through the twelve dormant problems. Most miss. A few land exactly. Taste is grown in this long, low-hit-rate scan — not in a single flash.

Closing: The Process Is the Reward

We tend to treat "success" as the terminal reward at the end of monastic suffering — grind it out first, and only then are you allowed to enjoy.

The Feynman mode flips this more completely:

To experience, in an ordinary body, real and joyful intellectual growth — that process itself is the highest reward.

Curiosity needs no permit. The Feynman technique needs no resources. The twelve problems don't need you to win at the starting line.

Put the three algorithms together and they form a tight loop:

  • Curiosity supplies the fuel — enough to stay on a single problem for a decade.
  • The Feynman technique supplies quality control — puncturing every fake "I get it."
  • The twelve problems supplies direction — pulling every new input into the same magnetic field and rearranging it.

It is anti-mythological, anti-utilitarian, and within reach. It will not guarantee that you become a Nobel laureate. But it does guarantee that every day you are walking the path of genuinely understanding the world — and that the walk itself is pleasurable, without needing any external trophy.

This loops back to something I wrote in Predictive Coding, Music, and the Vector Space of Conversation: pleasure is the moment your brain is successfully compressing information that was previously incompressible. Taste is your high-level predictive model of "what counts as a good problem." The Feynman technique is online debugging of that model's blind spots. And curiosity-driven play makes the process itself produce dopamine — no Nobel required.

Three "genius narratives." Only the third one doesn't ask you to be chosen.

Reference / inspiration: "[杂谈] On Genius and Taste," by 教令院的大闲者 on Xiaohongshu — original post.