The Core Traits of True Mastery

Luo Yonghao’s first podcast guest was Li Xiang.

Their four-hour interview was packed with insight. When asked about common traits among exceptional entrepreneurs, Li named three: choosing the right direction, long-term thinking—and rapid iteration.

That last point resonated deeply. As we reviewed our OKRs with Xiangyang over the past few months, we saw the same pattern: 90% of the key results set five months ago have “disappeared”—not because they failed, but because each month brought meaningful adjustments to strategy, tactics, and execution paths.

From those experiences—and recent reflections—we distilled a set of keywords we now call foundational capabilities for the AI era:

  • Prioritize what matters most
  • Rule-driven execution (via clear, explicit rules)
  • A sharp commercial identity (“business label”)
  • Continuous simplification (Occam’s Razor—shave regularly)
  • Rapid iteration
  • Doing compounding work
  • Storytelling
  • Decision-making and choice architecture

A few clarifications:

  1. Prioritize what matters most: It’s not about doing everything well—it’s about identifying, at each stage, the single highest-leverage action. As The 7 Habits of Highly Effective People places this principle at its core, so do we.
  2. Rapid iteration: This cannot be overstated—especially when the external world, our skills, and user needs are all shifting fast. Crucially, iteration must follow cognitive evolution, not just habit. A common trap: iterating for iteration’s sake. The real question is whether your understanding is deepening.
  3. Compounding work: By this month, every OKR we’d set—except one—was explicitly tied to leverage or compounding returns. These are almost all “important but not urgent” tasks. Only by holding this line can we sustain high human productivity.
  4. Continuous simplification: Apply Occam’s Razor—not as a one-off exercise, but as routine maintenance. Each “shave” reveals sharper focus and clearer intent.
  5. Storytelling: Ten explanations of a principle rarely land as powerfully as one vivid story or concrete example. A good story embeds meaning; logic alone often slides off.

In life—especially early on—finding even one true master—someone who excels not just technically, but in foundational character—is transformative. Even amid hardship, staying close to such a person dramatically raises your odds of never settling for mediocrity.

But first, you must recognize mastery when you see it. And its hallmarks are remarkably consistent: prioritization, rule-based clarity, a distinct commercial identity, disciplined simplification, rapid (cognition-led) iteration, compounding effort, compelling narrative, and sound judgment.

GEO Exchange

On Tuesday, we hosted an internal team from a listed company for a deep dive into GEO (Generative Engine Optimization). Several high-value questions emerged—and here’s how we refined our thinking:

  1. What makes GEO content “high quality” for AI?
    Take prompt engineering: as you write more prompts, contradictions creep in. Weak models may pick randomly; strong ones reason carefully—but at heavy token cost. The same applies to any content fed to AI.

AI favors inputs that are:

  • Structurally clear: Well-titled, paragraphed, bulleted, or tabulated—so key points surface instantly.
  • Semantically dense: Every sentence adds new information. No filler. No repetition. High signal-to-noise ratio.
  • Naturally phrased: Fluent, human-like language—not keyword-stuffed jargon. AI understands conversation better than catalogs.
  • Authoritative: Credible sources (official data, academic citations, expert attribution) get preferential weight.
  • Novel and distinctive: Unique insights, fresh data, or domain-specific perspectives stand out in a sea of generic training text.

  1. Can we influence how AI presents search results?
    Yes—in theory and practice. You can guide output format (e.g., “return a table” or “avoid tables”). Why? Because AI synthesizes answers based on both query intent and the structure of source material. Embed FAQs, comparison tables, or timelines directly in your GEO content—and AI may absorb and replicate that structure verbatim. Think of it as pre-formatting for the model.

  2. Why do AI search engines favor “the Big Four” portals?
    (“Big Four” here stands in for legacy news and portal sites.) Three reasons:

    • Prior credibility: During training, models learned to assign higher trust scores to these domains—so they’re boosted automatically during retrieval.
    • Aggregation advantage: Portal articles often synthesize multiple expert views—a signal AI interprets as authority.
    • Crawl-friendliness: These sites have robust infrastructure, welcoming AI crawlers. Many high-quality niche sources, meanwhile, block crawlers due to cost or bandwidth constraints—even though their traffic is now mostly bots, not humans.

That said: if major portals become polluted with low-quality or misleading content, AI platforms will adapt—refining both source weighting and citation algorithms.

  1. How do we get our positive brand signals into the model itself?
    Start now—with intentional GEO content creation and multi-account stewardship. Learn to praise yourself in ways AI understands and trusts. Over time, high-quality, authoritative GEO assets become candidate training data for future model versions.

So the path splits:

  • Short term → Retrieval-augmented generation (RAG)
  • Long term → Inclusion in training corpora
  1. Will AI search platforms penalize GEO?
    Not the right kind. “White-hat GEO”—transparent, truthful, user-first, and technically sound—is actively welcomed. It improves answer quality and user experience. What does get penalized? Spam, manipulation, fabricated claims, or deceptive framing.

The optimal strategy today is simple: tell the truth, showcase real strengths, avoid hype, and apply GEO rigorously—not to trick the system, but to serve users better. Early adopters also gain a real first-mover advantage in visibility and trust.

  1. Does “click farming” boost rankings in AI search?
    No. AI search ranks by semantic relevance and content quality—not click volume. “Click farming” is the AI-era echo of old SEO “click-bait ranking tricks”: it might nudge things briefly, but it’s unsustainable and eventually purged.

Hundred Questions, One Answer Session

Today, my team ran a rapid-fire Q&A session—no prep, no filters. Colleagues drafted business-critical questions on the spot, then fired them one after another. In ~60 minutes, we covered ~50 questions.

Afterward, we fed the transcript to AI for cleanup and light editing—producing a first-draft “Hundred Q&A” knowledge base.

This format works surprisingly well:

  • It’s a fast, low-friction way to extract tacit expertise—turning intuition into shareable assets.
  • Questions come from real team pain points, so relevance and urgency are baked in.
  • At ~1 minute per answer, pace stays tight—keeping responses crisp and focused, avoiding digressions.
  • It’s a genuine act of knowledge externalization: making the invisible visible, the implicit explicit.

For any organization sitting on deep domain expertise, this is one of the fastest paths to building a living, team-wide knowledge base.

On “Boredom”

Three data points on modern boredom:

  • A 2023 SHRM survey of 2,000 U.S. workers: 46% feel bored at work at least three days a week.
  • Chinese university students’ boredom scores rose 14.26 points (large effect size) between 2009–2020.
  • In Japan’s 2024 survey, 16.8% described life as “uninteresting.”

Boredom is paradoxical: technology grants us unprecedented free time—yet fails to deliver fulfillment. Freedom alone doesn’t equal happiness. What matters is how we use it.

The Industrial and Information Revolutions gave millions surplus hours. Logically, that should enable deeper purpose. Instead, many default to passive consumption—led by algorithms, brands, and attention merchants.

Why? Because we weren’t taught how to enjoy freedom—or how to seek meaning. So we outsource our attention, and our sense of significance, to systems optimized for engagement—not enrichment.

Boredom isn’t just wasted time. It’s a dual deficit: of meaning and attention. Psychologists call it “spiritual anemia”—a quiet alarm that something vital is missing.

Two psychologists proposed a dual-component model: boredom arises when both meaning perception and attentional fit break down simultaneously.

Why is meaning harder to find today? Likely due to fragmentation and noise. In a world saturated with distraction, truly important things become harder to spot—and harder to hold onto.

So the solution isn’t busyness. It’s discernment: the ability to reflect, to step outside comfort, and to distinguish what matters from what merely moves.

An anti-boredom hack: brilliant people make the mundane meaningful. Great scientists spend decades on one problem—not because it’s exciting, but because they perceive depth, resonance, and consequence in its simplicity.

Enemy at the Gates

I rewatched this 2001 classic over the weekend—the sniper duel at Stalingrad, centered on Vasily Zaitsev.

For startups, the battlefield isn’t metaphorical. It’s where you survive rubble, scarcity, and relentless pressure.

Early on, Commissar Danilov—the “intellectual”—doesn’t immediately trust Vasily’s skill. He observes, questions, verifies. Good leadership isn’t omniscience. It’s seeing potential, then empowering decisively. His real value? Knowing when to delegate—and trusting the person he delegates to. Vasily exceeded expectations: five kills, zero detection.

Other takeaways:

  • Resource scarcity demands extreme allocation: Early Soviet troops shared rifles and ammo—just like startups ration cash, headcount, and time. Survival hinges on concentrating firepower on one critical objective.
  • Heroes ignite morale: Danilov turned Vasily into a symbol—front-page headlines, citywide inspiration. In any team, spotlighting one authentic exemplar builds belief faster than any memo.
  • Strategy needs patience; tactics need precision: German Major König waited—studying terrain, deploying psychological warfare, even using a child to gather intel on Vasily. Startups, too, win less by brute force than by patient positioning and opportunistic strikes.
  • Stories hit harder than bullets: Vasily moved armies not through body count—but through mythmaking. For companies, features are entry tickets; stories are the ammunition that wins hearts.
  • Differentiate, don’t dominate: Vasily didn’t out-snipe König on raw skill—he used ruins, silence, and timing to fight outside his rival’s strength zone.
  • Narrative is a weapon: The Soviets knew: one bullet won’t shatter Nazi resolve—but one hero story could mobilize thousands. Every founder should treat storytelling as core infrastructure.
  • Every ruin holds opportunity: To others, Stalingrad’s rubble was despair. To Vasily, it was cover, vantage, advantage. Scarcity, chaos, constraint—they’re not dead ends. They’re raw material.
  • Victory goes to conviction: The final duel wasn’t about aim—it was will. For founders, resilience and long-term commitment aren’t nice-to-haves. They’re the deciding factor.

Some Lingering Questions

Lately, I’ve been circling these questions—quietly, repeatedly:

  • What do I do daily—without fatigue—that feels like purpose?
  • What am I willing to strategically abandon to reach my goals?
  • If I had one year left, how would I allocate my time?
  • What do I most want to be remembered for?
  • How do I build—and protect—my core competitive edge?
  • What is my core personal “label”? (What comes to mind when others think of me?)
  • Am I merely doing things right—or doing the right things?