Glean 拾遗
中 EN
Submit a link
Album · 202pieces

Paul Graham

01

Superlinear Returns

超线性回报：理解世界的真正驱动力

Paul Graham explains why 'you get out what you put in' is rarely true: in business, science, art, and many other fields, performance yields superlinear returns. He traces this to two fundamental causes: exponential growth (learning compounds, startups scale) and thresholds (winner-take-all in sports, science, fame). As technology advances and institutions weaken, more people can now pursue superlinear returns independently. The essay provides practical heuristics: work on what genuinely interests you, keep learning, take calculated risks, and don't conflate a job with your real work. Graham argues that curiosity, not just ambition, is the most powerful way to find those rare opportunities where the reward curve steepens dramatically. A must-read for any ambitious engineer, founder, or creator.

www.paulgraham.com · 25 min
02

What Languages Fix

编程语言都在解决什么问题？

Paul Graham shares a witty perspective on programming languages by describing each in terms of the problem it was designed to fix. From Algol's reaction to low-level assembly to Python's distaste for Perl's complexity, the list captures the evolutionary logic of language design. This concise, humorous comparison offers engineers a fresh way to think about language selection and the historical context of their tools.

www.paulgraham.com · 1 min
03

Power

权力与魅力：为什么正确的人反而招致批评

Paul Graham argues that powerful yet uncharismatic people tend to be disliked because their power attracts criticism that they lack the charm to deflect. Using Hillary Clinton and builder-type CEOs as examples, he notes that such leaders are often the best for the role despite attracting criticism. Graham suggests there is no solution, but that being a magnet for criticism can be a sign of the right person. This is a social psychology and leadership piece, not technical.

www.paulgraham.com · 1 min
04

What Doesn't Seem Like Work?

什么对你来说不像工作？

Paul Graham uses his father's childhood passion for solving math problems to propose a heuristic for choosing work: if something seems like work to others but not to you, that's a strong sign you're suited for it. He gives examples like enjoying debugging (common among programmers) and writing papers for friends in college. The essay encourages readers to explicitly ask themselves what feels effortless to them but laborious to others.

www.paulgraham.com · 3 min
05

Why It's Safe for Founders to Be Nice

创业者为何可以友善：增长率的复利效应

Paul Graham explains why startup founders can safely be nice without hurting their chances of success. The key insight: as long as the product is good enough to spread by word of mouth, growth follows a superlinear curve. Extracting less money per user merely multiplies revenue by a constant factor, leaving the shape of the curve unchanged. A simple calculation shows: at 5% weekly growth, $1,000/month becomes $160k/month in two years; halving revenue only lags by 15 weeks. But increasing growth to 6% while halving revenue yields $214k/month after two years, ahead of the rapacious founder. Thus, optimizing for growth rate, not revenue extraction, is the winning strategy. This essay is targeted at founders and anyone interested in startup growth dynamics.

www.paulgraham.com · 4 min
06

Don't Talk to Corp Dev

创业公司千万别过早接触 Corp Dev

Paul Graham warns startups not to talk to Corporate Development (corp dev) unless you genuinely want to sell the company immediately. Conversations with corp dev are highly distracting and demoralizing, involving tactics like lowball offers and renegotiation. The most dangerous situation is for promising but still small startups. The best defense is to decline the first meeting outright and focus on growth.

www.paulgraham.com · 7 min
07

Having Kids

生儿育女的真相与误解

Paul Graham reflects on his transformation from fearing parenthood to discovering its joys. He admits that his earlier observations were biased by loud children in airports and his own childhood troublemaking. After having kids, he experienced chemical changes, a deep protective instinct, and countless quiet moments of peace. Kids become friends, though they reduce productivity and ambition. Yet the happiness they bring far outweighs the downsides.

www.paulgraham.com · 8 min
08

The Risk of Discovery

科学家为何需要冒险：牛顿的赌注

This essay argues that biographies of famous scientists tend to edit out their mistakes, leading us to underestimate the risks they took. Using Newton as an example, it shows that he worked on physics, alchemy, and theology, all of which seemed equally promising at the time. Only physics paid off, but all three were risky bets. The piece encourages a more nuanced understanding of scientific discovery as a gamble rather than a straight path.

www.paulgraham.com · 1 min
09

Change Your Name

没有 x.com？创业公司该改名了

Paul Graham advises startups that if you have a US startup called X and don't own x.com, you should probably change your name. The reason is not just discoverability but signaling weakness; a marginal domain suggests a marginal company. Founders are in denial due to identity and lack of imagination: they overestimate their name's intrinsic value and underestimate their ability to find better ones. Graham suggests acknowledging that naming is a separate skill and that many cheap or untaken domain pairs exist, as Stripe demonstrated. Data from YC: 100% of top 20 companies by valuation have the matching .com, 94% of top 50, but only 66% of the current batch. Fixing this takes discipline and a couple of days.

www.paulgraham.com · 4 min
10

Let the Other 95% of Great Programmers In

释放全球95%优秀程序员

Paul Graham argues against the idea that training more American programmers can fill the talent gap. He contends that exceptional programming ability is innate and globally distributed: with the US having less than 5% of the world's population, 95% of great programmers are born elsewhere. Training can produce competent programmers but not exceptional ones. The fact that startups jump through legal hoops to hire foreign programmers at the same salary proves the shortage is real, not a plot to lower wages. Graham warns that if the US doesn't liberalize high-skilled immigration, the best programmers may cluster elsewhere, jeopardizing America's tech dominance.

www.paulgraham.com · 6 min
11

The Refragmentation

碎片化的回归：二战后美国社会聚合力量的瓦解

Paul Graham proposes that the current fragmentation in US politics, culture, geography, and economic inequality is not caused by new divisive forces but by the erosion of unifying forces from WWII and large corporations. WWII compressed income gaps through wage controls, high taxes, and national service; post-war oligopolistic firms paired with unions to overpay labor and underpay executives, creating economic and cultural cohesion. Since the 1970s, deregulation, globalization, technology (especially computers), hostile takeovers, and the rise of startups have broken up the Duplo economy, pushing wages toward market prices and amplifying productivity-based income variation. Cultural choices also diversified. Graham argues fragmentation is a return to the historical norm; we cannot replicate the conditions of mid-century cohesion without war or oligopoly, and should focus on mitigating symptoms like inequality rather than fighting fragmentation itself. The essay is a historical and sociological analysis, not technical engineering.

www.paulgraham.com · 42 min
12

The Lesson to Unlearn

别再为了高分而‘刷题’：学校教给我们的错误成功法则

Paul Graham argues that the most damaging thing learned in school is not any subject, but the habit of hacking tests to get good grades — a skill that conflates learning with performance. This mindset persists beyond school: young startup founders instinctively seek shortcuts (how to raise money, get exposure) rather than building great products. Using YC experiences, Graham exposes the deep training we receive in gaming artificial tests, and advocates for both individuals and society to unlearn this lesson. Written for entrepreneurs, educators, and anyone reflecting on achievement.

www.paulgraham.com · 23 min
13

Jessica Livingston

YC幕后灵魂：Jessica Livingston的贡献与隐退

Paul Graham writes about Jessica Livingston, the co-founder of Y Combinator, arguing that she was the key to YC's culture and founder selection. Her 'social radar' for character, combined with her aversion to publicity, meant she often got written out of YC's history. The essay highlights her central role in building the YC community and her expertise in evaluating startup founders.

www.paulgraham.com · 11 min
14

How to Present to Investors

向投资人做演示的实战守则

Paul Graham's classic guide for Y Combinator startups on presenting to investors, based on observations from multiple Demo Day rehearsals. The core advice: in a ten-minute slot, focus on two goals—explain what you're building and why users will want it. Key tips include: state your product in the first sentence (e.g., 'We built an easy web-based database'); demo quickly (assume network failure); prefer a narrow but gripping description over vague possibilities; have one person speak while another runs the computer; avoid deep dives into business models (they'll change anyway); speak slowly and deliberately; use specific numbers and user stories; and craft a memorable soundbite (e.g., 'the Microsoft Word of ecommerce'). The article is practical, packed with concrete examples, and useful for any engineer or founder pitching to a non-technical audience.

www.paulgraham.com · 16 min
15

The High-Res Society

后大企业时代：小团队如何重塑社会规则

Paul Graham argues that for most of history, success correlated with the ability to form large, disciplined organizations. But starting in the late 20th century, economies of scale have been trumped by the speed advantages of smaller groups. He predicts a future dominated by networks of small, autonomous units, using Silicon Valley startups as the model. While this essay is more socio-economic commentary than technical engineering, it offers a foundational perspective on why small teams and startups matter in the modern economy.

www.paulgraham.com · 9 min
16

What I Did this Summer

YC夏季创始人计划：年轻黑客创业的夏天

Paul Graham recaps Y Combinator's first Summer Founders Program (SFP), with 8 teams (ages 18-28, avg 23), minimal funding ($20k), and surprising results: 7 of 8 had prototypes in 10 weeks, 3-4 are expected to succeed. Key insights: startup success depends more on intelligence & energy than age/experience; young founders exhibit strong responsibility when given autonomy; initial versions must be lightweight to evolve. The article also critiques incubators and notes the power of independence.

www.paulgraham.com · 15 min
17

Web 2.0

Web 2.0 的本质：Ajax、民主与善待用户

Paul Graham dissects the term 'Web 2.0' in this 2005 essay, arguing that while originally a conference branding, it has come to represent three key elements: Ajax enabling desktop-like web apps, democracy (user-generated content and voting surpassing professional editors), and treating users well (free, no registration, no abuse). The common thread is using the web as it was meant to be used, exemplified by Google. The essay also offers sharp observations about the dot-com bust and startup culture.

www.paulgraham.com · 16 min
18

Why YC: A Hacker's Hack on the Economy

为什么创办 YC：一位黑客对经济的巧妙改动

In this essay, Paul Graham explains the real motivation behind founding Y Combinator. He denies it was for money or pure altruism, instead describing it as a 'hack' — applying a small amount of force at the right point to spring a stream of startups into existence that otherwise wouldn't, thereby improving the global economy. He also reflects on Viaweb, noting that startups are mostly schleps, and that YC feels like a band reunion with old friends.

www.paulgraham.com · 2 min
19

The Venture Capital Squeeze

风投的夹缝：资金过剩、创业成本降低与IPO停滞

Paul Graham analyzes the squeeze on venture capital from four directions: excess capital from the bubble, plummeting startup costs due to open source/Moore's law/the web/better languages, Sarbanes-Oxley killing IPOs, and acquirers buying early (like Google). He suggests lobbying to loosen Sarbanes-Oxley and allowing founders to cash out partially in Series A. The piece is business/startup strategy, not AI/engineering.

www.paulgraham.com · 9 min
20

Hiring is Obsolete

雇佣已过时：为何创业比打工好

Paul Graham's classic 2005 essay argues that the drastically lowered cost of starting a startup—now essentially just food and rent—has shifted power from employers and investors to young people. Instead of being hired as a proxy for customer valuation, talented individuals can go directly to users by founding their own companies. Large corporations increasingly buy startups as a form of hiring, fusing recruitment with product development. Graham stresses that risk is proportional to reward and that youth is the ideal time to take career risks, as even failure provides valuable experience that employers appreciate. This essay is a foundational read for engineers and students considering entrepreneurship.

www.paulgraham.com · 27 min
21

Design and Research

设计与研究：编程语言设计的艺术

Paul Graham dissects the difference between design (focused on 'good') and research (focused on 'new'), arguing that programming language design should be treated as a design problem, centered on user needs (not wants). He advocates rapid prototypes (Worse is Better), iterative refinement, always having working code for morale, and insists that good design requires a single dictator. A classic essay on the art of design, applicable beyond languages.

www.paulgraham.com · 15 min
22

The Roots of Lisp

Lisp 的根源：McCarthy 如何用七个操作符定义一种语言

Paul Graham revisits John McCarthy's seminal 1960 paper on Lisp, explaining how a complete programming language can be built from just seven primitive operators (atom, quote, car, cdr, cons, eq, cond) and function definition. By translating McCarthy's mathematical notation into working Common Lisp code, the article demonstrates Lisp's defining quality—it can be written in itself—and argues that Lisp and C represent the two cleanest models of programming, with modern languages steadily moving toward the Lisp model. Aimed at readers interested in programming language history and language design principles.

www.paulgraham.com · 2 min
23

Lisp for Web-Based Applications

Paul Graham 谈 Viaweb 为何选择 Lisp：实战中的宏与灵活性

In a 2001 talk, Paul Graham explains the specific technical advantages of using Lisp for Viaweb, a web-based application. He focuses on how Lisp macros allow developers to extend the language syntax, achieving a level of abstraction that other languages struggle to match, leading to faster development and more maintainable code. Though dated, the talk offers valuable insights into how language design impacts engineering productivity in practice. Recommended for engineers interested in programming language history, Lisp philosophy, or technology trade-offs.

www.paulgraham.com · 1 min
24

Better Bayesian Filtering

贝叶斯垃圾邮件过滤实战：五个关键改进

Paul Graham's 2003 talk at the Spam Conference presents five improvements to Bayesian spam filtering that boosted catch rate from 92% to 99.75% and reduced false positives from 1.16% to 0.06%. Key changes: using email headers, preserving case, no stemming, using only the 15 most significant tokens, and biasing against false positives by double-counting nonspam corpus occurrences. He introduces token degeneration to handle unseen tokens and contextual markers (e.g., 'Subject*') to improve discrimination. The article analyzes four missed spam types and three false positives, treating false positives as bugs rather than statistical errors. A practical deep-dive for engineers interested in text classification implementation.

www.paulgraham.com · 25 min
25

Paul Graham's Homepage

保罗·格雷厄姆的个人主页

This is the homepage of Paul Graham's personal website, containing links to his latest essays (e.g., 'How to Earn a Billion Dollars' and 'The Brand Age') and an ad for Y Combinator. The page itself has no technical content, serving only as a navigation index.

www.paulgraham.com · 1 min
26

Five Questions about Language Design

编程语言设计五问：为人、为友、为简洁、为实用

Paul Graham's notes for a 2001 MIT panel on programming language design, still insightful today. He argues that languages must be designed for human weaknesses, not just mathematical elegance; the best languages are designed by their authors for themselves; give programmers maximum control; aim for brevity in both code and manual; and acknowledge that hacking is about making great programs, not just producing research papers. He also discusses server-based software's implications (support for frequent releases, continuations), library organization, the connection between syntax and semantics, why OOP isn't the one true model, and the dangers of design by committee. A must-read for engineers interested in language design and software engineering philosophy.

www.paulgraham.com · 17 min
27

How to Make Wealth

创业致富的逻辑：测量、杠杆与用户验证

Paul Graham's classic essay argues that starting a startup is the best way to get rich because it compresses a lifetime of work into a few years via efficiency multipliers (e.g., working harder, smarter, and in a small team). Wealth is what people want, not money; it can be created, not just divided. Startups provide measurement through small teams and leverage through new technology. The ultimate test of wealth creation is user adoption. The essay also covers risks, acquisition strategy, and the historical context of wealth creation.

www.paulgraham.com · 50 min
28

How to Do Great Work

如何做出伟大工作：Paul Graham 的实用指南

Paul Graham distills a four-step framework for doing great work: choose a field you have natural aptitude and deep interest in, learn enough to reach its frontier, notice gaps in knowledge, and explore promising ones. He emphasizes curiosity as the driving force, and advises honesty, boldness, and acceptance of failure. The essay is filled with personal anecdotes and counterintuitive ideas, such as 'don't over-plan' and 'follow your interests over imitation.' Suitable for anyone aspiring to achieve greatness in technology, entrepreneurship, or academia.

www.paulgraham.com · 67 min
29

How to Get New Ideas

发现新想法的方法：关注异常与知识前沿的缝隙

The essay argues that new ideas come from noticing anomalies—things that seem strange, missing, or broken. The best place to look is at the frontiers of knowledge, which grow fractally: from a distance edges look smooth, but up close they are full of gaps. These gaps are obvious to those who have learned enough to get near them, yet unexplored. Exploring them can yield entirely new fractal buds.

www.paulgraham.com · 1 min
30

The Need to Read

写作是思考的不可替代途径

Paul Graham argues that reading and writing are not just conduits for information but essential tools for thinking. He debunks the sci-fi fantasy of direct knowledge upload, asserting that even if feasible, it would be insufficient because writing itself generates new ideas — it's a process of discovery. Good writers don't simply transcribe pre-formed thoughts; they uncover new ones while writing. This is irreplaceable for tackling complex, ill-defined problems. Without writing skill, one is always at a disadvantage in deep reasoning. And good writing stems from good reading. The essay is a defense of traditional literacy in an age of efficiency, relevant to any engineer who relies on deep thinking — architecture design, debugging, or technical documentation — as writing amplifies reasoning quality.

www.paulgraham.com · 3 min
31

What You (Want to)* Want

你无法控制最终想要什么：一个哲学悖论的计算化表述

Paul Graham revisits a puzzle he pondered since age 9: if we are merely machines made of predictable matter, how can we have free will? He starts with the common wrong answer — "You can do what you want, but you can't want what you want" — and progressively refines it. The key insight: people can change what they want (drug addicts quit, people learn to like broccoli), but the recursive chain of "want to want to want..." always terminates at a desire you cannot control. Graham formalizes this using a regex: `(want to)* want`. Just as 0.999... approaches 1, you have freedom at finite orders of wanting, but there is always an nth-order desire beyond your control. This is a philosophical essay about consciousness and determinism, not a technology/engineering piece. It will appeal to readers interested in free will, the nature of consciousness, and mathematical framing of philosophical problems, but is off-topic for this publication's focus on AI and systems engineering.

www.paulgraham.com · 3 min
32

Alien Truth

异星真理：寻找与外星智慧共享的普遍原则

Paul Graham introduces 'alien truth' — truths that any intelligent beings, beyond those of math and physics, would share with us, such as the value of controlled experiments, practice, and Occam's razor. He suggests it as a heuristic for seeking general truths, akin to Erdos's 'God's book'. The essay also touches on AI as potential alien intelligence, but stresses philosophy's role in using this target rather than defining its precise threshold.

www.paulgraham.com · 4 min
33

What I've Learned from Users

保罗·格雷厄姆：从 YC 创业公司身上学到的教训

Paul Graham reflects on lessons from Y Combinator’s startups. Most startups share similar problems, yet advising cannot be automated; individual office hours are essential. Founders frequently misdiagnose their issues (e.g., attributing funding struggles to external factors rather than product weakness) and often ignore advice because startup truths are counterintuitive. YC’s key value lies in sharpening focus on what matters most, enabling rapid weekly experimentation and course correction. Even more impactful than partner advice is the peer community; YC deliberately fosters a cluster of ambitious founders who support each other. This essay offers insights for startup founders, YC applicants, and those curious about the startup ecosystem.

www.paulgraham.com · 13 min
34

Putting Ideas into Words

写作是思想的最严苛试炼：从模糊到精确的必经之路

Paul Graham argues that writing is not merely recording pre-existing thoughts, but a rigorous process that tests, completes, and generates ideas. He notes that first drafts are invariably wrong, that half the ideas in an essay emerge during writing, and that reading one's own work as a stranger reveals gaps and imprecision. While formal domains like chess allow purely mental refinement, most complex ideas require the strict test of committing to an optimal sequence of words. He asserts that anyone who hasn't written about a topic has no fully formed ideas about it, and those who never write have none about anything nontrivial. This short essay is a compelling read for engineers and thinkers who rely on deep understanding and clear communication.

www.paulgraham.com · 7 min
35

Is There Such a Thing as Good Taste?

存在好品味吗？——一次归谬论证

This essay argues that there is such a thing as good taste, using reductio ad absurdum. If good taste doesn't exist, then good art doesn't exist, meaning no artist can be better than a random eight-year-old—an absurd conclusion. The author, Paul Graham, shares his own experience learning to paint and realizing that masters like Leonardo and Bellini are objectively better. He addresses common objections: that taste is too subjective and that experts often disagree. He concludes that while perfect taste is impossible, good taste is real, and art can be judged by how effectively it works on an audience with deep knowledge and clarity. The piece is a philosophical reflection, not a technical engineering article.

www.paulgraham.com · 6 min
36

Weird Languages

怪异语言的价值：从Lisp宏看编程的另一种可能

Paul Graham argues that 99.5% of programming consists of gluing library calls together, and all popular languages are equally good at that. The remaining 0.5% is disproportionately interesting. The weirdness of languages like Lisp is not accidental; it points to techniques beyond glue programming—for example, Lisp macros let you first write a language for a problem domain and then write your application in that language. Learning weird languages expands your concept of what programming can be and teaches you to think thoughts you couldn't think before.

www.paulgraham.com · 2 min
37

Heresy

异端复苏：当代社会何以重拾中世纪的言论禁忌

Paul Graham examines the secular rebirth of heresy, where expressing certain opinions is treated as a crime, overriding truth and past contributions. He identifies the aggressively conventional-minded as enforcers, mobilized by a purity‑based ideology that emerged in US universities in the late 1980s. Though absolute freedom of expression remains wider than centuries ago, the trend has reversed since 1985 — a problem analogous to declining vaccination rates. Graham argues for sustained pushback to develop social antibodies and believes the current wave may be peaking. The essay intentionally avoids specific examples to remain timeless. Useful for readers interested in free speech, social dynamics, and workplace culture.

www.paulgraham.com · 13 min
38

Beyond Smart

超越聪明：智力与创造性之间的真正差距

Paul Graham challenges the common notion that intelligence is the key to achievement. He argues that being smart is not the same as having new ideas — many highly intelligent people fail to produce anything novel. The real differentiators are qualities that can be cultivated: obsessive interest in a topic, independent-mindedness, writing ability (as a mode of thinking), and practical habits like hard work and sleep. This essay invites readers to shift focus from 'being smart' to 'generating new ideas' and explores the gap between the two.

www.paulgraham.com · 8 min
39

A Project of One's Own

滑冰式工作：自主项目的乐趣与生产力

Paul Graham explores the profound difference between working on a project of your own and ordinary assigned work, likening the former to skating versus walking. He argues that this self-driven mode is not only more enjoyable but far more productive, and traces how childhood passions can evolve into adult careers if not stifled by schooling. Using examples from his father’s approach to math textbooks to the original Macintosh team, he shows that autonomy and a sense of ownership can persist even within collaborative environments. The essay warns against the tyranny of high standards that prevent starting new projects and advocates recapturing the careless confidence of a child.

www.paulgraham.com · 14 min
40

How to Work Hard: Talent, Practice, and Honest Self-Assessment

努力工作的真相：天赋、练习与诚实的自我评估

In this reflective essay, Paul Graham distills what he has learned about working hard since his school days. He argues that great work requires a combination of natural ability, practice, and effort, and that no amount of brilliance can substitute for hard work. Key insights include: learning to distinguish real work from the fake versions in school, finding one's personal limit for intense work (e.g., ~5 hours for writing/programming), and the necessity of honest self-assessment to avoid both laziness and overwork. He emphasizes that deep interest is a better guide than external rewards, and that the decision to persist or switch fields is an ongoing part of working hard. The essay draws on examples from Bill Gates, Lionel Messi, G.H. Hardy, and his own startup experience.

www.paulgraham.com · 18 min
41

An NFT That Saves Lives

一枚拯救生命的 NFT：Noora Health 的慈善拍卖

Paul Graham writes about a new NFT launched by Noora Health, a nonprofit he has long supported. Noora Health reduces infant mortality in South Asia by teaching mothers post-discharge care; their data shows they save 9 babies per 1000 live births. Each life costs only $1,235, and the NFT's proceeds are tracked transparently. The reserve price of ~$2.5M corresponds to saving 2,000 lives, and every additional dollar goes directly to saving more lives. The article is a charitable call to action, highlighting NFT's potential for social impact.

www.paulgraham.com · 2 min
42

Fierce Nerds

凶猛极客：沉默外表下的好斗灵魂

Paul Graham defines 'fierce nerds'—people who appear shy but are intensely competitive, overconfident, independent-minded, and impatient in their domain. Using Watson and Crick's discovery of DNA as an example, he argues that these traits, if channeled into ambitious projects, lead to great success; otherwise, they turn into bitterness and online trolling. Graham notes that the modern world increasingly rewards getting the right answer, giving fierce nerds an edge (7 of 8 richest Americans are fierce nerds). The essay is a call to young nerds to embrace their fierceness and work on hard problems.

www.paulgraham.com · 8 min
43

Crazy New Ideas

如何对待‘疯狂’的新想法

Paul Graham argues that we should be cautious about dismissing ideas that sound preposterous when proposed by reasonable domain experts, as history shows that major breakthroughs often start as crazy-sounding notions. He explains the psychological and social reasons people dismiss new ideas—envy, sophistication signalling, vested interests, and factionalism—and suggests that the wise response is to ask questions rather than make statements. He also offers advice for those seeking to generate new ideas: study how they look at birth.

www.paulgraham.com · 8 min
44

The Real Reason to End the Death Penalty

废除死刑的真正原因：无辜者被处决

Paul Graham argues that intellectual debates about the death penalty—whether the state has the right to take life, whether it deters crime, whether it's applied unfairly—miss the real issue: in practice, a significant fraction of death row prisoners are innocent (at least 4%). He presents multiple documented cases of wrongful convictions, including Kenneth Adams (18 years), Keith Harward (33 years), and Cameron Willingham (executed despite discredited arson evidence). The core problem is systemic: police tunnel vision, prosecutorial misconduct, incompetent defense attorneys, false witnesses, and junk science. Capital punishment in practice means executing innocent people.

www.paulgraham.com · 5 min
45

Write Simply

写作应如 Saltimbocca：让思想跃入读者脑中

Paul Graham advocates writing with ordinary words and simple sentences, arguing that reducing friction in reading allows readers to focus more on ideas. He introduces the concept of 'saltintesta' (jumping into the head) as the goal, analogous to the Italian dish saltimbocca. Simple writing aids non-native speakers, prevents hiding lack of content, and lasts longer. Graham admits his main reason for simplicity is that complex sentences feel clumsy to him.

www.paulgraham.com · 3 min
46

Donate Unrestricted

无限制捐款：把钱交给非营利组织去花

Paul Graham argues that restricted donations, where donors specify how money must be used, are typically suboptimal because nonprofits understand their own needs better. Exceptions include umbrella organizations like universities and expert donors like the Gates Foundation. He urges donors to give unrestricted, trusting the nonprofit to allocate funds effectively. The essay also explores donors' mixed motives and nonprofits' reluctance to speak up due to financial anxiety.

www.paulgraham.com · 3 min
47

How People Get Rich Now

当代致富之路：创业取代继承与资源

Paul Graham compares the Forbes 400 lists of 1982 and 2020, revealing a dramatic shift in how people become rich. In 1982, 60 of the top 100 fortunes were inherited, and at least 24 of the 40 new fortunes came from oil or real estate. By 2020, only 27 fortunes were inherited, and of the 73 new fortunes, about 30 came from tech companies, with 8 of the top 10 being tech founders. Graham argues this is a return to the historical norm, suppressed in the mid-20th century by oligopolies and industrial consolidation. Falling startup costs, faster growth, and cultural acceptance now make it easier to get rich by starting a company. The article offers a data-driven perspective on wealth creation and economic history.

www.paulgraham.com · 15 min
48

Earnestness

创业者为何需要真诚——硅谷价值观解析

Paul Graham argues that the highest compliment for startup founders in Silicon Valley is to be called 'earnest'—meaning they have the right motives and try their hardest. This quality stems from a genuine interest in the problem itself, a trait shared with 'nerds.' While earnestness may be a social disadvantage in youth, it becomes a powerful asset when problems are hard. Graham notes that earnestness doesn't matter in fields like politics or crime, and that a bit of naivete can be beneficial in fast-changing domains. He concludes by highlighting the growing alignment between intellectual curiosity and money as one of the most fascinating changes of our time.

www.paulgraham.com · 10 min
49

The Truth About YC Interviews and Billionaires

YC 面试诀窍与亿万富翁的真实驱动力

Paul Graham combines two topics: how to succeed in a Y Combinator interview and why the political claim that billionaires exploit people is false. He explains that YC interviews are not pitches; partners seek evidence that founders deeply understand user needs. The most convincing answer is 'we and our friends want it,' ideally with a prototype already in use. Using Airbnb as an example, he shows that even without immediate growth, deep user insight and founder resourcefulness can secure funding. Graham argues that the defining trait of billionaires is genuine interest in building, not exploitation, and the most reliable path to wealth is delighting users.

www.paulgraham.com · 19 min
50

The Airbnbs

Airbnb 创始人的执着：从濒临放弃到 IPO 的历程

Paul Graham recounts the early days of Airbnb, highlighting the earnestness and relentless effort of founders Brian Chesky, Joe Gebbia, and Nathan Blecharczyk. Despite near-zero growth, maxed-out credit cards, and a walking-out investor, they persisted because of the unexpected joy and value they discovered in hosting. By focusing on New York City, personally photographing listings, and achieving ramen profitability ($4,000/month) in early 2009, they finally gained traction. The story emphasizes their tireless work ethic and the pivotal role of Y Combinator.

www.paulgraham.com · 6 min
51

Modeling a Wealth Tax

财富税复利效应：创始人股票被侵蚀的数学模型

Paul Graham uses simple compound interest calculations to model the long-term impact of a wealth tax on startup founders. Assuming a founder holds stock for 60 years, he computes the share taken by the government at various tax rates (0.1% to 5%) and with a $50 million exemption threshold. The key finding: because a wealth tax is applied annually to the same asset, its compounding effect means even a 2% tax consumes about two-thirds of a successful founder's stock over a lifetime. The article also explores sensitivity to initial valuation and growth rates, noting that actual effects may be worse due to dividend taxes. Relevant for readers interested in the trade-offs between wealth taxation and entrepreneurial incentives.

www.paulgraham.com · 3 min
52

How to Think for Yourself

独立思考的三块肌肉：对真理的考究、反灌输与好奇心

Paul Graham examines the essence and cultivation of independent-mindedness. He argues that in fields like science, investing, and startups, being correct is not enough—one must also have novel ideas. Independent-mindedness is partly innate but can be strengthened by surrounding yourself with nonconformists, reading history, and practicing skepticism. He breaks it down into three interchangeable components: fastidiousness about truth, resistance to being told what to think, and curiosity. The essay also explores how independent thinking gets diluted in growing companies and how exposure to diverse groups can shield against conformism.

www.paulgraham.com · 21 min
53

Early Work

如何克服对早期作品的恐惧：保罗·格雷厄姆的实用指南

Paul Graham explores the fear of making something lame that holds people back from ambitious projects. He argues that great work often goes through an unimpressive early phase, and offers practical strategies to push through it: overestimate the importance of your project, be slightly overconfident, leverage ignorance, surround yourself with supportive colleagues, focus on learning rather than output, and treat early work as sketches or quick hacks. Drawing on Y Combinator experience, he also discusses how Silicon Valley's culture of optimism evolved from self-interest into a beneficial custom. This essay is valuable for engineers and creators seeking to overcome self-doubt and start ambitious work.

www.paulgraham.com · 14 min
54

Orthodox Privilege

正统思维盲点：为何安全表达观点本身就是一种特权

Paul Graham coins "orthodox privilege" to describe the blindness of conventional-minded people who cannot conceive of true statements that would get one in trouble. Because their own opinions align with current acceptability, they assume everyone can speak safely. The essay urges politeness: if someone says they can hear an inaudible high-pitched noise, you should take their word for it—likewise for claims of unsayable truths.

www.paulgraham.com · 4 min
55

The Four Quadrants of Conformism

循规蹈矩的四象限：人格分类新视角

Paul Graham presents a 2x2 framework classifying people by their degree of conformity (conventional vs. independent) and aggression (passive vs. aggressive), yielding four types: aggressively conventional, passively conventional, passively independent, and aggressively independent. He argues these types are universal, appear in childhood, and are more personality-driven than rule-driven. The aggressively conventional are a disproportionate source of trouble, while the independent generate new ideas but need protective customs (like free inquiry). Graham notes universities are declining as safe havens but is hopeful that the independent will create new institutions.

www.paulgraham.com · 12 min
56

Coronavirus and Credibility

疫情是公信力的试金石

Paul Graham argues that the coronavirus pandemic serves as an unusually powerful and unequivocal test of credibility. Unlike vague topics where false predictions go unpunished, epidemics rapidly falsify claims with hard data. Drawing on Warren Buffett's analogy, he notes how the event exposes those who 'swim naked'—public figures who confidently speak on matters they don't understand. The article urges readers to remember the results as the most accurate credibility test we're likely to have. Relevant for those interested in public trust and cognitive biases.

www.paulgraham.com · 1 min
57

Being a Noob

新手感觉的价值

Paul Graham reflects on the positive value of feeling like a noob. He observes that the more you feel like a noob locally, the less of a noob you are globally. This discomfort is an evolutionary relic useful in stable environments but now misaligned in a rapidly changing world where novelty is abundant. Embracing the noob feeling is a sign of learning and growth.

www.paulgraham.com · 2 min
58

Haters

黑粉与脑残粉：成名后的两种负面效应

Paul Graham argues that haters and fanboys are fundamentally the same type of person, just with the sign flipped. Fanboys are obsessive and uncritical, idealizing everything you do; haters are equally obsessive and uncritical, demonizing everything you do. Both create a distorted image of you and treat you as a fraud. Graham advises not to engage with haters as if they were rational disputants—instead, ignore them just as you would an obsessive fan. He notes that haters are usually underachievers who project their own frustration. This essay is for anyone who becomes famous enough to attract haters and wants a practical mindset for handling them.

www.paulgraham.com · 8 min
59

Fashionable Problems

冷门问题的价值：别让流行趋势左右研究方向

Paul Graham observes that even the smartest people tend to work on fashionable problems, leaving much of the possibility space unexplored. He suggests looking at fields thought to be fully explored (essays, Lisp, venture funding) and finding new approaches. The best protection is genuinely loving what you do, so you'll persist even if you mistakenly think it's too marginal.

www.paulgraham.com · 1 min
60

How to Write Usefully

如何写出有用的文章

Paul Graham proposes a formula for useful writing: importance + novelty + correctness + strength. The key method is the 'Morris technique' — only publish what you're sure is worth hearing. He advises writing simply, using qualifications skillfully, and treating yourself as a proxy for the reader. He also notes that the internet has revived the essay form, making it possible for anyone to publish and discover new ideas.

www.paulgraham.com · 16 min
61

The Two Kinds of Moderate

故意温和与偶然温和：政治光谱上的两种人

Paul Graham distinguishes two types of political moderates: intentional moderates who deliberately position themselves midway between extremes, and accidental moderates who independently form opinions on each issue, resulting in a broad distribution averaging to center. He argues that the former are akin to ideologues in adopting ready-made positions, while the latter exhibit genuine independent thinking. The essay concludes that most impressive people working with ideas are accidental moderates, as independence is crucial for excellence. This piece is about political philosophy, not relevant to AI/engineering.

www.paulgraham.com · 4 min
62

Novelty and Heresy

新颖与异端：创新必遭指责？

Paul Graham argues that any truly novel discovery is likely to be branded as heresy because good ideas are often hidden in the shadow of cherished mistaken assumptions. Suppressing heresy not only protects those assumptions but also eliminates ideas that indirectly contradict them. Every cherished mistaken assumption creates a dead zone of unexplored ideas around it; the more preposterous the assumption, the larger the dead zone. Conversely, seeking out heresies is a powerful way to find new ideas. Using Galileo and Darwin as examples, he warns that organizations or societies overzealous in pouncing on heresy risk stifling innovation.

www.paulgraham.com · 2 min
63

How to Make Pittsburgh a Startup Hub: A City's Recipe for Tech Gravity

如何让匹兹堡成为创业中心：少干预、多包容的城市生态建议

In this 2016 talk at Opt412, Paul Graham argues that Pittsburgh can become a startup hub not by copying Silicon Valley, but by doubling down on its existing strengths: cheap housing, a rising share of 25-29 year olds, an independent food scene, historic walkable neighborhoods, a world-class university (CMU), and a tolerant, pragmatic culture. He advises against large real estate projects or forced innovation programs; instead, the city should make it easier to open restaurants, preserve old buildings, become the most bike-friendly place in the US, let CMU focus on research excellence, and tolerate strangeness. Investors will come slowly, but lower startup costs and remote funding (Kickstarter, YC) reduce the need. The talk is insightful for urban planners and startup advocates, but has limited direct engineering takeaways.

www.paulgraham.com · 15 min
64

The Bus Ticket Theory of Genius

天才的巴士票理论：痴迷于无用之事

Paul Graham argues that genius requires not only natural ability and determination, but also a third, often overlooked ingredient: an obsessive interest in a particular topic, like a bus ticket collector. This disinterested obsession drives figures like Darwin and Ramanujan to follow paths that seem unpromising but lead to great discoveries. Graham suggests that instead of pursuing conventionally ambitious projects, one should cultivate a collector-like depth of interest, even if it seems wasteful. He also discusses implications for education and maintaining creativity with age. This essay is a philosophical reflection on the nature of genius and creativity.

www.paulgraham.com · 15 min
65

A Way to Detect Bias

检测选择偏差：表现对比法

Paul Graham describes a method for detecting bias in selection processes without knowing the applicant pool: if a selection process is biased against a group, those selected from that group must be better and thus will outperform other successful applicants. Using the example of venture capital bias against female founders, he notes that First Round Capital's own data showed startups with female founders outperformed by 63%, inadvertently revealing bias. The method works when a random sample of selected applicants exists, their subsequent performance is measured, and the groups have roughly equal ability distributions.

www.paulgraham.com · 3 min
66

The Ronco Principle

罗恩科原则：为何在创业世界，真正的好人才能成功

Paul Graham uses angel investor Ron Conway as a paragon to argue that in a transparent and unpredictable startup world, being genuinely good is not just a moral choice but a winning strategy. Bad behavior gets exposed quickly, and rapid change makes it impossible to know whom to treat well; thus, only authentic benevolence can sustain long-term success. The principle extends beyond startups to any domain trending toward transparency and unpredictability.

www.paulgraham.com · 4 min
67

Life is Short

生命短暂，别浪费在无谓之事上

Paul Graham argues that life is actually short, not just finite, using the discrete nature of time with young children as evidence. He suggests relentlessly pruning bullshit (unnecessary meetings, arguments, bureaucracy, online fights), actively doing what matters, and savoring time. He advises not to wait to climb mountains or visit mothers, and to be impatient about important things. The tone is personal and philosophical, with practical advice for time management and life priorities.

www.paulgraham.com · 9 min
68

Write Like You Talk

像说话一样写作

Paul Graham offers a simple trick to make writing more readable: write in spoken language. He observes that most people adopt a more complex, formal style when writing, which creates distance and makes reading harder. Complex ideas don't require complex sentences—experts discussing abstruse topics use simple sentences. His advice: after your first draft, examine each sentence and ask 'Would I say this to a friend?' If not, replace it with what you would say. For heavily written pieces, try explaining your draft to a friend verbally, then use that explanation as your draft. Graham reads every essay aloud before publishing to fix anything that doesn't sound like conversation. This approach can put you ahead of 95% of writers.

www.paulgraham.com · 4 min
69

How You Know

阅读如何内化为思维模型

Paul Graham reflects on why he forgets most of what he reads but still finds reading valuable. Using a biography of Hilbert as a catalyst, he realizes that reading and experience train your mental model of the world; even when you forget the specifics, their effect persists. Your mind is like a compiled program whose source code is lost. The real harvest of reading isn't remembered facts but the lasting change in your mental framework. Specific implications: important books are worth rereading (your brain compiles them differently at different stages of life), and technology may one day let us relive, index, and edit our experiences.

www.paulgraham.com · 4 min
70

Mean People Fail

善良者的竞争优势：为什么友善的人更容易成功

Paul Graham observes that almost none of the most successful people he knows are mean, despite meanness being common online. He argues that meanness makes you stupid (fights waste mental energy on tactical tricks rather than big ideas), prevents attracting top talent (who have better options), and contrasts with the benevolence that drives founders to build enduring great things. He traces a historical shift from zero-sum games (where meanness helped) to positive-sum games of ideas and creation (where kindness becomes an advantage). The essay synthesizes observations from his wife Jessica Livingston and YC experience.

www.paulgraham.com · 7 min
71

How to Be an Expert in a Changing World

警惕过时的专业：Paul Graham谈如何保持开放思维

Paul Graham argues that experts often become obsolete because they are experts on an earlier version of the world. Drawing on his decade of early-stage startup investing at Y Combinator, he suggests strategies to avoid outdated beliefs: don't try to predict the future; instead be aggressively open-minded and sensitive to change. Focus on people rather than ideas, since good new ideas come from earnest, energetic, independent-minded individuals. Turn your comments into public bets (e.g., writing articles) to force yourself to be more careful. He uses Airbnb as an example of a seemingly bad idea that succeeded because of the founders. This essay is for entrepreneurs, investors, and anyone wanting to stay intellectually agile.

www.paulgraham.com · 6 min
72

The Fatal Pinch

创业死亡之钳：融资幻觉与自救路线图

Paul Graham defines the 'Fatal Pinch,' a common startup death trap: significant bank balance but high monthly burn and stagnant revenue, with founders delusionally relying on another fundraising round. He dissects the three converging forces (higher spending, higher investor standards, company perceived as failing) and prescribes a clear triage: assume fundraising probability is zero, then either shut down, cut costs (layoffs or salary cuts), or pivot to more consulting-like revenue generation. The key is to stop denial and act immediately.

www.paulgraham.com · 9 min
73

Before the Startup: Just Learn

大学里该做什么？Paul Graham：别急着创业，先学习

In this 2014 essay based on a Stanford guest lecture, Paul Graham advises college students against starting startups while in school. He outlines six counterintuitive points: startups defy intuition; expertise in startups matters less than understanding users; gaming the system doesn't work; startups are all-consuming; you can't predict your own fitness; and the best startup ideas emerge from genuine curiosity, not forced brainstorming. His bottom line: just learn.

www.paulgraham.com · 25 min
74

Investor Herd Dynamics

投资人的从众心理

Paul Graham breaks down the herd mentality in startup investing, where one investor's interest triggers a cascade. He outlines four drivers: fundraising decreases risk and increases a company's value; founder confidence rises, impressing investors; mediocre investors rely on others' judgment; and competing term sheets impose deadlines. He cautions that a stampede doesn't guarantee success, and shares practical negotiation tactics, such as never revealing other VCs' identities. Drawing from firsthand Y Combinator experience, the essay offers a clear-eyed view of fundraising dynamics for founders.

www.paulgraham.com · 6 min
75

How to Raise Money

创业融资实操指南：从规则到策略的完整地图

Paul Graham provides a comprehensive guide for startup founders navigating phase 2 fundraising (angel to pre-Series A). The core thesis: fundraising is a means to an end, not an end itself—founders should finish it quickly and return to building product and talking to users. Drawing on YC experience, PG dissects investor psychology (fear of missing out vs. fear of investing in a flop), common founder delusions (wishful thinking when investors never explicitly say no), and offers actionable rules: don't fundraise unless you're in fundraising mode; run parallel breadth-first search weighted by expected value; accept offers greedily (take the best acceptable offer now, don't optimize for hypothetical future offers); never sell more than 25% in phase 2; undervalue how much you need initially to create momentum. The essay also covers how to get intros, handle "valuation sensitive" investors, know when to stop, and avoid the trap of raising too much or becoming addicted to fundraising.

www.paulgraham.com · 53 min
76

Do Things that Don't Scale

做不规模化的事

Paul Graham's classic 2013 essay argues that startups take off not because the product sells itself, but because founders do unscalable things to initiate growth. Using examples from Stripe, Airbnb, Meraki, and others, he details tactics like manually recruiting users, obsessively delighting early customers, and focusing on a narrow market. He warns against the ineffectiveness of big launches and corporate partnerships. These laborious early-stage practices, though counterintuitive, compound and become part of the company's DNA. The essay is essential reading for founders and engineers interested in early-stage growth.

www.paulgraham.com · 25 min
77

How to Convince Investors

说服投资者的正确姿势：先说服自己

The key to convincing investors is to first convince yourself that your startup is worth investing in, then explain that clearly with matter‑of‑fact language. Founders need to be domain experts, understand the market, and appear formidable. Many fail because they pitch before being genuinely convinced. If rejected, candidly explaining why investors were wrong projects confidence and resonates with top investors. The time to raise money is when you can convince investors, not before.

www.paulgraham.com · 21 min
78

Startup Investing Trends

YC创始人谈创业投资趋势：成本下降、权力转移与VC的适应

Paul Graham's 2013 talk to Y Combinator investors, based on data from 564 funded startups (total valuation ~$11.7B), analyzes structural shifts in startup investing. Key theses: falling startup costs shift power to founders, investors get less equity but more quality deals; traditional Series A model demanding 20%+ equity is under market pressure — the first top-tier VC to adapt will win; angel investors can gain advantage by making decisions within 24 hours. Includes specific numbers, YC batch size adjustment details (due to an n² algorithm bottleneck), and falsifiable predictions. For readers interested in venture capital and early-stage investing trends.

www.paulgraham.com · 17 min
79

The Hardware Renaissance

保罗·格雷厄姆：硬件复兴正当时，别让投资者偏见拦路

Based on the latest Y Combinator batch, Paul Graham reports a surge in hardware startups: 7 out of 84 companies were hardware, and they outperformed the rest. He cites crowdfunding, tablets, improved motors, ubiquitous wireless, and easier prototyping (Arduino, 3D printing, etc.) as drivers. Investors are biased against hardware, but founders see the future better. Software's advantage as a business may be temporary; if shipping hardware becomes as easy as shipping software, we'd see many more hardware startups. The essay ends with an encouragement: don't be deterred by investor prejudice, and especially apply to YC with hardware ideas.

www.paulgraham.com · 2 min
80

Black Swan Farming

黑天鹅捕手：风险投资的反直觉本质

Paul Graham reflects on the most counterintuitive aspects of startup investing after a decade at Y Combinator. The key insight: virtually all returns come from a tiny number of big winners (Dropbox and Airbnb account for ~75% of YC's portfolio value), and the best ideas initially look like bad ideas. He likens the required mindset to instrument flying in clouds—you must ignore what feels right and follow what you know intellectually. Graham argues that a Demo Day where only 30% of startups raise funding would be the correct risk profile, yet the psychological pressure to avoid visible failure keeps the actual rate far higher. The article is a masterclass in decision-making under power-law distributions, relevant not only to investors but to anyone navigating high-variance outcomes.

www.paulgraham.com · 12 min
81

Startup = Growth

增长：创业公司的唯一核心

Paul Graham argues that a startup is a company designed to grow fast, and growth is the only essential thing. He explains how growth drives every aspect: from the definition (a barbershop isn't a startup because it can't scale) to decision-making (use a target growth rate as a compass), to the economics (high risk, high expected value). He provides concrete metrics: during Y Combinator, a good growth rate is 5-7% per week; at 5% weekly, a company grows 12.6x per year. He also covers why VCs prefer startups (capital gains over dividends, ease of oversight) and why acquisitions happen (fast growth is both valuable and dangerous). The essay encourages founders to focus relentlessly on growth as an optimization problem.

www.paulgraham.com · 31 min
82

How to Get Startup Ideas

创业点子不是想出来的：Paul Graham 谈如何发现需求

Paul Graham argues that the best startup ideas come not from brainstorming but from noticing problems, especially ones you yourself have. Drawing on examples from Microsoft, Apple, Google, and Facebook, he explains that top ideas share three traits: founders want the product, can build it, and few others see its worth. The ideal demand shape is a 'well'—a small group of users who need it urgently, not a large group with lukewarm interest. He advises living at the leading edge of a fast-changing field, building what's interesting, and turning off filters like 'unsexy' and 'schlep.' The essay offers practical tips for students and those in need of an immediate idea, all grounded in Y Combinator's startup experience. It's a must-read for engineers and product thinkers aiming to start a company.

www.paulgraham.com · 41 min
83

The Top of My Todo List

待办清单顶端的五条人生律令

In this 2012 essay, Paul Graham reflects on the five biggest regrets of the dying, as compiled by a palliative care nurse. He recognizes the pattern of errors of omission—mistakes made by default. To avoid them, he inverts the regrets into five commands: don't ignore your dreams; don't work too much; say what you think; cultivate friendships; be happy. He places these at the top of his todo list as a daily reminder. The piece is brief and personal, emphasizing the importance of deliberate living.

www.paulgraham.com · 1 min
84

How Y Combinator Started

Y Combinator 创立始末：一次散步如何催生创业加速器

On Y Combinator's 7th anniversary, Paul Graham recounts its founding in March 2005 after a walk with Jessica Livingston. They started with $200k, intending to standardize seed funding. The first batch was a summer program for undergrads, initially not taken seriously, but the founders' performance impressed everyone. This led to the synchronous batch model that became YC's signature. The essay details early skepticism, the naming, and the move to Silicon Valley, offering a personal view on the origins of the influential accelerator.

www.paulgraham.com · 8 min
85

Defining Property

财产的边界：当数字气味无法被锁在瓶子里

Paul Graham uses an ancient Japanese story about a student 'stealing' cooking smells as an analogy for today's copyright wars. He argues that property is not a fixed concept but depends on what 'works'—i.e., what can be enforced without warping society. The internet has made digital data flow freely like smells, while record labels and movie studios try to sue and lobby to maintain an obsolete model. Graham criticizes their legal tactics and concludes that societies must gracefully adapt to the accelerating changes in property definitions driven by technology. The essay is philosophical and policy-oriented, with no technical engineering content.

www.paulgraham.com · 6 min
86

A Word to the Resourceful

一点就通：为何优秀创始人总是一点就透

Paul Graham observed that the most successful YC startups had founders who were 'resourceful' — able to take a small lead and independently close it, whether for funding, users, or ideas. In contrast, the least successful founders were 'hard to talk to': not unintelligent, but subconsciously resistant to following inconvenient implications. This conversational resourcefulness is a critical trait. The essay illustrates with YC anecdotes and a partner’s observation, offering insight into effective communication and decision-making in startups. Originally published in 2012, it remains relevant for entrepreneurs and engineers.

www.paulgraham.com · 4 min
87

Schlep Blindness

那些你视而不见的脏活累活

Paul Graham introduces the concept of 'schlep blindness', the unconscious tendency to ignore startup ideas that involve tedious, unpleasant tasks. He uses Stripe as the prime example: for years every hacker knew how painful online payments were, yet no one tried to fix them because the idea required dealing with banks, fraud, and regulations. This makes ambitious problems doubly valuable—they are like undervalued assets, avoided by most founders. PG offers two antidotes: ignorance (young founders don't know what they're getting into, and that's an advantage) and perspective-shifting (ask 'what problem do I wish someone else would solve for me?'). The essay argues that plenty of broken infrastructure remains unnoticed precisely because its fix requires facing unpleasant work.

www.paulgraham.com · 5 min
88

Snapshot: Viaweb, June 1998

回顾 Viaweb：1998 年 6 月的创业快照与早期网络技术

A retrospective look at Viaweb's website snapshot taken hours before the Yahoo acquisition announcement in June 1998. Paul Graham recounts the technical and business realities of the early web: tiny screens, text rendered as images due to lack of antialiased fonts, $16,000/month PR spend to get press coverage, relying on only seven search engines (Google was months away), struggling with Cybercash online payments, and hosting all stores on six tower servers in a co-founder's office with just two T1 lines serving over 10 million page views per month. The article also covers RTML (a Common Lisp‑based template language), fake version numbers for the trade press, and the Shopfind shopping search engine. Rich with first‑hand details and technical lore.

www.paulgraham.com · 5 min
89

Frighteningly Ambitious Startup Ideas

保罗·格雷厄姆：最可怕的创业点子，正是最大的机会

Paul Graham argues that the most ambitious startup ideas are the most frightening and thus often ignored. He lists examples: a new search engine that wins over top hackers, replacing email with a todo list protocol, breaking the university monopoly on learning, internet-native drama, the next Apple, bringing back Moore's Law via automatic parallelization, and ongoing medical diagnosis. The key insight is that such ideas seem too big and threaten one's identity. He advises startups to approach them obliquely: start with something small and deceptively harmless, and only reveal the grand vision after it's a fait accompli.

www.paulgraham.com · 21 min
90

Writing and Speaking

好思想是写作的灵魂，却是演讲的点缀

Paul Graham reflects on the fundamental difference between writing and speaking: having good ideas is central to good writing, but plays an alarmingly small role in speaking. He recalls a conference where a charismatic speaker left the audience roaring with laughter, but a transcript would have revealed he said very little. Graham argues that the demands of engaging an audience—eye contact, ad‑libbing, flattery, jokes—actively work against careful thought. Prewritten talks inhibit connection, while impromptu speaking limits reflection. Audiences compound the problem, as collective mood degrades individual judgment. Though talks are inferior for conveying ideas, they excel at making distant figures feel accessible and motivating action. Ultimately, Graham cares less about becoming a better speaker than about having better ideas—the real foundation of writing well. Suitable for engineers interested in the nature of expression.

www.paulgraham.com · 7 min
91

The Patent Pledge

Paul Graham 提出“专利承诺”：不对少于25人的公司首先使用软件专利

In 2011, Paul Graham proposed a non-legally-binding industry self-regulation: tech companies should publicly pledge not to be the first to use software patents against startups with fewer than 25 employees. He argued that this downstream approach could reduce the abuse of patents to suppress small competitors without waiting for government reform. Companies making the pledge would attract better talent, while those refusing would stand out as litigators rather than innovators, repelling employees and investors. The pledge doesn’t stop patent trolls but targets the root of innovation suppression, and individuals can ask employers where they stand.

www.paulgraham.com · 4 min
92

The Emails That Show How VCs Almost Missed Airbnb

风投几乎错过 Airbnb：内部邮件记录

Paul Graham publishes the private email thread with Fred Wilson from 2009, showing how YC tried to convince an investor about Airbnb when it was still called AirBed&Breakfast. The exchange reveals Wilson’s initial skepticism, internal partner debates where junior members were enthusiastic but “three old guys” didn’t get it, and Graham’s persistent arguments: the YC batch poll landslide, the counter-cyclical nature, and the path toward hotels. Real data points like founders funding themselves via cereal and users renting to pay their own rent are included. This behind-the-scenes look is a rare glimpse into how early-stage investors almost miss a massive opportunity and how conviction is built.

www.paulgraham.com · 8 min
93

Tablets: The Next Wave of Ephemeralization

平板时代：去物质化的下一波浪潮

In this 2010 essay, Paul Graham argues that tablets, like the iPad, will replace many specialized physical devices by turning their functions into software apps. He introduces Buckminster Fuller's concept of ephemeralization—the trend of replacing physical machinery with information. Examples include GPS devices, music players, keys, and even scales. He advises startups to look for objects that can be made obsolete by a tablet app. He also expresses concern about Apple's potential dominance and suggests that competitors should become better platforms for ephemeralization.

www.paulgraham.com · 3 min
94

Why Startup Hubs Work

创业圣地之所以存在：密度、文化与意外邂逅

In this classic 2011 essay, Paul Graham frames startup hubs not as places that kill startups less, but as places that offer an antidote to startups' natural tendency to fail. The antidote has two components: a social environment where starting a startup is seen as cool, and chance meetings with people who can help. Both are driven by density of startup people—enough density to generate random helpful encounters, supply diverse needs of different startups, and shift social norms toward optimism. Graham also explains Y Combinator's role as a 'Valley within the Valley,' artificially amplifying that density. Essential reading for anyone building or funding early-stage startups.

www.paulgraham.com · 10 min
95

What We Look for in Founders

保罗·格雷厄姆：我们寻找创始人时看重的五种品质

Paul Graham outlines five key qualities Y Combinator looks for in founders: determination, flexibility, imagination, naughtiness, and friendship. Intelligence matters, but beyond a threshold, grit trumps all. Founders must pivot without losing resolve—like a running back. Imagination yields ideas that seem just crazy enough to work. A piratical streak helps bend inconsequential rules. Strong co-founder bonds resist the startup's inevitable strain. Examples include WePay, Greplin, Airbnb, and Justin.tv.

www.paulgraham.com · 4 min
96

High Resolution Fundraising

高分辨率融资：创业公司如何用可转债打破投资者僵局

Paul Graham explains why startups increasingly use convertible notes in angel rounds: they allow different prices for different investors, breaking the deadlock where everyone waits to see who else invests. Convertible notes enable 'high-resolution fundraising' by using valuation caps, rewarding bolder investors with lower effective valuations while later investors pay more. This also frees startups from picking a fixed round size upfront, letting them adjust based on investor reactions. The essay predicts markets evolve toward higher resolution pricing and that this model will favor decisive investors over followers.

www.paulgraham.com · 4 min
97

Where to See Silicon Valley

硅谷地理志：九个地标看透科技腹地

Paul Graham's 2010 guide to Silicon Valley's nine key locations—not tech itself, but the physical and social infrastructure that makes innovation possible. From Stanford's climate and campus layout, to cafes on University Ave where founders pitch investors, the Lucky Office at 165 University Ave that housed Google, Paypal, and Wepay, the VC maze of Sand Hill Road, the shift from Palo Alto to Mountain View's Castro Street, Google's utopian campus, the redwoods and fault lines along Skyline Drive, and the choice between Highway 101 and 280. Each spot reveals a spatial logic: restaurants near offices retain talent, expensive Palo Alto real estate ($1000/sq ft) signals post-exit wealth, and redwoods mark where coastal fog comes inland. It's anthropology of the startup world as much as geography.

www.paulgraham.com · 6 min
98

The New Funding Landscape

创业融资新格局：超级天使与VC的角力

In this 2010 essay, Paul Graham analyzes the transformation of startup fundraising from a bipolar angel-VC world to one disrupted by 'super-angels' — lightweight, fast-moving funds that fill the gap between small angel checks and large VC rounds. He explains why super-angels emerged (market demand for ~$400k rounds), how they force VCs to compete on speed and valuation, and the trade-offs for founders: less dilution and more control in angel rounds, but potentially less hands-on support. Graham notes super-angels may favor quick exits over home runs, and predicts a short-term founder-friendly landscape of fast closes and high valuations, with eventual convergence of the two investor types. Essential historical context for anyone studying startup finance evolution.

www.paulgraham.com · 20 min
99

What Happened to Yahoo

雅虎衰落探因：短视收入与身份错位

Paul Graham draws on firsthand experience at Yahoo to diagnose two fatal flaws. First, Yahoo profited handsomely from banner advertising—a bubble sustained by brand advertisers and a Ponzi scheme among internet startups—so it ignored the value of search, even after Revenue Loop demonstrated the potential. Search was only 6% of traffic and executives saw no need to improve. Second, Yahoo insisted it was a 'media company,' marginalizing programmers as mere coders of product specs, which repelled top talent and triggered a death spiral of technical mediocrity. Founders and managers should note: any company needing good software must build a hacker-centric culture.

www.paulgraham.com · 12 min
100

The Acceleration of Addictiveness

成瘾的加速

Paul Graham argues that technological progress concentrates both good and bad things, making addictive substances and behaviors more potent. He observes that the process is accelerating, leading to more addictive products (e.g., Facebook, World of Warcraft). As society's customs take too long to adapt, individuals must invent personal strategies—like avoiding iPhone or taking long hikes—to stay productive and sane.

www.paulgraham.com · 8 min
101

The Future of Startup Funding: Founder-Led Rounds and the Rise of Super-Angels

创业融资的未来：创始人主导融资与超级天使的崛起

Paul Graham analyzes the growing disconnect between VCs' need for large investments and startups' need for smaller amounts, leading to the rise of 'super-angels' (mini VC funds). He predicts founder power will increase, shifting fundraising toward founder-led, fast, and simple rounds: rolling closes, standardized paperwork, and less reliance on investor herd behavior. Super-angels are the invaders; VCs must adapt by taking fewer board seats. The essay uses YC experience to forecast that rounds will close faster, traction requirements will diminish, and investors may ultimately benefit from ceding control.

www.paulgraham.com · 22 min
102

The Top Idea in Your Mind

警惕你头脑中最重要的那件事

Paul Graham argues that at any given time, everyone has a single 'top idea' in their mind that their thoughts naturally drift toward when left idle. This idea gets the full benefit of unconscious thinking, while others starve. If the top idea is something undesirable (e.g., fundraising, disputes), it hijacks your creativity and productivity. The solution is to indirectly control what becomes critical to you, avoiding money matters and disputes. He illustrates with personal startup fundraising experiences and Newton's distraction by criticism. Practical advice for knowledge workers.

www.paulgraham.com · 7 min
103

Organic Startup Ideas

最好的创业想法来自你自己的生活

There are two types of startup ideas: organic ones that grow out of your own life, and made-up ones aimed at others. Organic ideas, like the Apple I and Facebook, tend to be more successful—they often start as personal projects. Young founders are in the best position to find organic ideas because they use the latest tech and can spot fixable brokenness early. The essay advises focusing on ideas rather than the startup label, fixing small annoyances, and not worrying if your product is dismissed as a toy. It's a timeless read for anyone building something new.

www.paulgraham.com · 6 min
104

How to Lose Time and Money

如何损失时间和金钱

In this 2010 essay, Paul Graham argues that the most common way to lose money is not excessive spending but bad investments, and similarly, the most dangerous way to lose time is not having fun but doing fake work. He notes that while luxuries trigger alarms, investments bypass them, just as dealing with email feels productive but wastes a day. He suggests developing new alarms to detect these subtle traps that mimic virtue. The piece offers a concise parallel between financial and temporal prudence, relevant to anyone interested in productivity and decision-making.

www.paulgraham.com · 4 min
105

Persuade xor Discover

说服与发现：写作的两种对立目标

In this essay, Paul Graham distinguishes between two purposes of writing: to persuade or to discover. He argues that most writers write to persuade, so they add rhetorical flattery to avoid offending readers, but this can hinder true discovery. He illustrates with a rewritten paragraph about labor unions, showing how the same point can be phrased to please or to confront. Graham states that he writes to discover, not to persuade, and values conciseness as a way to get closer to the ideas. This piece is for readers interested in writing methodology and intellectual honesty.

www.paulgraham.com · 8 min
106

Post-Medium Publishing

内容经济的真实逻辑：当媒介消失，什么值得付费

Paul Graham analyzes the economic structure of publishing: consumers have always paid for the medium (paper, CDs, etc.) not the content. As digital technology erases the medium, the industry is left with nothing to sell. He argues that selling pure content to consumers is historically difficult (exceptions like Bloomberg terminals serve business needs), and that iTunes works as a tollbooth rather than a store. The future lies in indirect monetization (ads, concerts) or embodying content in physical form (luxury books, cinema experiences). A sobering look at the end of the content-as-product era.

www.paulgraham.com · 10 min
107

The List of N Things

列表式文章的构造与陷阱

Paul Graham dissects why listicles (e.g., '7 Secrets of Success') are so popular: they spare readers the work of understanding structure, offering an exoskeleton of numbered points that are independent, randomly accessible, and fault-tolerant. He notes the form is also writer-friendly—you can quickly sketch a framework and fill it in, reducing the risk of dead ends. However, its greatest weakness is that it leaves little room for new thought, since the title locks the topic in advance and the compartmentalized structure cannot accommodate discoveries made during writing. Graham admits using the form himself when close to a deadline and suggests that teachers should openly acknowledge that the five-paragraph essay is essentially a listicle so students learn to escape it.

www.paulgraham.com · 8 min
108

The Anatomy of Determination

决心剖析：意志力、纪律与雄心如何决定成败

Drawing on Y Combinator's early-stage investing experience, Paul Graham argues that determination—not intelligence—is the strongest predictor of startup success. He deconstructs determination into willfulness, discipline, and ambition, using a 'melon seed model' to show that willfulness and discipline must stay in balance. Discipline and ambition can be cultivated (e.g., through ambitious peers and achievements), while willfulness is more innate. A classic essay for founders, investors, and anyone curious about the real drivers of achievement.

www.paulgraham.com · 9 min
109

What Kate Saw in Silicon Valley

硅谷新人的9个意外发现：创业文化的真实面貌

Paul Graham's 2009 blog post records architect Kate Courteau's nine surprising observations after 9 months immersed in Y Combinator's startup world: how many startups fail, how much ideas change, how little money it takes, how scrappy founders are, how tech-saturated Silicon Valley is, how consistent speakers' advice is (launch fast, listen, iterate), how casual successful founders are, how important advice networks are, and how solitary the work is. The post contrasts startup culture with the 'normal' world and suggests it may be a leading indicator of future work.

www.paulgraham.com · 5 min
110

What Startups Are Really Like

创业公司的真实面目

Based on a survey of YC-funded startup founders, Paul Graham summarizes the 19 most common surprises about starting a startup. The core thesis: startups are far more different from regular jobs than most people imagine. Key findings include: cofounder relationship matters more than ability; startups consume your entire life; extreme emotional roller-coaster; persistence trumps intelligence; launch with the minimal viable product; engage users and be willing to pivot; investors are often clueless; luck plays a huge role; founders get little respect. The essay is rich with direct founder quotes, revealing both the hardship and the freedom. Suitable for aspiring founders and engineers interested in startup realities.

www.paulgraham.com · 29 min
111

The Trouble with the Segway

Segway 的麻烦：设计者的自满与用户的耻感

Paul Graham examines why Segway failed to deliver on its promise, attributing it to the rider's smug appearance: standing upright while moving makes them look too effortless. He contrasts the Segway with Trevor Blackwell's one-wheeled Eunicycle, which gets smiles, while the Segway draws shouts like 'Too lazy to walk, ya fuckin homo?' from cars. Graham argues that Segway's easy funding allowed them to work in secret, insulated from real-world feedback. A thought experiment suggests a skateboard-like stance would be cooler. The article offers insight into product design and the social acceptance of technology.

www.paulgraham.com · 2 min
112

Ramen Profitable

拉面盈利：创业公司的最低生存策略

Paul Graham defines 'ramen profitable' as a startup that generates just enough revenue to cover the founders' living expenses, typically a few thousand dollars a month. This is not traditional profitability but a tactic to buy time by minimizing costs, freeing founders from the pressure to raise money immediately. He outlines four advantages: better investor terms, proof of product-market fit, morale boost, and avoiding distraction from fundraising. The essay also warns against slipping into consulting, which does not scale. A must-read for early-stage entrepreneurs.

www.paulgraham.com · 10 min
113

Maker's Schedule vs Manager's Schedule: How Meetings Wreck Your Day

制造者时间表 vs 管理者时间表：一场会议如何毁掉你的一天

In this 2009 classic, Paul Graham distinguishes two schedule types: the manager’s schedule, cut into one-hour slots, and the maker’s schedule, which requires long, uninterrupted half-day blocks. Meetings destroy a maker’s productivity—they fragment afternoons and throw an exception into deep work. Graham shares personal tricks, like coding late at night and grouping all business tasks in the morning, and explains how Y Combinator uses clustered office hours at day’s end to protect maker time. He notes the high cost of speculative meetings and asks managers to recognize the maker’s needs. Although anecdotal, the essay resonates with engineers who need focused time for complex tasks.

www.paulgraham.com · 6 min
114

The Founder Visa

创始人签证：一个几乎零成本的经济刺激方案

Paul Graham proposes establishing a 'founder visa' to allow 10,000 foreign entrepreneurs per year to start companies in the US. The visa would prohibit holders from working for existing firms, only for their own startups. To define a startup, the government could piggyback on accredited investors' expertise: only companies funded by recognized startup investors qualify. Graham argues this policy would not take jobs from Americans but create many, at virtually zero cost. The piece is an opinion essay with no hard data, typical of PG's blog style.

www.paulgraham.com · 2 min
115

Why Twitter is a Big Deal

Twitter：不指定收件人的新消息协议

In this 2009 essay, Paul Graham argues that Twitter is a big deal because it represents a new messaging protocol where users broadcast to no one in particular. New protocols that take off are rare—only a handful like TCP/IP, SMTP, and HTTP exist. Twitter's status as a privately-owned protocol is even rarer. The founders' slow approach to monetization may have helped Twitter feel like an open protocol, boosting adoption. Aimed at readers interested in the early internet and social media dynamics.

www.paulgraham.com · 1 min
116

A Local Revolution?

创业会是局部革命吗？

Paul Graham explores whether startups represent a new economic phase and, if so, whether they will concentrate in a few hubs like the movie industry rather than spread widely like the Industrial Revolution. He argues startups are a social phenomenon that require communities of expertise, lack market demand to spread, and are hard to replicate artificially. The pull of existing hubs and random factors will likely keep the revolution localized.

www.paulgraham.com · 8 min
117

Five Founders

五位影响深远的创始人

Paul Graham lists the five startup founders who have influenced him most over the last 30 years: Steve Jobs, TJ Rodgers, Larry & Sergey, Paul Buchheit, and Sam Altman. He highlights each founder's unique impact—Jobs on design, Rodgers on candor and pragmatism, Google's founders on hacker-centric culture, Buchheit on Gmail, AdSense, and 'Don't be evil', and Altman on sheer willpower. The piece is a personal reflection without technical depth.

www.paulgraham.com · 4 min
118

Relentlessly Resourceful

不屈不挠且足智多谋：创业者的核心特质

In this short essay, Paul Graham distills the essence of a great startup founder into two words: relentlessly resourceful. Unlike the hapless who passively accept circumstances, good founders are both determined and flexible, adapting to novel difficulties by constantly trying new approaches. Graham argues this trait can be taught — especially to those who have been under authority — and serves as a practical self-test for founders and cofounder selection. Though non-technical, it offers a clear, memorable framework for anyone navigating startup challenges.

www.paulgraham.com · 6 min
119

Why TV Lost

电视为何输给计算机：开放平台、摩尔定律、盗版与社交网络

In this 2009 essay, Paul Graham explains why television lost to computers. He identifies four forces: the open platform of the Internet, Moore's Law driving bandwidth, user convenience of piracy (both free and better UX), and social applications that made everyone want a computer. The TV industry's attempts to preserve synchronicity and locality are futile; the Internet dissolves both. A prescient take on media disruption.

www.paulgraham.com · 9 min
120

Can You Buy a Silicon Valley? Maybe.

花十亿买个硅谷？或许真可以

This essay explores whether a city could manufacture a Silicon Valley by funding startups. Paul Graham argues that seed funding doesn't keep startups local because they're mobile. To anchor them, a city would need to invest at least $500k per startup, so they can resist VC relocation offers. At $1M each, $1B could draw 1,000 startups and potentially make a city the second-largest hub. The hardest part is picking winners: governments can't identify skilled investors. He proposes tracking investments by top Silicon Valley angels and offering $1M to early-stage startups to relocate. An experiment with 30 startups, given freedom and no restrictions, could test the idea. Success depends on a city's resemblance to San Francisco—weather, tolerance, universities—and the political will to try.

www.paulgraham.com · 11 min
121

Startups in 13 Sentences

Paul Graham的13条创业经验

Paul Graham distills startup wisdom into 13 concise sentences covering cofounder selection, rapid launch, iterative evolution, user understanding, prioritizing a few loving users, surprisingly good customer service, measurement, frugality, ramen profitability, avoiding distractions, morale, persistence, and deal fall-through. He argues that understanding users is the central principle. This is general startup advice, not AI or engineering-focused.

www.paulgraham.com · 7 min
122

What I've Learned from Hacker News

从Hacker News学到的社区运营经验

Paul Graham reflects on two years of running Hacker News. He argues community quality hinges on user behavior rather than user identity, applying the broken windows theory to online communities. He analyzes dilution as a behavioral problem, introduces the Fluff Principle explaining why easy-to-judge links dominate, and notes comment quality correlates with length—short comments tend to be low-effort jokes. Graham emphasizes combining technical countermeasures (e.g., killing bad comments, response delays) with human moderation (e.g., rewriting titles, banning habitual linkjackers). He also candidly discusses addiction as an unsolved problem. This piece is valuable for community managers, product managers, and engineers interested in online culture.

www.paulgraham.com · 17 min
123

How to Be an Angel Investor

天使投资实战指南：从入门到选对项目

Paul Graham shares his practical advice on angel investing based on personal experience. He argues that success depends on picking the right founders—those who are 'relentlessly resourceful'—rather than obsessing over term sheets or valuations. The essay covers how to get started (via syndicates or using standard legal docs), how to evaluate startups, and how to build a deal flow. Graham advises making small initial investments as tuition fees and emphasizes being a good person to attract better opportunities. This is a must-read for anyone considering angel investing.

www.paulgraham.com · 22 min
124

Keep Your Identity Small

少一些身份，多一些思考

In this classic 2009 essay, Paul Graham examines why discussions about religion and politics are uniquely useless. He argues that the problem is not the topics themselves, but that they become part of people's identity. Once a topic engages identity, productive debate becomes impossible—participants are by definition partisan. Programming languages, for example, can be discussed fruitfully as long as you exclude those who respond from identity. The deeper insight is that the fewer labels you have for yourself, the smarter you become. The essay is a short, sharp call to keep your identity small to think clearly.

www.paulgraham.com · 5 min
125

Could VC be a Casualty of the Recession?

VC可能成为经济衰退的牺牲品吗？

Paul Graham argues in this 2008 piece that the cost of starting a startup has plummeted due to Moore's law, open source software, free web distribution, and more powerful programming languages. Many startups can now be profitable on as little as $3,000/month in revenue, making VC funding a 'nice-to-have' rather than a 'must-have'. During the recession, even if VCs pull back, founders may keep launching companies, and VCs could become irrelevant if they stop writing checks. Evidence from Y Combinator's application surge and founder attitudes supports this divergence.

www.paulgraham.com · 8 min
126

After Credentials

后证书时代：绩效如何取代文凭

This essay explores the shift in American society from relying on academic credentials to valuing actual performance. The author traces the history of credentialism, noting its original intent to block direct transmission of power, but it gets subverted by cram schools. A better solution is an economy of more, smaller organizations, which allows direct measurement of performance, reducing the importance of credentials. The rise of startups and yuppies illustrates this trend, forcing large organizations to adapt. The essay is relevant to those interested in social trends, organizational dynamics, and talent evaluation.

www.paulgraham.com · 14 min
127

Why to Start a Startup in a Bad Economy

经济差时创业的真相：创始人比经济重要得多

Should you start a startup during a recession? Paul Graham argues that the economy is a poor predictor of startup success; what matters is the founders. Microsoft and Apple were founded in the mid-1970s, a bad economy. Technology progresses independently of the stock market, so waiting is costly. Startups should run cheaply (be cockroaches) to survive. Investors are reluctant in bad times, but founders should invest their effort now. The key is the team, not the timing.

www.paulgraham.com · 6 min
128

The Pooled-Risk Company Management Company

把收购看作风险汇集型公司管理公司

In this classic essay, Paul Graham responds to David Heinemeier Hansson's advice that startup founders should build profitable companies and live off revenues rather than hope for an acquisition. Graham argues that the ultimate goals of founders are freedom and security — ideally, you could hand your company to a professional management firm and still receive income. He then reveals that such a firm already exists: a public company acquiring your startup. By trading your company for shares in a larger entity, you gain professional management and pooled risk. The catch is that the acquisition market is fickle (like Mr. Market), so you must be profitable and patient to get the average outcome. The essay concludes that building a revenue-generating company and selling it are not opposites; the latter is often the optimal case of the former.

www.paulgraham.com · 7 min
129

A Fundraising Survival Guide

融资生存指南：创业者如何扛过最残酷的第二关

Paul Graham, a Y Combinator partner, shares first-hand survival techniques for startup fundraising, which he calls the second hardest part of starting a startup. He explains why the process is brutal—inefficient markets, indecisive investors, and the danger of demoralization—and offers nine concrete tactics: lower expectations, keep building product, be conservative, be flexible about amount, achieve ramen profitability, handle rejection analytically, consider consulting as a backup, avoid inexperienced investors, and constantly gauge where you stand. The essay includes specific numbers ($140/month/user, $2K/month ramen profitability) and the Viaweb anecdote. Though from 2008, its core insights on investor psychology remain relevant for today's tech founders.

www.paulgraham.com · 28 min
130

Disconnecting Distraction

Paul Graham 谈分心：工作电脑永不联网的极端实验

In this 2008 essay, Paul Graham describes how distractions fuel procrastination and how technology evolves to become more addictive, like drug-resistant bacteria. He recounts slipping into compulsive Internet use and eventually adopting a radical solution: a dedicated web-only laptop across the room, while his main work computer stays offline. The physical distance made him much more aware of time spent online, limiting it to about an hour a day and restoring focus. He later updated the essay to note that this strategy also eventually failed, underscoring the ongoing struggle against distraction. Suitable for knowledge workers seeking practical tactics to regain concentration.

www.paulgraham.com · 6 min
131

Cities and Ambition

城市与抱负

Great cities attract ambitious people, but each sends a distinct message: New York urges you to make more money, Cambridge to be smarter, and Silicon Valley to have more impact. Using historical evidence (e.g., Leonardo’s Milan vs. Florence), personal anecdotes, and data, Paul Graham argues that physical environment strongly influences achievement. Cities shape ambition through subtle cues like overheard conversations and evening window sights, and chaotic fields like arts and technology benefit most from urban peer density. He advises young, unfocused ambitious people to try living in several cities to discover which ambition resonates. The essay offers a timeless reflection on place, culture, and personal drive.

www.paulgraham.com · 20 min
132

Why There Aren't More Googles

为什么没有更多谷歌：VC保守主义扼杀创新

In this 2008 essay, Paul Graham argues that the real reason there aren't more Googles is not that startups sell out too early, but that venture capitalists are too conservative to fund truly innovative ideas. He points out that Google and Facebook were willing to be acquired, but acquirers lowballed them. VCs, driven by consensus and fear of risk, miss the most outlier ideas. As the cost of starting a startup has plummeted, a funding gap has emerged: startups need $250-500k, but VCs want to invest $2M+. Graham calls for either VCs to adapt by making smaller, more numerous bets, or new investors to fill the gap. The essay offers insights into startup dynamics and investment psychology, though it is not technical.

www.paulgraham.com · 8 min
133

Be Good

善意驱动：创业公司的隐秘优势

In this essay, Paul Graham argues that startups should operate like charities: focus on making something people want, not on making money early. Using examples like Craigslist, Google's early days, and Microsoft's initial benevolence, he shows that being good provides three key advantages: boosts morale (users become a 'tamagotchi' that keeps you going through tough times), attracts help from others (investors, users, and top hackers who are idealistic), and serves as a stateless decision compass (just do what's best for users). He suggests 'don't be evil' is not enough; aim to be good as a practical strategy, not a moral stance. This is a classic piece for tech entrepreneurs.

www.paulgraham.com · 17 min
134

How to Disagree

分歧的艺术：从辱骂到驳斥核心点的七级阶梯

Paul Graham's 2008 essay proposes a disagreement hierarchy (DH0–DH6) ranging from name-calling to refuting the central point. Each level describes the form of a rebuttal, not its correctness: low levels (DH2 and below) are never convincing, while high levels (DH4 and above) can be. Most intellectual dishonesty is unintentional; awareness of the hierarchy helps debaters improve and reduces meanness. Useful for anyone engaging in online debate or technical commentary.

www.paulgraham.com · 9 min
135

Some Heroes

英雄谱：保罗·格雷厄姆的个人影响力清单

In this 2008 essay, Paul Graham lists his personal heroes, including Jack Lambert, Kenneth Clark, Larry Mihalko, Leonardo da Vinci, Robert Morris, P.G. Wodehouse, Alexander Calder, Jane Austen, John McCarthy, the Spitfire, Steve Jobs, and Isaac Newton. He identifies two common qualities: an almost excessive devotion to their work and absolute honesty (never pandering). The essay is a personal reflection on those who influenced him, not an objective ranking. Readers interested in Paul Graham's mindset may find it insightful.

www.paulgraham.com · 15 min
136

A New Venture Animal

Y Combinator：创业生态的新物种

In this 2008 essay, Paul Graham argues that Y Combinator is not merely a seed fund but a new species of venture firm that acts as a 'steam catapult' for startups. YC focuses on the earliest stage, providing intensive help (co-founder issues, product focus, legal setup) rather than just money. Using biological scaling analogies, he explains why early-stage startups need proportionally more assistance and less capital. YC takes 7% equity, and founders need only improve their next round valuation by 7.5% to break even—a threshold YC easily exceeds.

www.paulgraham.com · 11 min
137

Lies We Tell Kids

成年人向孩子撒谎的全面剖析：保护、控制与成长的代价

In this 2008 essay, Paul Graham examines the many ways adults lie to children, from protective cocooning (suburbia) to concealing sex, death, religious identity, and school authority. He argues that these lies, while often well-intentioned or aimed at keeping peace, distort children's understanding of reality and create a 'truth debt' that persists into adulthood. Graham uses personal anecdotes and logical reasoning to urge adults to consciously unlearn the lies they were told as kids. Suitable for readers interested in education, psychology, and social observation.

www.paulgraham.com · 29 min
138

You Weren't Meant to Have a Boss

你不该有老板：大公司对程序员是毒药

Paul Graham argues that large organizations are inherently toxic for programmers, using the analogy of zoo lions vs wild lions and comparing corporate jobs to junk food. He claims humans evolved to work in small groups (8-20 people), and the tree structure of big companies compresses individual freedom inversely with size. For programmers, whose essence is to build new things, this suppression kills creativity and learning. Graham advises programmers to start their own startups or work for small companies, even if they fail, as the experience is more natural and rewarding. The essay is philosophical and based on anecdotal evidence, lacking technical depth.

www.paulgraham.com · 14 min
139

Trolls

Trolls：网络喷子的四个成因与社区治理实验

Paul Graham reflects on a Hacker News comment about trolls, analyzing four causes: the distance of anonymity, the social awkwardness common in tech crowds, the incompetence of insult over argument, and forum culture. He shares practices from News.YC (the early Hacker News) to curb trolling: explicit guidelines against rude behavior, community self-policing, ruthless bans, and a karma system that makes reputation loss visible. He describes a 'Gresham's Law of trolls' where trolls drive away thoughtful users, and notes that at 8,000 daily uniques, the forum's tone remains good, but scaling is the real test. The essay is a firsthand account of running a technical community and resisting its decay.

www.paulgraham.com · 5 min
140

Six Principles for Making New Things

制造新事物的六条原则

Paul Graham distills six principles that have guided his work on Viaweb, Y Combinator, Arc, and his essays: find simple solutions to overlooked problems that actually need solving, deliver them informally, start with a crude version 1, and iterate rapidly. He explains why this approach always invites initial contempt but ultimately wins. A general philosophy for creating new things, not a technical deep dive.

www.paulgraham.com · 7 min
141

Why to Move to a Startup Hub

创业公司为何应迁往创业中心

In this 2007 essay, Paul Graham argues that startups should move to a startup hub like Silicon Valley for better odds of success. He uses the example of Facebook, which moved after Boston VC passed, and contrasts the aggressive, knowledgeable Silicon Valley investors with more conservative Boston ones. He acknowledges exceptions (family, immigration, industry ties) but maintains that the density of expertise, deal flow, and bold capital in hubs provide a decisive advantage. Written for entrepreneurs and those interested in regional startup dynamics.

www.paulgraham.com · 8 min
142

The Future of Web Startups

Web创业的廉价化与生态变迁

In 2007, Paul Graham predicted that as web startup costs drop to just courage, the number of startups will explode, leading to standardization, simplified acquisitions, riskier strategies, younger/nerdier founders, persistent startup hubs, better judgement by investors, college changes, more competition, and faster innovation. Based on Y Combinator's experience, the essay offers no hard data but a coherent vision. Relevant for those interested in startup trends and tech business history, but not aligned with this publication's focus on AI engineering and cutting-edge technology.

www.paulgraham.com · 20 min
143

How Not to Die: A Startup Survival Guide

如何不死：创业公司生存指南

Paul Graham's talk at the final Y Combinator dinner argues that startups succeed by simply not dying. He identifies common causes of death: demoralization, distraction, and founders quitting. Advice includes staying in regular contact with YC (using social pressure as a driver), iterating until you find a core of users who love your product, and avoiding qualifiers like 'but we're going to keep working on the startup.' He notes that public commitment (e.g., being featured in news) makes giving up unthinkable. The tone is blunt and practical, emphasizing persistence over brilliance.

www.paulgraham.com · 11 min
144

How to Do Philosophy

哲学的困境与出路：从语言陷阱到实用性真理

Paul Graham reflects on his philosophy education and critiques the discipline for being mired in verbal confusions. He argues that Aristotle's separation of theoretical knowledge from utility was a critical mistake, leading to centuries of unproductive speculation. He proposes a practical approach: start with useful observations and gradually generalize, avoiding the trap of meaningless abstraction.

www.paulgraham.com · 28 min
145

News from the Front

来自前线的报告：名校光环并不像你想的那样重要

Drawing from Y Combinator's experience funding startups, Paul Graham argues that where someone goes to college matters far less than commonly believed. The market test of startup success shows no correlation with alma mater prestige. He criticizes corporate hiring's reliance on brand-name schools and advises focusing on learning and doing rather than grades and college admissions. A provocative essay on meritocracy and education's role in real-world achievement.

www.paulgraham.com · 13 min
146

Holding a Program in One's Head

把程序装入头脑：程序员深度工作的艺术

Paul Graham argues that the best programmers can hold an entire program in their mind, like a mathematician holds a problem, enabling deep understanding and flexible manipulation. He offers eight practices: avoid distractions, work in long stretches, use succinct languages (especially via bottom-up programming), keep rewriting, write rereadable code (dense for yourself), work in small groups, avoid multiple people editing the same code, and start small. He notes that typical organizational practices—meetings, multiple authors, specs-first—undermine this state, and great programmers often rebel against them. For startups, exploiting this mismatch can be a competitive advantage against large companies.

www.paulgraham.com · 11 min
147

An Alternative Theory of Unions

工会的另一种解读：不过是增长期的超支泡沫

Paul Graham argues that the high-paying union manufacturing jobs of the mid-20th century were not a triumph of the labor movement but a temporary bubble created by rapidly growing companies that prioritized speed over cost. He compares it to the overpayment for web consultants during the Internet Bubble. When growth slowed, companies cut costs and unions shrank. The article rejects moral decline narratives in favor of economic cycle explanations.

www.paulgraham.com · 3 min
148

Stuff: The Burden of Excess Possessions

拥有即负担：论物质过剩与消费主义陷阱

Paul Graham reflects on the overvaluation of material possessions in modern life. He argues that stuff has become a burden rather than an asset, citing his own experience of living with just a backpack for a year. He contends that most purchases provide little real value and that clutter is mentally draining. He advises asking whether an item will be used constantly and considering the overall cost of ownership. While insightful, this essay is personal philosophy, unrelated to AI or engineering.

www.paulgraham.com · 7 min
149

The Equity Equation

股权交易公式：如何判断融资或招人是否划算

Paul Graham presents a simple equity equation 1/(1-n) for founders to evaluate whether accepting investment, hiring an employee, or making a deal is worth giving up a percentage of the company. The rule: a deal is good if it improves the company's average outcome by more than 1/(1-n). Examples include Y Combinator's typical 7% deal and Sequoia's 30% deal. The article also derives the formula for granting stock to employees (n=(i-1)/i) and discusses adjusting for salary and overhead. While real decisions require judgment, the equation provides a rational framework.

www.paulgraham.com · 6 min
150

Two Kinds of Judgement

两种评判：别把非个人的拒绝太当真

Paul Graham distinguishes two types of judgment: one where accurate evaluation is the goal (exams, court cases), and another where judgment is merely a means to select a set (college admissions, hiring, dating). The latter is not about you personally. Recognizing this helps depersonalize rejection and encourages proactive behavior rather than passive acceptance.

www.paulgraham.com · 4 min
151

Microsoft is Dead

微软已死：保罗·格雷厄姆2007年的技术历史论断

In 2007, Paul Graham declared Microsoft dead. He argued that Google's rise, Ajax making web apps viable, broadband Internet, and Apple's comeback (OS X, iPod) ended the desktop era. Microsoft's monopoly was inherited from IBM and lasted from the 1950s to around 2005. Graham suggests Microsoft could buy Web 2.0 startups to recover but doubts they will. A historical tech commentary, not an engineering piece.

www.paulgraham.com · 7 min
152

The Hacker's Guide to Investors

黑客融资指南：投资者世界的秘密与策略

Paul Graham distills 23 brutally honest lessons about the investor world for hacker-founders. Key insights: angels are the critical fuel for startups, VCs are momentum investors who follow crowds not tech, valuations are fiction, high valuations hurt exit chances, fundraising destroys product momentum, and investors fear looking bad more than losing money. Graham advises staying cockroach-like—cheap, urgent, with a backup plan—and identifies an opportunity for a new kind of fast-decision fund.

www.paulgraham.com · 35 min
153

Is It Worth Being Wise?

智慧与智力：两种不同的人生算法

In this 2007 essay, Paul Graham distinguishes wisdom from intelligence: wisdom means knowing what to do in most situations (high average), while intelligence means excelling in a few (high peaks). As knowledge specializes, the gap grows, forcing a trade-off. Wisdom comes from discipline and humility; intelligence from curiosity and selective self-indulgence. Ancient societies valued wisdom because most work involved choosing among alternatives; modern creative work has no upper bound, leading to discontent. Graham warns against applying wisdom's standards to intelligence, and urges acceptance of frustration inherent in unbounded work.

www.paulgraham.com · 22 min
154

Learning from Founders

初创公司才是真正高产的地方：Paul Graham 眼中的创业真相

In the foreword to Jessica Livingston's 'Founders at Work', Paul Graham contrasts the real productivity of startups with the superficial professionalism of big companies. He argues that the earliest phase of a startup is the most productive, resembling a Formula 1 car rather than a sedan with fake spoilers. The book aims to show what actual high-output work looks like, urging corporations to aspire to startup-like efficiency instead of startups mimicking corporate conformity. This essay is suitable for founders, managers, and anyone interested in organizational effectiveness.

www.paulgraham.com · 5 min
155

Why to Not Not Start a Startup

你为什么还在犹豫创业？——PG逐一击破十六个借口

Paul Graham draws on Y Combinator data to dissect sixteen common reasons people hesitate to start startups, arguing most are misconceptions or solvable through action. Highlights: out of YC's first eight startups, at least four succeeded; zero percent regretted the experience. Real barriers include family obligations and lack of cofounder. A data-driven, honest pitch for entrepreneurial risk-taking, aimed at technical talent.

www.paulgraham.com · 35 min
156

How Art Can Be Good

艺术何以有优劣？

Paul Graham recounts his shift from believing taste is purely subjective to recognizing objective standards for art. He argues that art's purpose is to interest its audience, and since humans share many psychological and biological traits (e.g., face recognition, 3D perception), what makes art good is not random. However, brand recognition, artistic tricks, and social conformity distort judgment, making direct voting unreliable. He suggests traveling widely in time and space and learning to spot tricks to approach a more universal taste. The essay concludes that acknowledging good art exists empowers artists to strive for greatness. Suitable for readers interested in aesthetics and the philosophy of art.

www.paulgraham.com · 20 min
157

Copy What You Like

只模仿你真正喜欢的

In this personal essay, Paul Graham recounts his youthful mistakes of imitating the wrong models: bad short stories in high school, vacuous philosophy papers in college, and the faddish expert systems in grad school. He concludes that one should only copy what one genuinely likes, not what is admired by authorities or trends. He offers two tricks to distinguish true liking from mere impression: ignore presentation (e.g., museum lighting) and examine your guilty pleasures—what you read when you don't need to feel virtuous. He also warns against copying flaws while trying to copy virtues, using the example of Renaissance painters whose dirty varnish led imitators to use brownish colors.

www.paulgraham.com · 6 min
158

The 18 Mistakes That Kill Startups

创业公司的18个致命错误：Paul Graham经典复盘

Paul Graham identifies 18 avoidable mistakes that kill startups, centered on failing to build something users want. He covers founder dynamics, location, market scope, execution, technology choices, fundraising, and team conflicts, emphasizing learning from users, rapid iteration, and full commitment over perfect planning. While dated 2006, its insights remain relevant for early-stage founders.

www.paulgraham.com · 32 min
159

The Island Test

孤岛测试：找出你真正上瘾的东西

Paul Graham proposes a thought experiment: if you were stranded on a small island off the coast of Maine for a weekend, with no shops and no way to leave, what would you make a point of packing besides clothes? That item is what you're addicted to. His own list: earplugs (for guaranteed quiet to think), a notebook and pen (to wick ideas rather than archive them), and books (as distraction insurance against boredom). He admits his dependence on books is not virtuous – they are jam when he should be eating fruit. He recalls a hike where he carried no books and found that having ideas could entertain himself. The essay is a personal reflection on habits and the tools of creativity.

www.paulgraham.com · 4 min
160

A Student's Guide to Startups

学生创业指南：毕业前后的分水岭

Paul Graham offers a candid guide for student entrepreneurs, drawing on his Y Combinator experience. He explains why starting a startup after graduation—when you lose the escape hatch of being a student—provides stronger motivation than starting in college. He outlines young founders' advantages (stamina, poverty, rootlessness, peer pool, ignorance) and disadvantages (building class-project-like products, lacking appreciation of the work-money equation). His core advice: leverage school to find co-founders; learn by building things for real users; ignore most classes and books. The key is internalizing the brutal necessity of avoiding 'drowning' in the real world.

www.paulgraham.com · 36 min
161

How to Be Silicon Valley

硅谷配方：人才、大学与有机生长

In this classic 2006 talk, Paul Graham examines what it takes to reproduce Silicon Valley elsewhere. He argues that only two kinds of people are essential: rich investors (with tech experience) and nerds (technical talent). A first-rate university acts as a magnet to attract nerds, but it must be located in a city that appeals to the young—places with personality, tolerance, an intact center, and a walkable, non-suburban feel. Bureaucrats or large-scale developments cannot substitute for organic growth: startups beget startups, creating a self-sustaining chain reaction. Graham uses examples from Shockley Semiconductor to Google to illustrate the cycle, and identifies Portland and Boulder as the most promising candidates for new silicon valleys. This essay is valuable for those interested in startup ecosystems, urban planning, and technology policy.

www.paulgraham.com · 21 min
162

Why Startups Condense in America

美国为何成为创业公司的聚集地——Paul Graham 2006年经典分析

In this classic 2006 essay, Paul Graham examines why startups concentrate in America, particularly Silicon Valley. He lists ten advantages: immigration that welcomes talent, the country's wealth, political freedom (not a police state), superior universities, flexible labor laws, a cultural separation between work and employment, tolerance of marginal business practices, a large domestic market, abundant venture capital (especially angel investors), and a dynamic career system that allows late decisions. He also suggests how other nations could outperform the US, such as lower capital gains tax and smarter immigration policies. While dated, the analysis remains insightful for understanding startup ecosystems.

www.paulgraham.com · 28 min
163

See Randomness

别把一切都当成有意的——看见随机性

Paul Graham extends a line of Western philosophy, arguing that humans have a natural tendency to see purpose and malice in events, but the better principle is to see randomness. From Socrates' teleology to Dawkins' selfish gene, we have gradually abandoned the assumption that everything revolves around us. This principle applies both to grand ideas and everyday annoyances—like a housemate eating your cake. The essay offers a framework for critical thinking without technical details.

www.paulgraham.com · 3 min
164

A Forgotten Search Patent: Revenue Loop

一篇被遗忘的搜索排序专利：Revenue Loop的故事

Paul Graham discovers he was granted a patent in 2003 for a search algorithm that ranks results by bid price multiplied by transaction count, effectively combining relevance and revenue. He recounts a 1998 meeting with Yahoo where the idea was dismissed because advertisers were already overpaying. The patent eventually issued, but lawyers mangled his clear prose into incomprehensibility. A historical look at early search-ad optimization.

www.paulgraham.com · 4 min
165

The Hardest Lessons for Startups to Learn

创业者最难学会的七堂课

Paul Graham's classic talk distills seven counterintuitive lessons for startup founders: release early, keep pumping out features, make users happy, fear the right things, commitment is a self-fulfilling prophecy, there is always room, and don't get your hopes up. He argues that the true purpose of a startup is not to get rich but to work faster, compressing the drudgery of making a living to focus on meaningful work. Backed by concrete examples (Wufoo, Reddit, Microsoft, Apple) and metrics, the essay emphasizes flexibility, user obsession, and determination. While not directly about AI or systems engineering, it offers timeless wisdom for any builder.

www.paulgraham.com · 28 min
166

Are Software Patents Evil?

软件专利真的邪恶吗？——Paul Graham 的深度剖析

Paul Graham discusses whether software patents are evil. He argues that opposing software patents is opposing all patents, as the line between hardware and software blurs. The real problem is the USPTO's poor examination, granting overly broad patents like Amazon's one-click. In practice, patents have little impact on software innovation: startups rarely get sued, and patents mainly serve as acquisition currency. The true evil is patent trolls—firms that create nothing and litigate. Graham advises startups to pragmatically file patents for defensive reasons and focus on building great products. He also contrasts patents with secrecy, noting patents at least incentivize disclosure.

www.paulgraham.com · 27 min
167

How to Do What You Love

如何找到你热爱的工作：Paul Graham 的 2006 年经典指南

In this classic 2006 essay, Paul Graham examines the complexity of 'doing what you love.' He traces the roots of our misconceptions: childhood conditioning equating work with pain, and later being misled by prestige and money. He proposes two tests for genuinely loving your work: you'd do it for free and it's what you'd choose in your spare time. He compares two routes to achieving this—the 'organic route' vs. the 'two-job route'—and warns against deciding your career too early. Essential reading for any engineer or technical professional grappling with career choices.

www.paulgraham.com · 23 min
168

Good and Bad Procrastination

拖延的艺术：好与坏

Paul Graham reframes procrastination: not all of it is bad. The key is what you procrastinate on—errands or important work. Good procrastination means putting off trivial tasks to get large chunks of time and the right mood for hard problems. He argues that the most impressive people are often terrible procrastinators because they avoid errands to focus on what matters. Even the to-do list can be a form of bad procrastination. The solution is to let delight pull you rather than a list push you.

www.paulgraham.com · 10 min
169

Ideas for Startups

创业点子从何而来？——Paul Graham 的经典方法论

In this 2005 talk, Paul Graham argues that startup ideas are not million-dollar flashes but mistaken questions to explore. Most successful startups end up very different from their initial idea; the real value of the initial idea is to lead you toward the right one. To generate such questions, you need two things: exposure to promising new technologies and the right friends to bounce ideas off. He advocates the 'stay upwind' principle—choose work that maximizes future options rather than immediate pay. Specific techniques include focusing on problems you find intolerable (like his spam filter), turning luxuries into commodities (dramatically lower cost or complexity), and the 'Woz route'—building fun hacks with friends. The talk also covers exit strategies (aim for acquisition by multiple suitors) and warns against single-buyer traps. A must-read for aspiring technical founders.

www.paulgraham.com · 22 min
170

After the Ladder

职场阶梯之后：从安稳工作到直接变现

Paul Graham observes that the traditional corporate ladder—where employees worked their way up in exchange for job security—has been largely replaced by a startup model where individuals seek upfront payoff. This shift makes economic inequality appear more stark, but statistics ignore the implicit value of safe jobs, which were like annuities. Takeover trends since the 1980s eroded the ladder's appeal. While the new system is more liquid and efficient, it involves more risk; yet older generations also faced arbitrary project cancellations. Graham argues the financial difference between generations is smaller than it seems.

www.paulgraham.com · 3 min
171

Inequality and Risk

降低经济不平等如何扼杀创业与创新

Paul Graham argues that reducing economic inequality effectively means taking money from the rich, which disrupts the risk-reward balance and decreases willingness to take risks. Startups are inherently risky; capping potential returns discourages founders and investors, killing startups and slowing innovation. He suggests attacking the corruption that turns wealth into power, not wealth itself, through transparency and accountability.

www.paulgraham.com · 16 min
172

How to Fund a Startup

创业融资齿轮论：五类资金来源与轮次演进

Paul Graham compares venture funding to shifting gears, describing five sources of startup money: friends and family, consulting, angel investors, seed firms, and venture capital. Through a hypothetical startup's lifecycle, he walks through seed, angel, and Series A rounds, covering valuation, terms, dilution, and investor psychology. Key insights: frugality is power, deals often fall through, and founders must always maintain alternatives. A classic primer for entrepreneurs and anyone interested in early-stage finance.

www.paulgraham.com · 51 min
173

What Business Can Learn from Open Source

开源三课：做爱做的事，自下而上，告别办公室

In this essay adapted from a 2005 OSCON talk, Paul Graham argues that open source and blogging reveal three lessons business must learn: people work harder on things they love, conventional offices kill productivity (forced hours, meetings, fragmentation), and bottom-up organization outperforms top-down control. He critiques the master-servant legacy of employment and advocates for investor-founder relationships as more productive and less paternalistic. While not a technical engineering piece, it offers thought-provoking insights for startup founders and managers about motivation, workspace, and organizational design.

www.paulgraham.com · 25 min
174

The Submarine

公关潜艇：新闻界的SEO

This essay reveals how the PR industry secretly shapes mainstream news. Paul Graham argues that more than half of non-political, non-crime, non-disaster stories originate from PR firms. He recounts his own startup's experience paying a PR firm $16,000/month to generate press hits, and dissects the coordinated campaign behind the 'suits are back' trend story across multiple outlets. The piece contrasts the manufactured nature of traditional media with the authenticity of online writing, predicting a structural decline for print.

www.paulgraham.com · 13 min
175

Writing, Briefly

写作的意外收获：好的写作激发思想

Paul Graham accidentally wrote this short essay on writing while answering an email, taking only 67 minutes (23 writing, 44 rewriting). He argues writing is far more important than most realize: it doesn't just communicate ideas, it generates them. If you're bad at writing and dislike it, you'll miss out on most ideas writing would have generated. He offers a dense list of practical advice: write a bad version fast, rewrite repeatedly, cut everything unnecessary, use conversational tone, develop a nose for bad writing, imitate writers you like, talk through your plan first, expect 80% of ideas to emerge after starting, be confident to cut, have trusted friends read drafts, avoid detailed outlines, mull ideas for days, carry a notebook, start with the first sentence that comes, if on deadline say the most important sentence first, write about what you like, don't try to sound impressive, change topics freely, use footnotes for digressions, use anaphora, read aloud to find awkward phrases and boring parts, tell the reader something new and useful, work in large blocks of time, when restarting reread what you've written, leave yourself an easy starting point, accumulate notes at the bottom but don't feel obliged to cover them, write for a less careful reader, fix mistakes immediately, ask friends which sentence you'll regret, tone down harsh remarks, publish online to have an audience that pushes you to write more, print drafts, use simple Germanic words, learn to distinguish surprises from digressions, and recognize the approach of an ending and grab it.

www.paulgraham.com · 3 min
176

Return of the Mac

黑客回归Mac：从ThinkPad到Powerbook的迁徙

In 2005, Paul Graham observes that top hackers are switching back to Macs, driven by OS X's combination of elegant design and FreeBSD underneath. He recalls the Mac's golden era in the 1980s when it was the canonical hacker's machine, and the exodus to Intel boxes running Linux/FreeBSD. Graham argues the hacker market is small but influential, as future technology trends originate in CS departments. He provides OS share data from Y Combinator's visitors: Windows 66.4%, Mac 18.8%, Linux 11.4%.

www.paulgraham.com · 6 min
177

Why Smart People Have Bad Ideas

聪明人为何总出昏招——从失败的Artix看创业点子误区

Paul Graham uses his own failed startup Artix (building websites for art galleries) as a case study to analyze three common mistakes smart people make when coming up with startup ideas: the 'still life effect' (sticking with an initial idea because of sunk time), choosing something cool over something profitable, and picking a poor market out of fear of competition. He notes that many hackers are good at solving problems but bad at choosing which ones to solve, and this skill can be learned. The essence of a startup is getting brilliant people to do work that is beneath them, and learning what customers want is easier than writing an optimizing compiler.

www.paulgraham.com · 14 min
178

Undergraduation

大学里如何成为优秀的程序员——Paul Graham 的建议

Paul Graham offers advice to computer science undergraduates on how to become great hackers in college. Key points: work on engaging hard projects, not just class assignments; math is valuable for its metaphors but not strictly necessary; avoid bogus fields like social sciences; choose tech stacks based on target employer; grad school is paradise except for the dissertation; the key to success is genuine passion. The essay is personal, anecdotal, and candid.

www.paulgraham.com · 21 min
179

A Unified Theory of VC Suckage

风投为何讨人厌：统一理论

In this classic 2005 essay, Paul Graham traces the fundamental flaws of venture capital to the fee structure (roughly 2% annual management fee plus carry). Because VCs manage huge funds, each investment must be multi-million dollar, forcing them to be paranoid, sneaky, and controlling. This leads to high valuations, oversized investments, and pressure for IPOs or massive exits, often harming startups. Graham illustrates with his own startup experience—a competitor raised $20M and failed, while his company spent only $2M total and sold for $50M to Yahoo. The essay argues that only the top-tier VCs (like Sequoia, Kleiner Perkins) can afford to behave decently because they have pick of the best deals. For founders and anyone curious about VC dynamics.

www.paulgraham.com · 8 min
180

Made in USA

美国制造的文化代价：为何擅长软件电影，却做不好汽车城市

Paul Graham argues that the American 'just‑do‑it' attitude, which prizes speed over deliberation, is a double‑edged sword. It gives the US an edge in malleable fields like software and movies (e.g., Dan Bricklin coding VisiCalc in a weekend), but leads to poor design in physical goods like cars and cities. He contrasts the clumsy AMC Matador with Japanese swords from 1200, and points to Apple’s iPod as proof that American companies can also achieve Japanese‑level craftsmanship when designers lead. He predicts that occupational character will increasingly override national character, allowing each country to adopt the best working style for each domain.

www.paulgraham.com · 11 min
181

What You'll Wish You'd Known

你将来会后悔没早知道的事

Paul Graham's talk to high school students argues that you shouldn't fixate on a life plan early; instead, explore interests and stay upwind. School and college admissions are flawed, but don't rebel—treat school as a day job and work on real projects in your spare time. Curiosity about hard problems matters more than discipline. Start with a small project and let your path emerge.

www.paulgraham.com · 28 min
182

It's Charisma, Stupid

魅力决定胜负？美国总统选举的简单理论

Paul Graham proposes a provocative theory: in US presidential elections since the advent of television, the more charismatic candidate always wins. He revisits elections from 1960 to 2004, arguing that conventional left/right shift analyses fail to explain the consistent pattern. Graham claims that the major parties' policy positions are so closely matched that charisma becomes the tiebreaker. The essay also explores implications for the Democratic Party, which tends to nominate earnest but dull candidates, and suggests that once both parties recognize this dynamic, charisma will cancel out and elections will be decided on issues again.

www.paulgraham.com · 9 min
183

Bradley's Ghost: Why Exit Polls Were Wrong in 2004

出口民调为何失准？——2004年大选中的“布拉德利效应”

Paul Graham examines why 2004 US election exit polls drastically overestimated Kerry's vote, particularly in swing states. He suggests a variant of the 'Bradley Effect': some Bush voters were unwilling to admit their choice due to social pressure from liberal elite values (e.g., NPR values). This led to non-response or dishonesty, skewing the polls. The article argues that when elite culture leans liberal, polls underestimate the conservatism of ordinary voters.

www.paulgraham.com · 4 min
184

How to Start a Startup

创业起步：三件事决定成败

Based on his experience founding Viaweb, Paul Graham argues that startup success hinges on three things: good people, making something customers want, and spending as little as possible. He emphasizes that ideas are less important than execution, advises rapid prototyping and user feedback, and warns against overspending. The essay covers funding, equity splits, IP issues, and the value of staying lean. Aimed at aspiring founders, especially engineers.

www.paulgraham.com · 54 min
185

A Version 1.0

好文章是改出来的——Paul Graham论写作的真谛

Paul Graham shows the earliest surviving draft of his essay 'The Age of the Essay', with surviving text in red and deleted text in gray, demonstrating that good writing is rewriting. He reveals that for every word in the final version, he writes three or four, and 90% of the content only occurs to him while writing. A real essay, unlike school assignments, starts with a question and aims to explore, not defend a thesis. He criticizes the historical conflation of writing instruction with literary studies, which forces students to write about uninteresting topics like symbolism in Dickens. This piece is for anyone interested in writing craft and educational critique.

www.paulgraham.com · 25 min
186

What the Bubble Got Right

泡沫中的真金：互联网泡沫隐藏的持久趋势

Paul Graham reflects on the Internet Bubble and argues that despite the crash, several trends it popularized will endure: retail VC (taking companies public early), the genuine importance of the Internet, choice-driven economics, the edge of young founders, informality, nerd culture, stock options, startups built to be sold, California's innovation ecosystem, and the multiplier effect of technology on productivity. The overarching theme: good ideas will count more over time.

www.paulgraham.com · 21 min
187

The Python Paradox

Python悖论：小众语言吸引顶尖程序员

Paul Graham argues that companies choosing to write software in a relatively esoteric language like Python can attract better programmers, because those who learn Python do so out of genuine interest, not just to get a job. This self-selection creates a pool of smarter, more dedicated developers. He also discusses the importance of language aesthetics and code readability. A thought-provoking piece for tech managers and language designers.

www.paulgraham.com · 3 min
188

The Age of the Essay

论真正的随笔：探索而非辩护

Paul Graham traces the origins of high school essay writing to medieval legal disputation and the 19th-century merger of composition with literary studies. He argues that a real essay (from Montaigne's 'essai' meaning 'attempt') is not about defending a thesis but about thinking on paper. Good essays follow the principle of 'flow interesting,' allowing meandering and surprise. The Web makes essays a democratic form again.

www.paulgraham.com · 26 min
189

Mind the Gap

收入差距：财富创造与技能价值的再审视

Paul Graham argues that income inequality is not inherently unjust; rather, it reflects differences in skill and productivity, just like in chess or art. He explains wealth as created, not distributed, and traces the historical shift from theft-based to creation-based wealth. Technology widens income gaps but narrows lifestyle and social gaps. In a modern democracy, increasing variation in income can be a sign of health. This essay is for engineers interested in socioeconomic dynamics and the role of technology.

www.paulgraham.com · 33 min
190

Great Hackers

优秀黑客的特质与管理之道

Paul Graham draws from personal experience to discuss the vast productivity gap (10-100x) between great hackers and average programmers. He analyzes core traits of great hackers: they love programming, demand good tools (open source, Python/Perl over Java), need quiet offices with doors (not cubicles), and crave interesting technical challenges over tedious bug fixes. Recognizing a great hacker is only possible through working together—resumes are useless. Companies should attract them via autonomy, interesting projects, and brilliant peers, not just high salaries. The essay also covers self-cultivation: curiosity, focus, political incorrectness, and the habit of questioning assumptions.

www.paulgraham.com · 29 min
191

If Lisp is So Great

如果Lisp那么棒，为什么用的人不多？

In this classic essay, Paul Graham answers the perennial question: if Lisp is so great, why don't more people use it? He draws an analogy with literature, where popularity and quality are often uncorrelated. He examines how popularity is self-perpetuating in programming languages — through library ecosystems, program compatibility, and manager bias — and argues that Lisp remains superior in many ways, especially in expressive power and runtime performance. Like Jane Austen's novels, it will never be the most popular but will always find devoted users.

www.paulgraham.com · 3 min
192

Why Arc Isn't Especially Object-Oriented

Arc语言为何不特别面向对象

Paul Graham questions the current mania for object-oriented programming (OOP), arguing that three and a half of the five reasons people like it are bad. Drawing on his experience with Lisp and the Arc language, he contends: OOP is a workaround for statically-typed languages without closures; it imposes discipline on mediocre programmers at big companies but produces bloated, duplicated code; it generates the appearance of work (e.g., turning a Lisp list operation into a file of classes). Graham himself has never needed CLOS and warns against adding features one has never used. The essay is a thought-provoking critique of OOP, but falls outside AI/engineering.

www.paulgraham.com · 3 min
193

Succinctness is Power

简洁即力量：编程语言设计的核心原则

In this classic 2002 essay, Paul Graham posits that succinctness is the primary measure of a programming language's power. He argues that high-level languages exist to shrink source code, allowing programmers to express more with fewer 'elements' (abstract syntax tree nodes). Citing Brooks' observation that programmers produce about the same lines of code per day regardless of language, and Ulf Wiger's study showing Erlang is 4-10x more succinct than C++, Graham contends that more succinct languages lead to higher productivity and fewer bugs. He discusses readability as a product of per-line readability and line count, so succinctness aids rather than hinders readability. Graham also distinguishes between language-level succinctness and program-level density, noting that while individual programs can be too dense, it is hard for a language to be 'too succinct'. This essay is for language designers, programmers interested in language comparison, and those concerned with code efficiency.

www.paulgraham.com · 17 min
194

Taste for Makers

Paul Graham：品味是设计者的核心能力

Paul Graham challenges the notion that taste is merely subjective, arguing that it is a core skill for designers. Drawing on examples from mathematics, engineering, architecture, and painting, he distills common principles of good design: simplicity, timelessness, solving the right problem, being suggestive, often funny, hard to achieve but appearing effortless, using symmetry, resembling nature, embracing redesign and copying, sometimes being strange, thriving in communities, and daring to defy convention. He contends that taste is cultivated by learning to recognize and fix what is ugly. This essay is for any maker—engineer, designer, or developer—seeking a deeper understanding of what makes work truly good.

www.paulgraham.com · 25 min
195

Beating the Averages

超越平庸：Lisp如何成为创业公司的秘密武器

Paul Graham recounts his startup Viaweb's experience to argue that programming languages vary in power and choosing a more powerful language like Lisp can provide a decisive competitive advantage for startups. He introduces the 'Blub paradox': programmers tend to think their current language is good enough and fail to appreciate more powerful ones. In Viaweb, roughly 20-25% of the editor code consisted of Lisp macros—programs that write programs—enabling features competitors couldn't match. For startups, this technical edge can be a secret weapon. The article is relevant for founders making tech choices and programmers curious about language design.

www.paulgraham.com · 25 min
196

New Essays: How to Earn a Billion Dollars & The Brand Age

Paul Graham 新随笔：亿万之路与品牌时代

Paul Graham published links to two new essays, 'How to Earn a Billion Dollars' and 'The Brand Age'. The page contains only titles and links, with no technical details. This is a pure announcement and does not fit the magazine's technical scope.

www.paulgraham.com · 1 min
197

Being Popular

编程语言流行之道

Paul Graham examines what makes a programming language popular from a hacker's perspective. He argues that a language's success depends not on theoretical elegance but on whether it meets hackers' actual needs: a free implementation, a well-written book (like K&R), and a system to hack (e.g., scripting language of Unix). Hackers value brevity and hackability—the ability to bypass abstractions. Writing throwaway programs is crucial because large systems often evolve from them. Libraries matter more than core language features. Syntax is not Lisp's main problem; weak libraries and lack of system integration are. He sketches the dream language: terse, interactive, transparent abstractions, excellent profiler, and well-designed libraries.

www.paulgraham.com · 44 min
198

Programming Bottom-Up

自底向上编程：Lisp 哲学如何重塑程序与语言的边界

In this classic essay from 'On Lisp,' Paul Graham explains the bottom-up design philosophy of Lisp programming: instead of decomposing a program top-down into subroutines, the programmer builds up the language toward the problem. Language and program co-evolve until the boundary settles along the natural frontiers of the problem. This yields smaller, clearer, and more reusable code, with fewer components and fewer connections. Graham argues that bottom-up design makes programs easier to read, promotes code reuse, and is especially powerful for small teams. The essay contrasts this approach with traditional top-down design and illustrates how augmenting the language transforms the program structure. While focused on Lisp, the ideas generalize to DSL design, framework creation, and API design.

www.paulgraham.com · 6 min
199

The Other Road Ahead

服务器端软件：程序员的未来与创业机会

In this classic 2001 essay, Paul Graham argues that web-based applications will dominate desktop software, drawing from his experience building Viaweb (later Yahoo Store). He details the advantages for users—no installation, universal access, automatic updates—and the radical shift for developers: continuous releases (3-5 per day), instant bug fixing, and the merging of developer and sysadmin roles. With a team of three, Viaweb achieved high reliability and low cost ($5 per user). Graham contends that web apps lower startup barriers and outcompete large companies, despite placing more stress on programmers. He urges hackers to embrace this model as a way to circumvent Microsoft's desktop monopoly.

www.paulgraham.com · 69 min
200

General and Surprising

通用与新奇——最有价值的见解

Paul Graham explores how to generate insights that are both general and surprising. He argues that the most valuable ideas combine these two qualities, yet most people produce either surprising but narrow content (gossip) or general but trivial platitudes. The key is small increments: add a touch of novelty to a general idea, or extract a general lesson from a surprising observation. He encourages writers not to fear repetition, as slight variations can lead to breakthroughs. The essay emphasizes that achieving even a small delta of novelty on top of general ideas is a real achievement. Suitable for readers interested in creativity, writing, and thinking methodologies.

www.paulgraham.com · 3 min
201

Economic Inequality

经济不平等并非单一问题——Paul Graham 论创业与财富创造

Paul Graham argues that economic inequality is not a single phenomenon but has multiple causes. The common 'pie fallacy' conflates different types of wealth creation and wealth taking. He contends that inequality driven by productivity differences (e.g., successful startups) is beneficial and cannot be eliminated without stifling innovation and driving away talent. Instead of targeting inequality metrics, society should focus on concrete problems like poverty, lack of social mobility, and rent-seeking. This essay is relevant for readers interested in startup economics and policy.

www.paulgraham.com · 20 min
202

Default Alive or Default Dead?

你的创业公司默认会死吗？——一个必须问的问题

In this classic essay, Paul Graham asks a critical question for any startup past 8-9 months: given current revenue growth and expenses, will they reach profitability before running out of money? He finds that most founders don't know. Graham introduces the concepts of 'default alive' (you survive without additional funding) and 'default dead' (you don't). He warns of the 'fatal pinch' — default dead + slow growth + insufficient time to fix. The biggest killer is hiring too fast; founders overhire thinking it drives growth, but the real problem is usually an unappealing product. VCs push for kill-or-cure strategies, but founders should prioritize survival. He advises asking the question early, separating hopes from facts, and doing things that don't scale (like Airbnb).

www.paulgraham.com · 9 min