All Vaults
PG

Paul Graham

Startups & Essays

30 essential Paul Graham essays — startups, makers, ideas, and how to think. The condensed reading list for indie hackers, founders, and curious minds.

30 articles·4 phases·Updated 4/14/2026·
Curated by@burn451
Get Burn 451

What if startup advice and life advice are the same thing — pick problems worth doing, do them simply, ignore what doesn't matter?

About this vault

Paul Graham wrote 200+ essays at paulgraham.com over 25 years. This vault picks the 30 that matter most for builders today — covering startups (How to Start a Startup, Do Things that Don't Scale, Default Alive or Default Dead, Founder Mode), thinking and working (How to Do Great Work, The Top Idea in Your Mind, How to Think for Yourself), programming (Hackers and Painters, Beating the Averages, Great Hackers), society (Cities and Ambition, The Refragmentation), and life (Life is Short, What You'll Wish You'd Known). Organized by theme so you can read by interest, not chronology. AI summary on every essay so you can decide what to read in 5 seconds.

30 articles

Startups: The Founder's Field Guide

Paul Graham's most-cited startup essays. From 'How to Start a Startup' to 'Founder Mode' — the playbook used by 5,000+ Y Combinator founders. Covers picking ideas, raising money, scaling, the 18 mistakes that kill startups, and why default-alive matters more than growth metrics.

How to Start a Startup

The canonical 4-step formula: good people, build something users actually want, spend as little as possible. Most startups fail at #2. Talk to users. Build the smallest thing that solves their problem. Charge if you can. Everything else is procrastination dressed up as strategy.

Do Things that Don't Scale

The biggest mistake new founders make: assuming startups are like 'getting a ship to fly', when they're like 'cranking a flywheel by hand'. Recruit users one by one. Provide insanely good service to your first 10 customers. Manual hacks that won't scale are how you find what does.

Startup = Growth

A startup is a company designed to grow fast. Not 'a small business', not 'an entrepreneur'. The growth rate that makes you a startup vs. a regular business: 5-7% weekly. Below 1% weekly = dying. PG's most rigorous definition.

Founder Mode

The conventional wisdom 'founders should learn to manage like trained executives' is wrong. Steve Jobs ran Apple by skip-level meetings, deep involvement, and refusing to delegate the things that mattered. 'Manager mode' optimizes for legibility; founder mode optimizes for the company. Triggered an industry-wide reckoning when published.

Default Alive or Default Dead?

The single most important question in early-stage startups: at current spending and growth, will you reach profitability before your money runs out? Most founders avoid running this calculation. Most who run it discover they're default dead. The good news: the answer is fixable if you know it.

The 18 Mistakes That Kill Startups

PG's autopsy of startup failure — single founder, bad location, marginal niche, derivative idea, obstinacy, hiring bad programmers, choosing wrong platform. Reads like a checklist 20 years later. Most fatal: launching too late, not making something users actually want.

Schlep Blindness

Why founders avoid the best ideas: they require unpleasant work (schleps). Stripe was a schlep. Airbnb was a schlep. The biggest opportunities sit in plain sight, ignored because they look painful. Train yourself to notice schleps and run toward them.

How to Get Startup Ideas

Don't try to think of startup ideas. Look for problems — preferably problems you yourself have. Live in the future, then build what's missing. The best ideas seem too narrow at first. The Microsoft starting on hobbyist computers, Apple on people who couldn't afford terminals.

Black Swan Farming

VC returns are dominated by extreme outliers. The best YC investments returned 1,000x; most returned 0. The skill isn't picking 'good' startups, it's spotting things that look weird but might be huge. Most great founders look strange initially. The good ones disagree with you and explain why.

Before the Startup

Counter-intuitive advice for college students: don't try to be 'startup-ready'. Build skills (tech), build connections, work on your own projects. The startup itself is best learned by doing one. The best preparation is being good at things and noticing problems.

How to Think and How to Work

PG's essays on the meta-skills behind great work — how to pick what to work on, how to think clearly, how to write to think better. Read these before any career or research decision.

How to Do Great Work

PG's longest essay (12,000 words) — the distilled wisdom from 50 years of working with talented people. Pick something you have a genuine interest in. Learn enough to get to the frontier. Notice gaps. Work hard but consistently. The hard part is not effort, it's choosing the right problem.

The Top Idea in Your Mind

You can have multiple things on your todo list, but only one thing on your mind — the thing your subconscious works on while you shower. Money disputes will dominate your top idea slot if you let them. So will personal drama. Choose carefully what occupies that slot.

The Top of My Todo List

5 regrets of the dying (from a hospice nurse): I wish I'd lived true to myself. I wish I hadn't worked so hard. I wish I'd expressed my feelings. I wish I'd stayed in touch with friends. I wish I'd let myself be happier. PG turns each into a personal todo list item. Read once a year.

How to Think for Yourself

Independent thinking requires three traits: caring more about truth than orthodoxy, picking unfashionable problems, and being able to discount fashion. Most education optimizes for the opposite. Surround yourself with other independent minds. Avoid environments where heresy is punished.

Putting Ideas into Words

Writing about something is the most rigorous test of whether you actually understand it. The act of converting half-formed thoughts to sentences exposes the gaps. Don't write to communicate ideas you have — write to discover whether your ideas hold up.

Keep Your Identity Small

The fewer labels you adopt as identity ('I'm a Republican', 'I'm a Mac user', 'I'm a vim person'), the better you can think. Once something is your identity, you can't think clearly about it — you're committed to defending it. Limit your identities.

The Need to Read

Reading isn't optional, even in the YouTube era. There's a category of ideas you can only access through dense, sustained text. Reading also trains the kind of attention you need for original thinking. Don't replace it with podcasts, videos, or 'audio summaries'.

Hackers, Programming, and Taste

The essays that made PG famous in the dev community. Hackers and Painters, Beating the Averages, the Python Paradox — written when most people still thought programming was a clerical job. Foundational reading for anyone who builds.

Hackers and Painters

The essay that named PG's first book. Programming is closer to painting than to math or engineering. You learn by doing, not by studying. You build things you don't fully plan. The best programmers are people who think of programming as a creative act, not a clerical one.

Beating the Averages

PG's case study from Viaweb: they used Lisp while competitors used Perl/C, and that single technical choice gave them a sustained advantage no competitor could match. The lesson generalizes: a powerful technical advantage compounds. Pick tools accordingly.

Great Hackers

What separates great programmers from merely good ones: they care about the problem, not just the technical exercise. They use better tools. They work in concentrated bursts. They despise being told what to do. They're often invisible to corporate hiring processes.

The Python Paradox

If you advertise for Python programmers, you'll get smarter applicants than if you advertise for Java programmers — not because Python makes people smart, but because choosing to learn an unfashionable language signals self-direction. A heuristic for technical hiring.

Taste for Makers

Good design has principles you can name: simplicity, timelessness, solves the right problem, suggestive rather than complete. Bad design tries to look impressive. Good designers cultivate taste through years of looking. Taste isn't subjective — it converges across cultures and centuries.

Why Nerds are Unpopular

Smart kids are unpopular in American high school not because being smart is uncool, but because they don't focus on popularity. Popularity is a job, requiring constant attention. Nerds opt out, then pay the social price. The payoff comes after high school.

Start reading, not hoarding.

Import this vault to Burn 451 and actually read what matters.

Content attributed to original authors. Burn 451 curates publicly available writing as a reading index. For removal requests, contact @hawking520.