How to Do Great Work

How to Do Great Work. An essay by Paul Graham, July 2023.

What to work on

Gather advice on great work from many fields: what do they have in common? Graham went looking, and found the overlap has a definite shape.

First, choose the work. It should fit your natural aptitude and your deep interest, and leave room for greatness. Ambitious people rarely miss that last part.

Not sure yet? Guess, and start. Pick projects of your own that feel excitingly ambitious, and follow what you're unusually curious about.

Four steps to the frontier

Next, learn enough to reach the edge of what's known. From far away, that edge looks smooth. Up close, it's full of gaps.

The trick is to notice them. Your brain prefers to paper over gaps, and many discoveries began by questioning what everyone assumed.

Then explore the promising ones, even odd ones everyone ignores. Pry a gap open and you may find a whole new world inside.

Four steps: field, frontier, gaps, explore. Steps two and four take hard work, and only real interest will drive you that hard.

Finding it by doing

Choosing is hard, because you only learn what work is like by doing it, sometimes for years. Yet schools expect you to commit as a teenager.

So make yourself a big target for luck: try many things, meet people, read, ask questions. When in doubt, follow what's most interesting.

Don't plan much. At each stage, do what's most interesting and keeps the best options open. Graham calls this staying upwind.

Starting, and compounding

Working has a technique, like sailing. The hardest kinds may last only four or five hours a day, ideally in one unbroken block.

Starting is harder than continuing, so it's fine to trick yourself over the hump. Graham tells himself he'll just reread what he has so far.

The dangerous delay is per project: putting off the ambitious thing for years, while busy. Now and then, ask: is this what I most want to do?

Be consistent: a page a day is a book a year. Work that compounds grows exponentially, but feels flat at first. Push through the start.

Aim high, earnestly

Learn what the best work in your field is. Then try to be the best: aim only for good, and you'll usually fall short, even of that.

Don't chase a style. Do your best, and style comes on its own. Aiming for it is affectation: pretending to be someone else.

Instead, be earnest. At its core that means intellectual honesty: to see new truths, you need a sharp eye for the truth.

Admit mistakes eagerly, and spend nothing on how you seem. Be the one putting ideas out, even at the risk of looking foolish.

Elegance and discovery

Great work hangs together. Facing a choice? Take the more consistent option. And be willing to redo and cut, even parts you're proud of.

Strip the work to its essence. It gets more concentrated, and you can't fool yourself about whether anything real is there.

The best work can look almost effortless, as if it was already there. When you can't tell if you're creating or discovering, that's a great sign.

Originality is a habit

A great idea must be true and new. Spotting new ones is its own skill, a habit of mind: original people throw off ideas like a grinder throws sparks.

Ideas come while working on something slightly too hard, or while writing, walking, or wandering into new topics.

Graham offers a test: “When an idea seems simultaneously novel and obvious, it’s probably a good one.”

Why are obvious ideas hard to see? We look through models of the world. Fixing a broken one is hard. After that, the idea looks obvious.

Strict, and willing to break rules

Broken models leave clues where they hit reality. Most people look away. Einstein saw where Maxwell's equations led because he was stricter.

You also need to be comfortable breaking rules, since a new model breaks the old one's. The sun-centered model took most of a century to win.

A promising new idea must look bad to most people, or it would be taken already. Look for the right kind of crazy: exciting, not depressing.

Cherished but mistaken beliefs cast a shadow of unexplored ideas. So ask: what is your field religious about, and what opens up if you drop it?

Questions over answers

People are far more original in solving problems than in choosing them. Even smart people follow fashion, so unfashionable problems are bargains.

If you took a break from serious work for something just because it's interesting, what would it be? It probably matters more than it seems.

We assume big ideas are answers. Often, though, the question was the insight. How do new species arise? Just asking it put you somewhere new.

So carry questions, even ones from childhood. Pull on a loose thread and see how far it goes. The best questions grow as you answer them.

Start small, then iterate

Pull a little on lots of threads. Big things start as experiments and side projects. You can't have many good ideas without many bad ones.

Grow things in successive versions. Begin with the simplest thing that might work, show it to people early, and let them steer the next one.

If people call an early version a toy, good sign: all it lacks is scale. Planning isn't a virtue. It's a necessary evil.

Take all the risk you can afford. If you never fail, you're too careful. Even a failure leaves you with questions few have asked.

Youth, age and school

Youth has energy, time, optimism and freedom; age has knowledge, efficiency, money and power. The young rarely see how rich in time they are.

Fresh eyes help too. When something in a new subject seems wrong, note it. If it still looks wrong later, it may be an undiscovered idea.

But school leaves nonsense behind. It makes you passive, hands you problems that are always solvable, and teaches you to hack the test.

Treat your education as your own project. Don't wait for a gatekeeper's big break, or take committees to heart. Just do good work.

Copying and colleagues

Copying is how beginners learn, and that's fine. Do it openly, not unconsciously. The most powerful kind carries an idea from one field to another.

But the easiest features to imitate are often the flaws. Some talented people are jerks. Talent is just how they get away with it.

Seek the best colleagues. They shape you as well as your work, so pick people you'd like to become. One or two great ones beat a building of good ones.

How do you know they're good enough? They keep surprising you. They see and do things you can't.

Guard your morale

Protect your morale; ambitious work rests on it. Good work lifts morale, which leads to better work. But the loop also runs in reverse.

Great work is a depth-first search. After a setback, try again, or back up a little and try again. Just never abandon the root.

Your audience can be tiny: a handful who love your work is enough. Keep company that raises your energy, and look after your body.

Impress people you respect; fame is mostly noise. Don't ask how prestigious a kind of work is. Ask how well it could be done.

Curiosity, and a question

If an oracle named the secret of great work in one word, Graham would bet on curiosity. It drives all four steps.

The factors are ability, interest, effort and luck. Ignore luck; assume effort. So find the work where your ability and interest combine.

Many more could try than do. Modesty and fear hold them back: aiming to be Newton seems presumptuous, and doomed. Few decide consciously.

So Graham asks it outright: do you want to do great work? You needn't tell anyone, and failing isn't so bad. Why not you?

In short

Choose work that fits your aptitude and interest, then let curiosity take you to the frontier and its gaps.

Work consistently on excitingly ambitious projects. Plan little; stay upwind.

Be earnest: honest, informal, strict, and willing to break rules.

Treasure good questions, start small, make versions, and guard your morale.

It is long, and much richer than this. Read it: paulgraham.com/greatwork.html