You Weren't Meant to Have a Boss

You Weren't Meant to Have a Boss. An essay by Paul Graham, from March 2008.

Normal isn't natural

Technology keeps pulling normal away from natural. Rich-country diets and so little exercise aren't what our bodies evolved for.

Graham suspects work is the same. An ordinary job might do to your mind what white flour and sugar do to your body.

Next to big-company programmers, the founders he works with seem better off, the way a body is on a long run, not slumped on the sofa.

Or take lions: in the wild they seem ten times more alive. A zoo is easier, but not the life they evolved for. Working for yourself may be our wild.

How big a group?

So what makes a big company unnatural? Its size. Humans didn't evolve to work in groups that big.

Each species has its own group size: impalas about 100, baboons 20, lions rarely 10. For humans, 8 works well, 20 gets hard, 50 is a struggle.

Yet many people now work in companies of hundreds or thousands. That's technology's doing, not human nature's.

A tree of bosses

Huge groups don't work, so companies split into small teams. Coordinating them takes a new role: the boss.

The teams form a tree, and your boss is where your team joins it. One level up, your boss stands in for your whole team.

Here's the odd part. For ten managers to work like ten ordinary people, each team below must act as one person, sharing one person's freedom.

Teams never fully manage it, but the pressure never lets up. So the bigger the tree, the less freedom each person has, even in the same team of 10.

A fake tribe

So a team of ten in a big company is a fake tribe. The headcount feels right, but individual initiative is missing.

Hunter-gatherer bands had leaders, with a little more say. But they rarely gave orders the way a boss does.

Don't blame the boss, though. One level up, your whole team counts as a single virtual person. The boss just passes that limit down.

So the job feels right and wrong at once. Graham compares it to high fructose corn syrup: a taste of what you crave, without what you need.

Junk food, junk jobs

Food shows how bad a default can be. Almost everything to eat in America is bad for you, made mostly of white flour, sugar, corn syrup and oil.

Why does junk win? It tastes great now, and it scales, so it's cheap and heavily marketed. Real food is costly, obscure, and pays off later.

Jobs work the same way. The typical MIT grad wants Google or Microsoft: a known brand, safe, good pay now. The cost comes later, as a vague malaise.

Founders are the health-food diehards of work: a tiny minority, living as humans evolved to. As Graham puts it: “In an artificial world, only extremists live naturally.”

Hardest on programmers

This hits programmers hardest. Salespeople repeat the same pitch, but code never needs writing twice. So programming is always about making new things.

One founder went to Google to learn, and learned less than he expected. Legacy code, overhead and other teams' interfaces blocked most of his ideas.

And a blockage downstream spreads upstream: if you can't act on ideas, you stop having them. Free to try anything, you have more.

Freedom by company size

You can tune your freedom by company size. Founders get the most, the first ten employees almost as much. Even 100 people feels different from 1,000.

Small doesn't guarantee freedom: a small company's boss can still be a tyrant. Size sets a ceiling, not a floor. But a big company's structure forces it to be one.

Stay small

For companies, bigger means slower, however hard they fight it. It comes with the tree.

The only escape Graham can imagine is no tree at all: independent groups cooperating like parts of a market. He knew no tech company that did it.

Short of that, stay as small as you can, and hire the best. Mediocre hires cost you twice: they do less, and it takes more of them.

For individuals, the same: aim small. Graham revised his old advice: work for someone else only at a small company. Itching to start a startup? Go ahead.

Coming back to life

Graham once told graduates to wait, since most startups fail. They do. But an ambitious programmer is better off failing at their own thing than joining a big company.

They learn more, and may even do better financially: early salaries often lead to debt. Fail at a startup, and your net worth is zero, not negative.

Y Combinator's founders show no benefit from big-company years. Those who come from one often seem conservative, but much of that is learned. It can burn off.

Founders often arrive looking like refugees. Three months later they seem inches taller: more worried, yet happier. Just like lions in the wild.

In short

Humans evolved to work in small groups, not companies of thousands.

Big companies are trees, so freedom shrinks as the tree grows.

A big-company job is like corn syrup: right on the surface, missing initiative.

So aim small: start your own thing, or join a small one.

The original has much more, and is better than this. Read it: paulgraham.com/boss.html