Paul Graham - Essays

[2] I know this is why I haven't released Arc. The moment I do, I'll have people nagging me for features.

[3] A web site is different from a book or movie or desktop application in this respect. Users judge a site not as a single snapshot, but as an animation with multiple frames. Of the two, I'd say the rate of improvement is more important to users than where you currently are.

[4] It should not always tell this to users, however. For example, MySpace is basically a replacement mall for mallrats. But it was wiser for them, initially, to pretend that the site was about bands.

[5] Similarly, don't make users register to try your site. Maybe what you have is so valuable that visitors should gladly register to get at it. But they've been trained to expect the opposite. Most of the things they've tried on the web have sucked-- and probably especially those that made them register.

[6] VCs have rational reasons for behaving this way. They don't make their money (if they make money) off their median investments. In a typical fund, half the companies fail, most of the rest generate mediocre returns, and one or two "make the fund" by succeeding spectacularly. So if they miss just a few of the most promising opportunities, it could hose the whole fund.

[7] The attitude of a running back doesn't translate to soccer. Though it looks great when a forward dribbles past multiple defenders, a player who persists in trying such things will do worse in the long term than one who passes.

[8] The reason Y Combinator never negotiates valuations is that we're not professional negotiators, and don't want to turn into them.

[9] There are two ways to do work you love: (a) to make money, then work on what you love, or (b) to get a job where you get paid to work on stuff you love. In practice the first phases of both consist mostly of unedifying schleps, and in (b) the second phase is less secure.

Thanksto Sam Altman, Trevor Blackwell, Beau Hartshorne, Jessica Livingston, and Robert Morris for reading drafts of this.

(This essay is derived from a keynote at Xtech.)

Could you reproduce Silicon Valley elsewhere, or is there something unique about it?

It wouldn't be surprising if it were hard to reproduce in other countries, because you couldn't reproduce it in most of the US either. What does it take to make a silicon valley even here?

What it takes is the right people. If you could get the right ten thousand people to move from Silicon Valley to Buffalo, Buffalo would become Silicon Valley. [1]

That's a striking departure from the past. Up till a couple decades ago, geography was destiny for cities. All great cities were located on waterways, because cities made money by trade, and water was the only economical way to ship.

Now you could make a great city anywhere, if you could get the right people to move there. So the question of how to make a silicon valley becomes: who are the right people, and how do you get them to move?

I think you only need two kinds of people to create a technology hub: rich people and nerds. They're the limiting reagents in the reaction that produces startups, because they're the only ones present when startups get started. Everyone else will move.

Observation bears this out: within the US, towns have become startup hubs if and only if they have both rich people and nerds. Few startups happen in Miami, for example, because although it's full of rich people, it has few nerds. It's not the kind of place nerds like.

Whereas Pittsburgh has the opposite problem: plenty of nerds, but no rich people. The top US Computer Science departments are said to be MIT, Stanford, Berkeley, and Carnegie-Mellon. MIT yielded Route 128. Stanford and Berkeley yielded Silicon Valley. But Carnegie-Mellon? The record skips at that point. Lower down the list, the University of Washington yielded a high-tech community in Seattle, and the University of Texas at Austin yielded one in Austin. But what happened in Pittsburgh? And in Ithaca, home of Cornell, which is also high on the list?

I grew up in Pittsburgh and went to college at Cornell, so I can answer for both. The weather is terrible, particularly in winter, and there's no interesting old city to make up for it, as there is in Boston. Rich people don't want to live in Pittsburgh or Ithaca. So while there are plenty of hackers who could start startups, there's no one to invest in them.

Do you really need the rich people? Wouldn't it work to have the government invest in the nerds? No, it would not. Startup investors are a distinct type of rich people. They tend to have a lot of experience themselves in the technology business. This (a) helps them pick the right startups, and (b) means they can supply advice and connections as well as money. And the fact that they have a personal stake in the outcome makes them really pay attention.

Bureaucrats by their nature are the exact opposite sort of people from startup investors. The idea of them making startup investments is comic. It would be like mathematicians running Vogue -- or perhaps more accurately, Vogue editors running a math journal. [2]

Though indeed, most things bureaucrats do, they do badly. We just don't notice usually, because they only have to compete against other bureaucrats. But as startup investors they'd have to compete against pros with a great deal more experience and motivation.

Even corporations that have in-house VC groups generally forbid them to make their own investment decisions. Most are only allowed to invest in deals where some reputable private VC firm is willing to act as lead investor.

If you go to see Silicon Valley, what you'll see are buildings. But it's the people that make it Silicon Valley, not the buildings. I read occasionally about attempts to set up "technology parks" in other places, as if the active ingredient of Silicon Valley were the office space. An article about Sophia Antipolis bragged that companies there included Cisco, Compaq, IBM, NCR, and Nortel. Don't the French realize these aren't startups?

Building office buildings for technology companies won't get you a silicon valley, because the key stage in the life of a startup happens before they want that kind of space. The key stage is when they're three guys operating out of an apartment. Wherever the startup is when it gets funded, it will stay. The defining quality of Silicon Valley is not that Intel or Apple or Google have offices there, but that they were started there.

So if you want to reproduce Silicon Valley, what you need to reproduce is those two or three founders sitting around a kitchen table deciding to start a company. And to reproduce that you need those people.

The exciting thing is, all you need are the people. If you could attract a critical mass of nerds and investors to live somewhere, you could reproduce Silicon Valley. And both groups are highly mobile. They'll go where life is good. So what makes a place good to them?

What nerds like is other nerds. Smart people will go wherever other smart people are. And in particular, to great universities. In theory there could be other ways to attract them, but so far universities seem to be indispensable. Within the US, there are no technology hubs without first-rate universities-- or at least, first-rate computer science departments.

So if you want to make a silicon valley, you not only need a university, but one of the top handful in the world. It has to be good enough to act as a magnet, drawing the best people from thousands of miles away. And that means it has to stand up to existing magnets like MIT and Stanford.