Paul Graham Essays Growth Hormones

September 2012

A startup is a company designed to grow fast. Being newly founded does not in itself make a company a startup. Nor is it necessary for a startup to work on technology, or take venture funding, or have some sort of "exit." The only essential thing is growth. Everything else we associate with startups follows from growth.

If you want to start one it's important to understand that. Startups are so hard that you can't be pointed off to the side and hope to succeed. You have to know that growth is what you're after. The good news is, if you get growth, everything else tends to fall into place. Which means you can use growth like a compass to make almost every decision you face.


Let's start with a distinction that should be obvious but is often overlooked: not every newly founded company is a startup. Millions of companies are started every year in the US. Only a tiny fraction are startups. Most are service businesses—restaurants, barbershops, plumbers, and so on. These are not startups, except in a few unusual cases. A barbershop isn't designed to grow fast. Whereas a search engine, for example, is.

When I say startups are designed to grow fast, I mean it in two senses. Partly I mean designed in the sense of intended, because most startups fail. But I also mean startups are different by nature, in the same way a redwood seedling has a different destiny from a bean sprout.

That difference is why there's a distinct word, "startup," for companies designed to grow fast. If all companies were essentially similar, but some through luck or the efforts of their founders ended up growing very fast, we wouldn't need a separate word. We could just talk about super-successful companies and less successful ones. But in fact startups do have a different sort of DNA from other businesses. Google is not just a barbershop whose founders were unusually lucky and hard-working. Google was different from the beginning.

To grow rapidly, you need to make something you can sell to a big market. That's the difference between Google and a barbershop. A barbershop doesn't scale.

For a company to grow really big, it must (a) make something lots of people want, and (b) reach and serve all those people. Barbershops are doing fine in the (a) department. Almost everyone needs their hair cut. The problem for a barbershop, as for any retail establishment, is (b). A barbershop serves customers in person, and few will travel far for a haircut. And even if they did the barbershop couldn't accomodate them. [1]

Writing software is a great way to solve (b), but you can still end up constrained in (a). If you write software to teach Tibetan to Hungarian speakers, you'll be able to reach most of the people who want it, but there won't be many of them. If you make software to teach English to Chinese speakers, however, you're in startup territory.

Most businesses are tightly constrained in (a) or (b). The distinctive feature of successful startups is that they're not.


It might seem that it would always be better to start a startup than an ordinary business. If you're going to start a company, why not start the type with the most potential? The catch is that this is a (fairly) efficient market. If you write software to teach Tibetan to Hungarians, you won't have much competition. If you write software to teach English to Chinese speakers, you'll face ferocious competition, precisely because that's such a larger prize. [2]

The constraints that limit ordinary companies also protect them. That's the tradeoff. If you start a barbershop, you only have to compete with other local barbers. If you start a search engine you have to compete with the whole world.

The most important thing that the constraints on a normal business protect it from is not competition, however, but the difficulty of coming up with new ideas. If you open a bar in a particular neighborhood, as well as limiting your potential and protecting you from competitors, that geographic constraint also helps define your company. Bar + neighborhood is a sufficient idea for a small business. Similarly for companies constrained in (a). Your niche both protects and defines you.

Whereas if you want to start a startup, you're probably going to have to think of something fairly novel. A startup has to make something it can deliver to a large market, and ideas of that type are so valuable that all the obvious ones are already taken.

That space of ideas has been so thoroughly picked over that a startup generally has to work on something everyone else has overlooked. I was going to write that one has to make a conscious effort to find ideas everyone else has overlooked. But that's not how most startups get started. Usually successful startups happen because the founders are sufficiently different from other people that ideas few others can see seem obvious to them. Perhaps later they step back and notice they've found an idea in everyone else's blind spot, and from that point make a deliberate effort to stay there. [3] But at the moment when successful startups get started, much of the innovation is unconscious.

What's different about successful founders is that they can see different problems. It's a particularly good combination both to be good at technology and to face problems that can be solved by it, because technology changes so rapidly that formerly bad ideas often become good without anyone noticing. Steve Wozniak's problem was that he wanted his own computer. That was an unusual problem to have in 1975. But technological change was about to make it a much more common one. Because he not only wanted a computer but knew how to build them, Wozniak was able to make himself one. And the problem he solved for himself became one that Apple solved for millions of people in the coming years. But by the time it was obvious to ordinary people that this was a big market, Apple was already established.

Google has similar origins. Larry Page and Sergey Brin wanted to search the web. But unlike most people they had the technical expertise both to notice that existing search engines were not as good as they could be, and to know how to improve them. Over the next few years their problem became everyone's problem, as the web grew to a size where you didn't have to be a picky search expert to notice the old algorithms weren't good enough. But as happened with Apple, by the time everyone else realized how important search was, Google was entrenched.

That's one connection between startup ideas and technology. Rapid change in one area uncovers big, soluble problems in other areas. Sometimes the changes are advances, and what they change is solubility. That was the kind of change that yielded Apple; advances in chip technology finally let Steve Wozniak design a computer he could afford. But in Google's case the most important change was the growth of the web. What changed there was not solubility but bigness.

The other connection between startups and technology is that startups create new ways of doing things, and new ways of doing things are, in the broader sense of the word, new technology. When a startup both begins with an idea exposed by technological change and makes a product consisting of technology in the narrower sense (what used to be called "high technology"), it's easy to conflate the two. But the two connections are distinct and in principle one could start a startup that was neither driven by technological change, nor whose product consisted of technology except in the broader sense. [4]


How fast does a company have to grow to be considered a startup? There's no precise answer to that. "Startup" is a pole, not a threshold. Starting one is at first no more than a declaration of one's ambitions. You're committing not just to starting a company, but to starting a fast growing one, and you're thus committing to search for one of the rare ideas of that type. But at first you have no more than commitment. Starting a startup is like being an actor in that respect. "Actor" too is a pole rather than a threshold. At the beginning of his career, an actor is a waiter who goes to auditions. Getting work makes him a successful actor, but he doesn't only become an actor when he's successful.

So the real question is not what growth rate makes a company a startup, but what growth rate successful startups tend to have. For founders that's more than a theoretical question, because it's equivalent to asking if they're on the right path.

The growth of a successful startup usually has three phases:

  1. There's an initial period of slow or no growth while the startup tries to figure out what it's doing.

  2. As the startup figures out how to make something lots of people want and how to reach those people, there's a period of rapid growth.

  3. Eventually a successful startup will grow into a big company. Growth will slow, partly due to internal limits and partly because the company is starting to bump up against the limits of the markets it serves. [5]
Together these three phases produce an S-curve. The phase whose growth defines the startup is the second one, the ascent. Its length and slope determine how big the company will be.

The slope is the company's growth rate. If there's one number every founder should always know, it's the company's growth rate. That's the measure of a startup. If you don't know that number, you don't even know if you're doing well or badly.

When I first meet founders and ask what their growth rate is, sometimes they tell me "we get about a hundred new customers a month." That's not a rate. What matters is not the absolute number of new customers, but the ratio of new customers to existing ones. If you're really getting a constant number of new customers every month, you're in trouble, because that means your growth rate is decreasing.

During Y Combinator we measure growth rate per week, partly because there is so little time before Demo Day, and partly because startups early on need frequent feedback from their users to tweak what they're doing. [6]

A good growth rate during YC is 5-7% a week. If you can hit 10% a week you're doing exceptionally well. If you can only manage 1%, it's a sign you haven't yet figured out what you're doing.

The best thing to measure the growth rate of is revenue. The next best, for startups that aren't charging initially, is active users. That's a reasonable proxy for revenue growth because whenever the startup does start trying to make money, their revenues will probably be a constant multiple of active users. [7]


We usually advise startups to pick a growth rate they think they can hit, and then just try to hit it every week. The key word here is "just." If they decide to grow at 7% a week and they hit that number, they're successful for that week. There's nothing more they need to do. But if they don't hit it, they've failed in the only thing that mattered, and should be correspondingly alarmed.

Programmers will recognize what we're doing here. We're turning starting a startup into an optimization problem. And anyone who has tried optimizing code knows how wonderfully effective that sort of narrow focus can be. Optimizing code means taking an existing program and changing it to use less of something, usually time or memory. You don't have to think about what the program should do, just make it faster. For most programmers this is very satisfying work. The narrow focus makes it a sort of puzzle, and you're generally surprised how fast you can solve it.

Focusing on hitting a growth rate reduces the otherwise bewilderingly multifarious problem of starting a startup to a single problem. You can use that target growth rate to make all your decisions for you; anything that gets you the growth you need is ipso facto right. Should you spend two days at a conference? Should you hire another programmer? Should you focus more on marketing? Should you spend time courting some big customer? Should you add x feature? Whatever gets you your target growth rate. [8]

Judging yourself by weekly growth doesn't mean you can look no more than a week ahead. Once you experience the pain of missing your target one week (it was the only thing that mattered, and you failed at it), you become interested in anything that could spare you such pain in the future. So you'll be willing for example to hire another programmer, who won't contribute to this week's growth but perhaps in a month will have implemented some new feature that will get you more users. But only if (a) the distraction of hiring someone won't make you miss your numbers in the short term, and (b) you're sufficiently worried about whether you can keep hitting your numbers without hiring someone new.

It's not that you don't think about the future, just that you think about it no more than necessary.

In theory this sort of hill-climbing could get a startup into trouble. They could end up on a local maximum. But in practice that never happens. Having to hit a growth number every week forces founders to act, and acting versus not acting is the high bit of succeeding. Nine times out of ten, sitting around strategizing is just a form of procrastination. Whereas founders' intuitions about which hill to climb are usually better than they realize. Plus the maxima in the space of startup ideas are not spiky and isolated. Most fairly good ideas are adjacent to even better ones.

The fascinating thing about optimizing for growth is that it can actually discover startup ideas. You can use the need for growth as a form of evolutionary pressure. If you start out with some initial plan and modify it as necessary to keep hitting, say, 10% weekly growth, you may end up with a quite different company than you meant to start. But anything that grows consistently at 10% a week is almost certainly a better idea than you started with.

There's a parallel here to small businesses. Just as the constraint of being located in a particular neighborhood helps define a bar, the constraint of growing at a certain rate can help define a startup.

You'll generally do best to follow that constraint wherever it leads rather than being influenced by some initial vision, just as a scientist is better off following the truth wherever it leads rather than being influenced by what he wishes were the case. When Richard Feynman said that the imagination of nature was greater than the imagination of man, he meant that if you just keep following the truth you'll discover cooler things than you could ever have made up. For startups, growth is a constraint much like truth. Every successful startup is at least partly a product of the imagination of growth. [9]


It's hard to find something that grows consistently at several percent a week, but if you do you may have found something surprisingly valuable. If we project forward we see why.


A company that grows at 1% a week will grow 1.7x a year, whereas a company that grows at 5% a week will grow 12.6x. A company making $1000 a month (a typical number early in YC) and growing at 1% a week will 4 years later be making $7900 a month, which is less than a good programmer makes in salary in Silicon Valley. A startup that grows at 5% a week will in 4 years be making $25 million a month. [10]

Our ancestors must rarely have encountered cases of exponential growth, because our intutitions are no guide here. What happens to fast growing startups tends to surprise even the founders.

Small variations in growth rate produce qualitatively different outcomes. That's why there's a separate word for startups, and why startups do things that ordinary companies don't, like raising money and getting acquired. And, strangely enough, it's also why they fail so frequently.

Considering how valuable a successful startup can become, anyone familiar with the concept of expected value would be surprised if the failure rate weren't high. If a successful startup could make a founder $100 million, then even if the chance of succeeding were only 1%, the expected value of starting one would be $1 million. And the probability of a group of sufficiently smart and determined founders succeeding on that scale might be significantly over 1%. For the right people—e.g. the young Bill Gates—the probability might be 20% or even 50%. So it's not surprising that so many want to take a shot at it. In an efficient market, the number of failed startups should be proportionate to the size of the successes. And since the latter is huge the former should be too. [11]

What this means is that at any given time, the great majority of startups will be working on something that's never going to go anywhere, and yet glorifying their doomed efforts with the grandiose title of "startup."

This doesn't bother me. It's the same with other high-beta vocations, like being an actor or a novelist. I've long since gotten used to it. But it seems to bother a lot of people, particularly those who've started ordinary businesses. Many are annoyed that these so-called startups get all the attention, when hardly any of them will amount to anything.

If they stepped back and looked at the whole picture they might be less indignant. The mistake they're making is that by basing their opinions on anecdotal evidence they're implicitly judging by the median rather than the average. If you judge by the median startup, the whole concept of a startup seems like a fraud. You have to invent a bubble to explain why founders want to start them or investors want to fund them. But it's a mistake to use the median in a domain with so much variation. If you look at the average outcome rather than the median, you can understand why investors like them, and why, if they aren't median people, it's a rational choice for founders to start them.


Why do investors like startups so much? Why are they so hot to invest in photo-sharing apps, rather than solid money-making businesses? Not only for the obvious reason.

The test of any investment is the ratio of return to risk. Startups pass that test because although they're appallingly risky, the returns when they do succeed are so high. But that's not the only reason investors like startups. An ordinary slower-growing business might have just as good a ratio of return to risk, if both were lower. So why are VCs interested only in high-growth companies? The reason is that they get paid by getting their capital back, ideally after the startup IPOs, or failing that when it's acquired.

The other way to get returns from an investment is in the form of dividends. Why isn't there a parallel VC industry that invests in ordinary companies in return for a percentage of their profits? Because it's too easy for people who control a private company to funnel its revenues to themselves (e.g. by buying overpriced components from a supplier they control) while making it look like the company is making little profit. Anyone who invested in private companies in return for dividends would have to pay close attention to their books.

The reason VCs like to invest in startups is not simply the returns, but also because such investments are so easy to oversee. The founders can't enrich themselves without also enriching the investors. [12]

Why do founders want to take the VCs' money? Growth, again. The constraint between good ideas and growth operates in both directions. It's not merely that you need a scalable idea to grow. If you have such an idea and don't grow fast enough, competitors will. Growing too slowly is particularly dangerous in a business with network effects, which the best startups usually have to some degree.

Almost every company needs some amount of funding to get started. But startups often raise money even when they are or could be profitable. It might seem foolish to sell stock in a profitable company for less than you think it will later be worth, but it's no more foolish than buying insurance. Fundamentally that's how the most successful startups view fundraising. They could grow the company on its own revenues, but the extra money and help supplied by VCs will let them grow even faster. Raising money lets you choose your growth rate.

Money to grow faster is always at the command of the most successful startups, because the VCs need them more than they need the VCs. A profitable startup could if it wanted just grow on its own revenues. Growing slower might be slightly dangerous, but chances are it wouldn't kill them. Whereas VCs need to invest in startups, and in particular the most successful startups, or they'll be out of business. Which means that any sufficiently promising startup will be offered money on terms they'd be crazy to refuse. And yet because of the scale of the successes in the startup business, VCs can still make money from such investments. You'd have to be crazy to believe your company was going to become as valuable as a high growth rate can make it, but some do.

Pretty much every successful startup will get acquisition offers too. Why? What is it about startups that makes other companies want to buy them? [13]

Fundamentally the same thing that makes everyone else want the stock of successful startups: a rapidly growing company is valuable. It's a good thing eBay bought Paypal, for example, because Paypal is now responsible for 43% of their sales and probably more of their growth.

But acquirers have an additional reason to want startups. A rapidly growing company is not merely valuable, but dangerous. If it keeps expanding, it might expand into the acquirer's own territory. Most product acquisitions have some component of fear. Even if an acquirer isn't threatened by the startup itself, they might be alarmed at the thought of what a competitor could do with it. And because startups are in this sense doubly valuable to acquirers, acquirers will often pay more than an ordinary investor would. [14]


The combination of founders, investors, and acquirers forms a natural ecosystem. It works so well that those who don't understand it are driven to invent conspiracy theories to explain how neatly things sometimes turn out. Just as our ancestors did to explain the apparently too neat workings of the natural world. But there is no secret cabal making it all work.

If you start from the mistaken assumption that Instagram was worthless, you have to invent a secret boss to force Mark Zuckerberg to buy it. To anyone who knows Mark Zuckerberg that is the reductio ad absurdum of the initial assumption. The reason he bought Instagram was that it was valuable and dangerous, and what made it so was growth.

If you want to understand startups, understand growth. Growth drives everything in this world. Growth is why startups usually work on technology—because ideas for fast growing companies are so rare that the best way to find new ones is to discover those recently made viable by change, and technology is the best source of rapid change. Growth is why it's a rational choice economically for so many founders to try starting a startup: growth makes the successful companies so valuable that the expected value is high even though the risk is too. Growth is why VCs want to invest in startups: not just because the returns are high but also because generating returns from capital gains is easier to manage than generating returns from dividends. Growth explains why the most successful startups take VC money even if they don't need to: it lets them choose their growth rate. And growth explains why successful startups almost invariably get acquisition offers. To acquirers a fast-growing company is not merely valuable but dangerous too.

It's not just that if you want to succeed in some domain, you have to understand the forces driving it. Understanding growth is what starting a startup consists of. What you're really doing (and to the dismay of some observers, all you're really doing) when you start a startup is committing to solve a harder type of problem than ordinary businesses do. You're committing to search for one of the rare ideas that generates rapid growth. Because these ideas are so valuable, finding one is hard. The startup is the embodiment of your discoveries so far. Starting a startup is thus very much like deciding to be a research scientist: you're not committing to solve any specific problem; you don't know for sure which problems are soluble; but you're committing to try to discover something no one knew before. A startup founder is in effect an economic research scientist. Most don't discover anything that remarkable, but some discover relativity.


[1] Strictly speaking it's not lots of customers you need but a big market, meaning a high product of number of customers times how much they'll pay. But it's dangerous to have too few customers even if they pay a lot, or the power that individual customers have over you could turn you into a de facto consulting firm. So whatever market you're in, you'll usually do best to err on the side of making the broadest type of product for it.

[2] One year at Startup School David Heinemeier Hansson encouraged programmers who wanted to start businesses to use a restaurant as a model. What he meant, I believe, is that it's fine to start software companies constrained in (a) in the same way a restaurant is constrained in (b). I agree. Most people should not try to start startups.

[3] That sort of stepping back is one of the things we focus on at Y Combinator. It's common for founders to have discovered something intuitively without understanding all its implications. That's probably true of the biggest discoveries in any field.

[4] I got it wrong in "How to Make Wealth" when I said that a startup was a small company that takes on a hard technical problem. That is the most common recipe but not the only one.

[5] In principle companies aren't limited by the size of the markets they serve, because they could just expand into new markets. But there seem to be limits on the ability of big companies to do that. Which means the slowdown that comes from bumping up against the limits of one's markets is ultimately just another way in which internal limits are expressed.

It may be that some of these limits could be overcome by changing the shape of the organization—specifically by sharding it.

[6] This is, obviously, only for startups that have already launched or can launch during YC. A startup building a new database will probably not do that. On the other hand, launching something small and then using growth rate as evolutionary pressure is such a valuable technique that any company that could start this way probably should.

[7] If the startup is taking the Facebook/Twitter route and building something they hope will be very popular but from which they don't yet have a definite plan to make money, the growth rate has to be higher, even though it's a proxy for revenue growth, because such companies need huge numbers of users to succeed at all.

Beware too of the edge case where something spreads rapidly but the churn is high as well, so that you have good net growth till you run through all the potential users, at which point it suddenly stops.

[8] Within YC when we say it's ipso facto right to do whatever gets you growth, it's implicit that this excludes trickery like buying users for more than their lifetime value, counting users as active when they're really not, bleeding out invites at a regularly increasing rate to manufacture a perfect growth curve, etc. Even if you were able to fool investors with such tricks, you'd ultimately be hurting yourself, because you're throwing off your own compass.

[9] Which is why it's such a dangerous mistake to believe that successful startups are simply the embodiment of some brilliant initial idea. What you're looking for initially is not so much a great idea as an idea that could evolve into a great one. The danger is that promising ideas are not merely blurry versions of great ones. They're often different in kind, because the early adopters you evolve the idea upon have different needs from the rest of the market. For example, the idea that evolves into Facebook isn't merely a subset of Facebook; the idea that evolves into Facebook is a site for Harvard undergrads.

[10] What if a company grew at 1.7x a year for a really long time? Could it not grow just as big as any successful startup? In principle yes, of course. If our hypothetical company making $1000 a month grew at 1% a week for 19 years, it would grow as big as a company growing at 5% a week for 4 years. But while such trajectories may be common in, say, real estate development, you don't see them much in the technology business. In technology, companies that grow slowly tend not to grow as big.

[11] Any expected value calculation varies from person to person depending on their utility function for money. I.e. the first million is worth more to most people than subsequent millions. How much more depends on the person. For founders who are younger or more ambitious the utility function is flatter. Which is probably part of the reason the founders of the most successful startups of all tend to be on the young side.

[12] More precisely, this is the case in the biggest winners, which is where all the returns come from. A startup founder could pull the same trick of enriching himself at the company's expense by selling them overpriced components. But it wouldn't be worth it for the founders of Google to do that. Only founders of failing startups would even be tempted, but those are writeoffs from the VCs' point of view anyway.

[13] Acquisitions fall into two categories: those where the acquirer wants the business, and those where the acquirer just wants the employees. The latter type is sometimes called an HR acquisition. Though nominally acquisitions and sometimes on a scale that has a significant effect on the expected value calculation for potential founders, HR acquisitions are viewed by acquirers as more akin to hiring bonuses.

[14] I once explained this to some founders who had recently arrived from Russia. They found it novel that if you threatened a company they'd pay a premium for you. "In Russia they just kill you," they said, and they were only partly joking. Economically, the fact that established companies can't simply eliminate new competitors may be one of the most valuable aspects of the rule of law. And so to the extent we see incumbents suppressing competitors via regulations or patent suits, we should worry, not because it's a departure from the rule of law per se but from what the rule of law is aiming at.

Thanks to Sam Altman, Marc Andreessen, Paul Buchheit, Patrick Collison, Jessica Livingston, Geoff Ralston, and Harj Taggar for reading drafts of this.

One balmy May evening, thirty of Silicon Valley’s top entrepreneurs gathered in a private room at the Berlinetta Lounge, in San Francisco. Paul Graham considered the founders of Instacart, DoorDash, Docker, and Stripe, in their hoodies and black jeans, and said, “This is Silicon Valley, right here.” All the founders were graduates of Y Combinator, the startup “accelerator” that Graham co-founded: a three-month boot camp, run twice a year, in how to become a “unicorn”—Valleyspeak for a billion-dollar company. Thirteen thousand fledgling software companies applied to Y Combinator this year, and two hundred and forty were accepted, making it more than twice as hard to get into as Stanford University. After graduating thirteen hundred startups, YC now boasts the power—and the peculiarities—of an island nation.

At the noisy end of the room, Graham was cheerfully encouraging improbable schemes. At the quiet end, Sam Altman was absorbed in private calculations. When founders came over to talk, he’d train his green eyes on them, listen to their propositions, then crisply observe, “What everyone gets wrong about that is . . .” In 2014, Graham chose Altman—who, at thirty-one, is twenty years his junior—to succeed him as Y Combinator’s president. The two men share a close friendship, a religious zeal for YC, and an inexplicable fondness for cargo shorts. But, where Graham proposes, Altman disposes. At Graham’s table, he and others discussed how to stop Donald Trump, then decided to reach out to an affiliated expert: Chris Lehane, a former White House lawyer now at the YC company Airbnb. Altman declared, “The best idea seems to be just to support Hillary.”

At a hundred and thirty pounds, Altman is poised as a clothespin, fierce as a horned owl. Even in a Valley that worships productivity, he is an outlier, plowing through e-mails and meetings as if strapped to a time bomb, his unblinking stare speeding up colleagues until they sound like chipmunks. Though he is given to gee-whizzery about anything “super awesome”—Small amounts of radiation are actually good for you! It’s called radiation hormesis!—he has scant interest in the specifics of the apps that many YC companies produce; what intrigues him is their potential effect on the world. To determine that, he’ll upload all he needs to know about, say, urban planning or nuclear fusion. Patrick Collison, the C.E.O. of the electronic-payments company Stripe, likened Altman’s brain to the claw machine on a carnival midway: “It roams around but has the ability to plunge very deep when necessary.”

A blogger recently asked Altman, “How has having Asperger’s helped and hurt you?” Altman told me, “I was, like, ‘Fuck you, I don’t have Asperger’s!’ But then I thought, I can see why he thinks I do. I sit in weird ways”—he folds up like a busted umbrella—“I have narrow interests in technology, I have no patience for things I’m not interested in: parties, most people. When someone examines a photo and says, ‘Oh, he’s feeling this and this and this,’ all these subtle emotions, I look on with alien intrigue.” Altman’s great strengths are clarity of thought and an intuitive grasp of complex systems. His great weakness is his utter lack of interest in ineffective people, which unfortunately includes most of us. I found his assiduousness alarming at first, then gradually endearing. When I remarked, after a few long days together, that he never seemed to visit the men’s room, he said, “I will practice going to the bathroom more often so you humans don’t realize that I’m the A.I.”

When he took over YC, he inherited a budding colossus. The venture capitalist Chris Dixon told me, “They created the greatest business model of all time. For basically no money”— YC gives each company just a hundred and twenty thousand dollars, to cover expenses—“they get seven per cent of a lot of the best startups in Silicon Valley!” Collectively, YC companies are worth eighty billion dollars, a valuation that has grown seventeenfold in the past five years.

Yet Altman decided to retool nearly everything. At the Berlinetta Lounge, as he picked at the vegetarian plate, he observed that a change in C.E.O.s works only if the new leader “re-founds” the company. “I very intentionally did that with YC,” he said. After conferring with the accelerator’s sixteen other partners, Altman launched an initiative to support startups even earlier in their life span, and a fund to continue investing in them as they grow. YC would no longer waft explorers out to sea in rickety ships but launch ironclad armadas to claim an empire. And it would mold not a few hundred companies a year but a thousand, then ten thousand.

Like everyone in Silicon Valley, Altman professes to want to save the world; unlike almost everyone there, he has a plan to do it. “YC somewhat gets to direct the course of technology,” he said. “Consumers decide, ultimately, but enough people view YC as important that if we say, ‘We’re super excited about virtual reality,’ college students will start studying it.” Soon after taking over, he wrote a blog post declaring that “science seems broken” and calling for applications from companies in energy, biotech, artificial intelligence, robotics, and eight other fields. As a result, the once nerdy Y Combinator is now aggressively geeky. Across the table from Altman at dinner, the C.E.O. of a nuclear-fission startup was urging the founder of a quantum-computing startup to get his artificial-atom-based machine to market: “These computers would shorten our product-development cycle 10 to 20x!”

Two YC partners sat Altman down last year “and told him, ‘Slow down, chill out!’ ” another partner, Jonathan Levy, told me. “Sam said, ‘Yes, you’re right!’—and went off and did something else on the side that we didn’t know about for a while.” That was YC Research, a nonprofit, initially funded with a ten-million-dollar personal gift, to conduct pure research into moon-shot ideas. Altman also co-founded, with Elon Musk, the C.E.O. of Tesla and SpaceX, a nonprofit called OpenAI, whose goal is to prevent artificial intelligence from accidentally wiping out humanity. The venture capitalist Marc Andreessen said, “Under Sam, the level of YC’s ambition has gone up 10x.” Paul Graham, who was leaving soon after the dinner for a sabbatical year in England, told me that Altman, by precipitating progress in “curing cancer, fusion, supersonic airliners, A.I.,” was trying to comprehensively revise the way we live: “I think his goal is to make the whole future.”

Altman is rapidly building out an economy within Silicon Valley that seems intended to essentially supplant Silicon Valley—a guild of hyper-capitalist entrepreneurs who will help one another fix the broken world. Everyone has cautioned him against it. The Valley prizes overweening ambition but expects it to be “rifle-focussed” on making the world’s best houseboat-rental platform or Cognac-delivery service. Reid Hoffman, a leading venture capitalist, warned, “It’s great they’re being ambitious, but classically, in the Valley, when people try to reinvent an area it’s ended very badly.” Altman, as he nursed a negroni after dinner, had his own warning for the timid: “Democracy only works in a growing economy. Without a return to economic growth, the democratic experiment will fail. And I have to think that YC is hugely important to that growth.”

Launching a startup in 2016 is akin to assembling an alt-rock band in 1996 or protesting the Vietnam War in 1971—an act of youthful rebellion gone conformist. Since 2005, the year Y Combinator began, accelerators have sprung up everywhere to help transform startups from a skein of code into a bona-fide company. In exchange for five to seven per cent of a startup’s equity, an accelerator usually provides between fifteen thousand and a hundred thousand dollars, advice for an intensive three-month period, introductions to mentors, and a Demo Day, when investors assess the finished product. The U.S. has a hundred and sixty accelerators—three in Chattanooga alone—and there are thousands more around the world, including Brainnovations, outside Tel Aviv, and the Startup Sauna, in locations from Minsk to Dar es Salaam.

In the nineties, before the accelerator era, startups were usually launched by mid-career engineers or repeat entrepreneurs, who sought millions in venture capital and then labored in secret on something complicated that took years to launch. As the price of Web hosting plummeted and PCs and cell phones proliferated, college and grad-school dropouts like Mark Zuckerberg or Larry Page and Sergey Brin could suddenly conjure unicorns on their laptops. Paul Graham, a gifted programmer who’d sold his own startup to Yahoo for fifty million dollars, was one of the first people to harness these trends. His 2005 essay “How to Start a Startup”—together with Steven Blank’s “The Four Steps to the Epiphany” and Eric Ries’s “The Lean Startup”—helped to codify the modern entrepreneur’s ethos: bootstrap; begin with a “minimum viable product” and iterate rapidly; prefer ten people loving what you make to ten thousand liking it.

Graham and his wife and two friends started Y Combinator (named for an obscure mathematical function) in Cambridge, Massachusetts, as both a summer-vacation experiment in investing and a radical stab at reimagining the summer job. In his book “Hackers & Painters,” Graham calculated that smart hackers at a startup could get 36x more work done than the average office drone—and that they would, therefore, eventually blow up employment as we know it. He made this sound patriotic and fun; how could an oligarchic technocracy go wrong? “Hackers are unruly,” he wrote. “That is the essence of hacking. And it is also the essence of American-ness.” [cartoon id="a20328"]

Graham could gauge applicants’ technical skills, and his wife, Jessica Livingston, was a remarkable judge of character. They prized people in their mid-twenties, an age at which, Graham wrote, your advantages include “stamina, poverty, rootlessness, colleagues, and ignorance.” The first group of eight companies—which included a mobile app, Loopt, founded by Sam Altman and two friends—got six thousand dollars per founder, along with Graham’s advice and home-cooked chicken fricassee, and the promise that at the end of the summer they could pitch his wealthy friends for fifteen minutes. That batch had Reddit, now valued at six hundred million dollars, and a batch two years later had Dropbox, valued at ten billion.

In a class that Altman taught at Stanford in 2014, he remarked that the formula for estimating a startup’s chance of success is “something like Idea times Product times Execution times Team times Luck, where Luck is a random number between zero and ten thousand.” The rise of Airbnb, now valued at thirty billion dollars, seems replete with luck. When it arrived at YC, in 2009, it had been making more money selling novelty cereals—Obama O’s and Cap’n McCains—than booking bed-and-breakfast reservations. Graham thought the founders’ idea was so unpromising that he tried to convince them to do payments instead. And the event that transformed the company, turning it into a combine that harvested living space around the globe, was a fluke: Barry Manilow’s drummer went on tour and asked if he could rent out his place without being present to provide breakfast.

Yet Airbnb’s C.E.O., Brian Chesky, attributes much of the company’s success to Y Combinator. “When we entered YC, it wasn’t at all clear that we would exist after it,” Chesky said. “And by the end it was: ‘Can we be the next marketplace, the next eBay?’ ” Part of this exponential increase in ambition occurred when the founders showed Altman, who was then an unpaid mentor and fund-raising expert at YC, a slide deck they hoped would secure a five-hundred-thousand-dollar round of seed, or initial, funding. (Companies usually raise a seed round after YC, an A round once they’ve reached a real milestone, then a B round, and so on.) “We had limited our projected revenue to thirty million dollars,” Chesky said. “Sam said, ‘Take all the “M”s and make them “B”s.’ ” Altman recalls telling them, “Either you don’t believe everything you said in the rest of the deck, or you’re ashamed, or I can’t do math.”

A 2012 study of North American accelerators found that almost half of them had failed to produce a single startup that went on to raise venture funding. While a few accelerators, such as Tech Stars and 500 Startups, have a handful of alumni worth hundreds of millions of dollars, Y Combinator has graduates worth at least a billion—and it has eleven of them. The angel investor Ron Conway, who has provided funding for hundreds of YC companies, told me that the accelerator was tech’s Tomorrowland: “When my team met Airbnb at YC, that was the first time we thought about the sharing economy. And when we met DoorDash and Instacart we said, ‘Oh, my God, there’s a thing called the on-demand economy!’ ”

As YC grew, it moved to an exurban office cube in the town of Mountain View, an hour south of San Francisco, where it shared space with a company called Anybots. (The founders had to keep alert to avoid being crushed by lumbering robots.) The accelerator soon expanded into a second cube, across the street, and now it’s near the fire-code limit in that building, too. It’s a measure of Altman’s ambition that he compares YC to Alphabet, Google’s parent company, which is also composed of independent units that collaborate, and likewise has a moon-shot division, the X research group. He recently tweeted that YC’s empire had reached fourteen per cent of the value of Alphabet—whose market cap is among the world’s largest—adding, “Long way to go. . . .” It’s a blatantly unfair comparison: YC’s average ownership of its companies, diluted by subsequent venture funding, is just three per cent. Yet Altman told me that, “unlike Google, we grow faster as we get bigger. We could catch them in ten years.”

As the sun set over Atherton, the loveliest town in unlovely Silicon Valley, Altman challenged Geoff Ralston, another YC partner, to a table-tennis match by Ralston’s pool. They were about to host a party for the thirty-two companies they were responsible for in the winter batch; the goal was to relieve the pressure of Demo Day, which was three weeks off, in late March. That was when the founders would have two and a half minutes to wow investors—six hundred in the room and twenty-five hundred more watching online. Two founders were already having anxiety attacks.

Altman was raising his arms in victory as the founders began to amble in and gaze around at startup Valhalla: a seven-thousand-square-foot mansion, catered food under a grapefruit tree festooned with lights, a back yard that seemed to stretch to Redwood City. (Ralston made his fortune building what became Yahoo Mail.) Luke Miles, the eighteen-year-old founder of Restocks, tried not to look overawed. Restocks is a messaging service for young “super-consumers” who want to know, five minutes ahead of everyone else, when small shipments of Supreme T-shirts and Yeezy Boost 350 shoes go on sale. Miles was accepted to YC after excelling in YC Fellowship, the new program for embryonic startups, in which they receive as much as twenty thousand dollars and the opportunity to consult with a partner over Skype. Miles said, “The money they gave me was enough to prove to my parents that I wasn’t wasting my life dropping out of college.”

For many founders, YC provides the university experience they wish they could have had. Michael Seibel, a YC partner who was recently put in charge of the batches, and who also went through the program twice, said, “P.G.”—as Paul Graham is known—“used to tell everyone at the beginning of the batch, ‘Some of these people will be in your wedding,’ which is a weird thing to say to three hundred strangers. But almost all my groomsmen were in YC. What does that remind you of? College.” Y Combinator’s founders come to the building for group office hours on alternate Tuesdays, as well as for individual office hours as needed with their assigned partners (the professors), then stick around to scarf down bowls of pasta at long Formica tables (the dining hall) and listen to eminences such as Marissa Mayer and Mark Zuckerberg (visiting scholars). At the end, they present on Demo Day (defending their thesis), and either raise money (pass) or don’t (fail).

The curriculum is deliberately spartan. Kevin Hale, a YC partner, said, “What we ask of startups is very simple but very hard to do. One, make something people want”—a phrase of Graham’s, which is emblazoned on gray T-shirts for the founders—“and, two, all you should be doing is talking to your customers and building stuff.” Chad Rigetti, the founder of YC’s quantum-computing startup, told me that he kept his office walls a matte white “so my team’s neurons aren’t accidentally firing because of outside stimulation.”

The ethics, too, have a collegiate clarity. YC prides itself on rejecting jerks and bullies. “We’re good at screening out assholes,” Graham told me. “In fact, we’re better at screening out assholes than losers. All of them start off as losers—and some evolve.” The accelerator also suggests that great wealth is a happy by-product of solving an urgent problem. This braiding of altruism and ambition is a signal feature of the Valley’s self-image. Graham wrote an essay, “Mean People Fail,” in which—ignoring such possible counterexamples as Jeff Bezos and Larry Ellison—he declared that “being mean makes you stupid” and discourages good people from working for you. Thus, in startups, “people with a desire to improve the world have a natural advantage.” Win-win.

A founder’s first goal, Graham wrote, is becoming “ramen profitable”: spending thriftily and making just enough to afford ramen noodles for dinner. “You don’t want to give the founders more than they need to survive,” Jessica Livingston said. “Being lean forces you to focus. If a fund offered us three hundred thousand dollars to give the founders, we wouldn’t take it.” (Many of YC’s seventeen partners, wealthy from their own startups, receive a salary of just twenty-four thousand dollars and get most of their compensation in stock.) This logic, followed to its extreme, would suggest that you shouldn’t even take YC’s money, and many successful startups don’t. Only twenty per cent of the Inc. 500, the five hundred fastest-growing private companies, raised outside funding. But the YC credential, and the promise that it will turn you into a juggernaut, can be hard to resist.

Near Ralston’s grapefruit tree, Omer Sadika and Sebastian Wallin nibbled hors d’oeuvres as they compared notes about the stress of trying to launch their security companies, Secful and Castle. “We are sleeping at most five hours a day,” Sadika said. Wallin muttered, “I’ve forgotten what day it is.” Both men were planning to move to the Valley; Sadika from Israel, and Wallin from Malmö*. “The customers are here,” Sadika said. “And you’re one step away from the entrepreneurs at Airbnb and Stripe,” Wallin pointed out. YC provides instant entrée to Silicon Valley—a community that, despite its meritocratic rhetoric, typically requires a “warm intro” from a colleague, who is usually a white man. All the early arrivals at the party were men; the batch’s female founders were attending a presentation on the challenges of being a female founder. YC is more diverse than many institutions in tech, but it knows that it has a ways to go. [cartoon id="a18530"]

On the far side of a fire pit, two founders of Shypmate, an app that links you to airline passengers who will cheaply carry your package to Ghana or Nigeria, were commiserating. Kwadwo Nyarko said, “We’re at the mercy of travellers who never have as much space in their luggage as they said.” Perry Ogwuche murmured, “YC tells us, ‘Talk to your customers,’ but it’s hard to find our customers.” Altman walked over to engage them, dutiful as a birthday-party magician. “So what are your hobbies?” he asked. Nonplussed, Ogwuche said, “We work and we go to the gym. And what are yours?”

“Well, I like racing cars,” Altman said. “I have five, including two McLarens and an old Tesla. I like flying rented planes all over California. Oh, and one odd one—I prep for survival.” Seeing their bewilderment, he explained, “My problem is that when my friends get drunk they talk about the ways the world will end. After a Dutch lab modified the H5N1 bird-flu virus, five years ago, making it super contagious, the chance of a lethal synthetic virus being released in the next twenty years became, well, nonzero. The other most popular scenarios would be A.I. that attacks us and nations fighting with nukes over scarce resources.” The Shypmates looked grave. “I try not to think about it too much,” Altman said. “But I have guns, gold, potassium iodide, antibiotics, batteries, water, gas masks from the Israeli Defense Force, and a big patch of land in Big Sur I can fly to.”

Altman’s mother, a dermatologist named Connie Gibstine, told me, “Sam does keep an awful lot tied up inside. He’ll call and say he has a headache—and he’ll have Googled it, so there’s some cyber-chondria in there, too. I have to reassure him that he doesn’t have meningitis or lymphoma, that it’s just stress.” If the pandemic does come, Altman’s backup plan is to fly with his friend Peter Thiel, the billionaire venture capitalist, to Thiel’s house in New Zealand. Thiel told me, “Sam is not particularly religious, but he is culturally very Jewish—an optimist yet a survivalist, with a sense that things can always go deeply wrong, and that there’s no single place in the world where you’re deeply at home.”

Altman makes a list of goals each year, and he looks at it every few weeks. It always includes a taxing physical objective—a hundred-mile bike ride each week; fifty consecutive pull-ups—and an array of work targets. This year, for YC, those included “Better partnership dynamic; decision on [expanding to] China; figure out how to scale another 2x.” The latest list also contains a reminder to fund videos that demonstrate counterintuitive concepts in physics and quantum mechanics (“QM experiment/physics explainers”) and a prompt to reread a Huffington Post article about the regrets of the dying (“I wish that I had let myself be happier”).

He was always precocious and efficient. As a child, in St. Louis, he grasped the system behind area codes in nursery school, and learned to program and disassemble a Macintosh at eight. The Mac became his lifeline to the world. “Growing up gay in the Midwest in the two-thousands was not the most awesome thing,” he told me. “And finding AOL chat rooms was transformative. Secrets are bad when you’re eleven or twelve.” When he came out to his parents, at sixteen, his mother was astonished. She told me, “Sam had always struck me as just sort of unisexual and tech-y.” After a Christian group boycotted an assembly about sexuality at his prep school, John Burroughs, Altman addressed the whole community, announcing that he was gay and asking whether the school wanted to be a repressive place or one open to different ideas. Madelyn Gray, Altman’s college counsellor, said, “What Sam did changed the school. It felt like someone had opened up a great big box full of all kinds of kids and let them out into the world.”

He attended Stanford University, where he spent two years studying computer science, until he and two classmates dropped out to work full time on Loopt, a mobile app that told your friends where you were. Loopt got into Y Combinator’s first batch because Altman in particular passed what would become known at YC as the young founders’ test: Can this stripling manage adults? He was a formidable operator: quick to smile, but also quick to anger. If you cross him, he’ll joke about slipping ice-nine into your food. (Ice-nine, in Kurt Vonnegut’s “Cat’s Cradle,” annihilates everything it touches that contains water.) Paul Graham, noting Altman’s early aura of consequence, told me, “Sam is extremely good at becoming powerful.”

Altman worked so incessantly that summer that he got scurvy. He excelled at wangling meetings with mobile carriers, and at closing deals with them to feature the app, whose valuation eventually soared to a hundred and seventy-five million dollars. Consumers, though, never bought in. “We had the optimistic view that location would be all-important,” Altman said. “The pessimistic view was that people would lie on their couches and just consume content—and that is what happened. I learned you can’t make humans do something they don’t want to do.” In 2012, he and the other founders sold the company for forty-three million dollars—a negative return for their V.C.s.

One of Altman’s co-founders at Loopt, Nick Sivo, was also his boyfriend; the two dated for nine years, but after the company sold they broke up. “I thought I was going to marry him—very in love with him,” Altman said. At loose ends, he launched a small venture fund, Hydrazine Capital. He raised twenty-one million dollars, including a significant investment from Peter Thiel and much of the five million dollars he’d made selling Loopt, then invested seventy-five per cent of the fund in YC companies. He had a knack for finding opportunity in chaos. Altman told me that he led the B round at Reddit, a chronically disorganized YC graduate, because “you want to invest in messy, somewhat broken companies. You can treat the warts on top, and because of the warts the company will be hugely underpriced.”

Hydrazine has grown tenfold in value in just four years, but, despite its success, Altman recoiled from venture capital. “You’re trying to find a company that will be successful with or without you, then convince them to take your money instead of somebody else’s, and at a lower price,” he said. “I didn’t like being oppositional to the entrepreneur.” Leery of tech’s culture of Golcondan wealth, in which a billion dollars is dismissed as “a buck,” he decided to rid himself of all but a comfortable cushion: his four-bedroom house in San Francisco’s Mission district, his cars, his Big Sur property, and a reserve of ten million dollars, whose annual interest would cover his living expenses. The rest would go to improving humanity.

Like a stymied startup, Altman then made a radical pivot. Paul Graham and Jessica Livingston, who had two young children, were worn out by the work of running YC and had begun looking for a successor. Livingston said, “There wasn’t a list of who should run YC and Sam at the top. It was just: Sam.”

Graham said, “I asked Sam in our kitchen, ‘Do you want to take over YC?,’ and he smiled, like, it worked. I had never seen an uncontrolled smile from Sam. It was like when you throw a ball of paper into the wastebasket across the room—that smile.”

Altman wanted to create a trillion-dollar conglomerate and to move the world forward. And, he realized, “you couldn’t have a trillion-dollar enterprise without major scientific advances.” So he opened the batches to hard tech, studied the science and engineering problems such companies faced, and recruited the most promising ones. Altman helped to persuade Kyle Vogt, the C.E.O. of the self-driving-car company Cruise, to do YC in 2014; afterward, when Cruise had trouble finding funding, he invested three million dollars in the company. In March, General Motors bought Cruise for $1.25 billion.

Altman had long wanted to start his own nuclear-energy company; instead, he had YC fund the best fission and fusion startups he could find. Then he personally invested in both companies and chaired their boards. Thousands of startups are devoted to social interaction, and fewer than twenty to fission and fusion, but, Altman said, “hard things are actually easier than easy things. Because people feel it’s interesting, they want to help. Another mobile app? You get an eye roll. A rocket company? Everyone wants to go to space.”

Graham has written that the two founders he most often invoked when advising startups were Steve Jobs and Altman: “On questions of design, I ask ‘What would Steve do?’ but on questions of strategy or ambition I ask ‘What would Sam do?’ ” Founders in a crisis call Altman first, relying on his knack for high-speed trading in the Valley’s favor bank—“I called Brian and got it sorted out,” he’ll say, referring to Brian Chesky—and his ability to see people as chess pieces and work out their lines of play. One YC founder told me, “Since Sam can see the future, we want him to tell us what’s coming.”

When the two founders of a Norwegian startup called Konsus arrived for office hours with Altman, at YC’s new outpost, in San Francisco’s South of Market district, they were as solemn as pilgrims approaching a mountaintop shrine. Konsus, which was in the winter batch, connects businesses to freelancers who perform tasks such as data entry or Web design. Despite having raised a robust $1.6 million after Demo Day, the founders were ridden with angst. Fredrik Thomassen said that they wanted to make their war chest last forever, and Sondre Rasch mentioned that he’d frugally chosen to live in a twelve-entrepreneur collective in a nearby forest. So: did they really have to buy their engineers computers? The two men, bearded and wraithlike, stared at Altman. [cartoon id="a20341"]

“It is one of the rarer mistakes to make, trying to be too lean,” Altman said, dryly. “But, if someone’s going to make it, it’s going to be Scandinavians. Buy the computers.” The founders nodded intently. They had revered Altman since they met him. At the time, their explanation of what Konsus did was “Companies send us office tasks and we instantly assign them to top-talent freelancers all over the world based on skill and availability.” Altman had immediately asked, “Aren’t you an on-demand temp agency?”

Thomassen now said, “We of course want to talk about quality, because the quality of our freelancers’ work is our differentiator. We need a metric to measure it, somehow.” Altman replied, “Repeat use or customer retention will track that. You don’t need to invent some complicated new metric, so don’t.” Thomassen consulted a list they’d made: “What is the most likely thing we’ll do wrong in the next three months?”

Altman found their hypervigilance encouraging: he believes that “the founders who do best are very paranoid, very full of existential crises.” He told them, “Founders by definition like to start new things. But starting a business means grinding away for ten years.” He continued, without irony, “Most people do too many things. Do a few things relentlessly.”

Altman’s terse prescience led one YC founder to call him “startup Yoda.” Entrepreneurs trudge in to see him burdened by half an hour’s worth of calamities and bounce out after fifteen minutes, springy with resolve. Much of his advice follows YC’s standard imperative to transparency: if you’re worried about investors’ responses to a setback, “just tell them”; if you’re mystified by what a potential customer’s silence portends, “just ask them.” It’s the knottier questions that elicit his cleaving judgments. “Don’t worry about a competitor until they’re beating you in the market,” Altman told the founders of Elucify at his dining table one afternoon. “Competitors are one of the last monsters that haunt your dreams.” A few minutes later, he was on speakerphone with Varden Labs, a Canadian self-driving-vehicle company. As the founders poured out concerns about fund-raising, Altman, in stocking feet, cargo shorts, and a gray hoodie, waved around a Bronze Age sword that he’d bought as a gift for Paul Graham. “To raise fifty million,” he said—slash, two-handed swipe—“you either need a major technological breakthrough or a large customer ready to go.” He parried invisible strokes, advancing inexorably. What about long-term? they asked. “Always think about adding one more zero to whatever you’re doing, but never think beyond that.” Stab to the heart.

Four years ago, on a daylong hike with friends north of San Francisco, Altman relinquished the notion that human beings are singular. As the group discussed advances in artificial intelligence, Altman recognized, he told me, that “there’s absolutely no reason to believe that in about thirteen years we won’t have hardware capable of replicating my brain. Yes, certain things still feel particularly human—creativity, flashes of inspiration from nowhere, the ability to feel happy and sad at the same time—but computers will have their own desires and goal systems. When I realized that intelligence can be simulated, I let the idea of our uniqueness go, and it wasn’t as traumatic as I thought.” He stared off. “There are certain advantages to being a machine. We humans are limited by our input-output rate—we learn only two bits a second, so a ton is lost. To a machine, we must seem like slowed-down whale songs.”

OpenAI, the nonprofit that Altman founded with Elon Musk, is a hedged bet on the end of human predominance—a kind of strategic-defense initiative to protect us from our own creations. OpenAI was born of Musk’s conviction that an A.I. could wipe us out by accident. The problem of managing powerful systems that lack human values is exemplified by “the paperclip maximizer,” a scenario that the Swedish philosopher Nick Bostrom raised in 2003. If you told an omnicompetent A.I. to manufacture as many paper clips as possible, and gave it no other directives, it could mine all of Earth’s resources to make paper clips, including the atoms in our bodies—assuming it didn’t just kill us outright, to make sure that we didn’t stop it from making more paper clips. OpenAI was particularly concerned that Google’s DeepMind Technologies division was seeking a supreme A.I. that could monitor the world for competitors. Musk told me, “If the A.I. that they develop goes awry, we risk having an immortal and superpowerful dictator forever.” He went on, “Murdering all competing A.I. researchers as its first move strikes me as a bit of a character flaw.”

It was clear what OpenAI feared, but less clear what it embraced. In May, Dario Amodei, a leading A.I. researcher then at Google Brain, came to visit the office, and told Altman and Greg Brockman, the C.T.O., that no one understood their mission. They’d raised a billion dollars and hired an impressive team of thirty researchers—but what for? “There are twenty to thirty people in the field, including Nick Bostrom and the Wikipedia article,” Amodei said, “who are saying that the goal of OpenAI is to build a friendly A.I. and then release its source code into the world.”

“We don’t plan to release all of our source code,” Altman said. “But let’s please not try to correct that. That usually only makes it worse.”

“But what is the goal?” Amodei asked.

Brockman said, “Our goal right now . . . is to do the best thing there is to do. It’s a little vague.”

A.I. technology hardly seems almighty yet. After Microsoft launched a chatbot, called Tay, bullying Twitter users quickly taught it to tweet such remarks as “gas the kikes race war now”; the recently released “Daddy’s Car,” the first pop song created by software, sounds like the Beatles, if the Beatles were cyborgs. But, Musk told me, “just because you don’t see killer robots marching down the street doesn’t mean we shouldn’t be concerned.” Apple’s Siri, Amazon’s Alexa, and Microsoft’s Cortana serve millions as aides-de-camp, and simultaneous-translation and self-driving technologies are now taken for granted. Y Combinator has even begun using an A.I. bot, Hal9000, to help it sift admission applications: the bot’s neural net trains itself by assessing previous applications and those companies’ outcomes. “What’s it looking for?” I asked Altman. “I have no idea,” he replied. “That’s the unsettling thing about neural networks—you have no idea what they’re doing, and they can’t tell you.”

OpenAI’s immediate goals, announced in June, include a household robot able to set and clear a table. One longer-term goal is to build a general A.I. system that can pass the Turing test—can convince people, by the way it reasons and reacts, that it is human. Yet Altman believes that a true general A.I. should do more than deceive; it should create, discovering a property of quantum physics or devising a new art form simply to gratify its own itch to know and to make. While many A.I. researchers were correcting errors by telling their systems, “That’s a dog, not a cat,” OpenAI was focussed on having its system teach itself how things work. “Like a baby does?” I asked Altman. “The thing people forget about human babies is that they take years to learn anything interesting,” he said. “If A.I. researchers were developing an algorithm and stumbled across the one for a human baby, they’d get bored watching it, decide it wasn’t working, and shut it down.”

Altman felt that OpenAI’s mission was to babysit its wunderkind until it was ready to be adopted by the world. He’d been reading James Madison’s notes on the Constitutional Convention for guidance in managing the transition. “We’re planning a way to allow wide swaths of the world to elect representatives to a new governance board,” he said. “Because if I weren’t in on this I’d be, like, Why do these fuckers get to decide what happens to me?”

Under Altman, Y Combinator was becoming a kind of shadow United Nations, and increasingly he was making Secretary-General-level decisions. Perhaps it made sense to entrust humanity to someone who doesn’t seem all that interested in humans. “Sam’s program for the world is anchored by ideas, not people,” Peter Thiel said. “And that’s what makes it powerful—because it doesn’t immediately get derailed by questions of popularity.” Of course, that very combination of powerful intent and powerful unconcern is what inspired OpenAI: how can an unfathomable intelligence protect us if it doesn’t care what we think?

This spring, Altman met Ashton Carter, the Secretary of Defense, in a private room at a San Francisco trade show. Altman wore his only suit jacket, a bunchy gray number his assistant had tricked him into getting measured for on a trip to Hong Kong. Carter, in a pin-striped suit, got right to it. “Look, a lot of people out here think we’re big and clunky. And there’s the Snowden overhang thing, too,” he said, referring to the government’s treatment of Edward Snowden. “But we want to work with you in the Valley, tap the expertise.”

“Obviously, that would be great,” Altman said. “You’re probably the biggest customer in the world.” The Defense Department’s proposed research-and-development spending next year is more than double that of Apple, Google, and Intel combined. “But a lot of startups are frustrated that it takes a year to get a response from you.” Carter aimed his forefinger at his temple like a gun and pulled the trigger. Altman continued, “If you could set up a single point of contact, and make decisions on initiating pilot programs with YC companies within two weeks, that would help a lot.”

“Great,” Carter said, glancing at one of his seven aides, who scribbled a note. “What else?”

Altman thought for a while. “If you or one of your deputies could come speak to YC, that would go a long way.”

“I’ll do it myself,” Carter promised. [cartoon id="finck-2015-09-28"]

As everyone filed out, Chris Lynch, a former Microsoft executive who heads Carter’s digital division, told Altman, “It would have been good to talk about OpenAI.” Altman nodded noncommittally. The 2017 U.S. military budget allocates three billion dollars for human-machine collaborations known as Centaur Warfighting, and a long-range missile that will make autonomous targeting decisions is in the pipeline for the following year. Lynch later told me that an OpenAI system would be a natural fit.

Altman was of two minds about handing OpenAI products to Lynch and Carter. “I unabashedly love this country, which is the greatest country in the world,” he said. At Stanford, he worked on a DARPA project involving drone helicopters. “But some things we will never do with the Department of Defense.” He added, “A friend of mine says, ‘The thing that saves us from the Department of Defense is that, though they have a ton of money, they’re not very competent.’ But I feel conflicted, because they have the world’s best cyber command.” Altman, by instinct a cleaner-up of messes, wanted to help strengthen our military—and then to defend the world from its newfound strength.

Nearly all YC startups enter the program with the same funding, and thus the same valuation: $1.7 million. After Demo Day, their mean valuation is ten million. There are several theories about why this estimation jumps nearly sixfold in three months. One is that the best founders apply to the best accelerator, and that YC excels at picking formidable founders who would become successful anyway. Paul Buchheit, who ran the past few batches, said, “It’s all about founders. Facebook had Mark Zuckerberg, and MySpace had a bunch of monkeys.”

The corollary is that Y Combinator makes its companies more desirable by teaching them how to tell their story on Demo Day. The venture capitalist Chris Dixon, who admires YC, said, “The founders are so well coached that they know exactly how to reverse-engineer us, down to demonstrating domain expertise and telling anecdotes about their backgrounds that show perseverance and courage.” In the winter batch, the pitches followed an invariable narrative of imminent magnitude: link yourself to a name-brand unicorn (“We’re Uber for babysitting . . . Stripe for Africa . . . Slack for health care”), or, if there’s no apt analogue, say, “X is broken. In the future, Y will fix X. We’re already doing Y.” Then express your premise as a chewy buzzword sandwich: We “leverage technology to achieve personalization in a fully automated way” (translation: individuated shampoo). Paul Graham cheerfully acknowledged that, by instilling message discipline, “we help the bad founders look indistinguishable from the good ones.”

The counter-theory is that YC actually does make its companies better, by teaching them to focus on growth above all, thereby eliminating distractions such as talking to the tech press or speaking at conferences or making cosmetic coding tweaks. YC’s gold standard for revenue growth is ten per cent a week, which compounds to 142x a year. Failing that, well, tell a story of some other kind of growth. On Demo Day, one company announced that it had enjoyed “fifty-per-cent word-of-mouth growth,” whatever that might be. Sebastian Wallin told me that his security company, Castle, raised $1.8 million because “we managed to find a good metric to show growth. We tried tracking installations of our product, but it didn’t look good. So we used accounts protected, a number that showed roughly thirty-per-cent growth through the course of YC—and about forty per cent of the accounts were YC companies. It was a perfect fairy-tale story.”

The truth is that rapid growth over a long period is rare, that the repeated innovation required to sustain it is nearly impossible, and that certain kinds of uncontrollable growth turn out to be cancers. Last year, after a series of crises at Reddit, Altman, who is on its board, convinced Steve Huffman, the co-founder of the company, to return as C.E.O. Huffman said, “I immediately told Sam, ‘Don’t get on my ass about growth. I’m not in control of it.’ Every great startup—Facebook, Airbnb—has no idea why it’s growing at first, and has to figure that out before the growth stalls. Growth masks all problems.”

Perhaps the most dispositive theory about YC is that the power of its network obviates other theories. Alumni view themselves as a kind of keiretsu, a network of interlocking companies that help one another succeed. “YC is its own economy,” Harj Taggar, the co-founder of Triplebyte, which matches coders’ applications with YC companies, said. Each spring, founders gather at Camp YC, in a redwood forest north of San Francisco, just to network—tech’s version of the Bohemian Grove, only with more vigorous outdoor urination. When Altman first approached Kyle Vogt, the C.E.O. of Cruise, Vogt had been through YC with an earlier company, so he already knew its lessons. He told me, “I talked to five of my friends who had done YC more than once and said, ‘Was it worth it the second time? Are you likely to receive higher valuations because of the brand, and because you’re plugging into the network?’ Across the board, they said yes.”

There really is no counter-theory. “The knock on YC,” Andy Weissman, a managing partner at Union Square Ventures, told me, “is that on Demo Day their users are just YC companies, which entirely explains why they’re all growing so fast. But how great to have more than a thousand companies willing to use your product!” It’s not just that YC startups can get Airbnb and Stripe to use their apps; it’s that the network’s alumni honeycomb the Valley’s largest companies. Many of the hundred and twenty-one YC startups that have been acquired over the years have been absorbed by Facebook, Apple, and Google.

Yet Altman worries that the network’s very potency has become problematic. In February, he sent an e-mail to recent batches warning that some founders had grown cocky and entitled. And he told me, “It’s bad for the companies and bad for Silicon Valley if companies can stay alive just because they’re YC. It’s better for everyone if bad companies die quickly.”

One evening at Altman’s house, his younger brothers, Max and Jack, were teasing him that he should run for President in 2020, when he’d be thirty-five: just old enough. Max, twenty-eight, said, “Who better than you, Sam?” As Altman tried not very vehemently to change the subject, Jack, twenty-seven, said, “It’s not purely little-brother trolling. I do think tech needs a good candidate.”

“Let’s send the Jewish gay guy!” Altman said. “That’ll work!”

Jack eyed a board game called Samurai on the bookshelf and said, “Sam won every single game of Samurai when we were kids because he always declared himself the Samurai leader: ‘I have to win, and I’m in charge of everything.’ ”

Altman shot back, “You want to play speed chess right now?,” and Jack laughed.

Max was working at the Y Combinator company Zenefits; Jack co-founded a performance-management company, Lattice, which had just gone through YC. The two brothers moved in with Altman temporarily three years ago and never left. Altman recently hired a designer to upgrade his gray IKEA


Leave a Reply

Your email address will not be published. Required fields are marked *