Microsoft is dead.

A few days ago, I suddenly realized that Microsoft is dead. It happened while I was explaining to a young entrepreneur the difference between Google and Yahoo. I said that Yahoo had always been hampered by fear of Microsoft. That was why they positioned themselves as a "media company" rather than a technology company. I looked at him and saw he didn't get it. It was like explaining to someone in their teens about how girls used to have big crushes on Barry Manilow in the mid-1980s. Barry who?

Microsoft? He didn't say anything, but I could tell he didn't believe anyone was still afraid of them.

For nearly twenty years after the late 1980s, Microsoft cast a long shadow over the software business. I remember when it was IBM that cast this shadow. Basically, I never paid much attention to it. I've never used Microsoft's software, so its effect on me was only indirect—like when I got attacked by zombie computers. Also, because I wasn't paying much attention, it wasn't till the shadow disappeared that I noticed it.

But now it's gone. I can feel it. No one is afraid of Microsoft anymore. They're still profitable, oh yeah, IBM is too. But they're not dangerous companies anymore.

When did Microsoft die? What killed it? I know it was still dangerous in 2001, because I wrote an article about how much less dangerous it had become. I think it died in 2005. When we started Y Combinator, I knew we didn't have to worry about Microsoft being a threat to the startups we funded. In fact, when we introduced the startups we funded to investors, we never even thought of inviting Microsoft people. We invited Yahoo and Google, and other Internet companies, but Microsoft never occurred to us. Nor, apparently, did we occur to them. They were another world.

What killed them? I think there were four causes, all dating from the middle of 2000.

The most obvious was Google. A village can only support one headman, and it's pretty clear now that Google is it. Right now, Google is the scariest company, in both senses of the word. Microsoft can only hobble along behind them.

When did Google take the lead? Conventional wisdom says August 2004, when they went public. But they weren't setting the pace then. I think they started leading in 2005. Several things contributed to their breakthrough, Gmail among them. Gmail proved that they could do more than search.

Gmail also showed people what online software could do if you really pushed AJAX—the second cause of Microsoft's death: everyone realized that the era of desktop applications was over. Storing programs on the Web is an irreversible trend—not just email, everything will be this way, even Photoshop. Even Microsoft saw this coming.

Ironically, Microsoft helped create AJAX without realizing it. The X in AJAX stands for XMLHttpRequest, the object that lets browsers communicate with servers in the background while displaying a page. (Previously, communicating with servers required requesting a new page.) XMLHttpRequest was invented by Microsoft in the late 1990s because Outlook needed it. But they didn't realize how useful it would be to others—to anyone who wanted to write web-based applications that felt like desktop apps.

Another important component of AJAX is JavaScript, a programming language that runs inside browsers. Microsoft ignored the problems with JavaScript as long as possible, seeing it as a threat. But the open-source world eventually won, writing JavaScript libraries that worked around IE's flaws, like trees growing through barbed wire.

The third cause of Microsoft's death was broadband Internet. Now everyone can have fast access to the Internet. The wider the pipe to the server, the less dependent you are on the desktop.

The final nail in Microsoft's coffin was Apple. Thanks to OS X, Apple came back in a way that's extremely rare in technology. Their victory has been so complete that I rarely see computers running Windows anymore. Everyone that Y Combinator funds uses Mac laptops. All the students in our startup school do too. People in tech either use Macs or Linux. Windows, like Apple in the 1990s, is something grandmothers use. So it's not just that desktop systems aren't important anymore; it's that anyone who cares about computers doesn't use Microsoft anymore.

Of course, in music, Microsoft is chasing Apple's tail, and soon TVs and phones will join the list.

I'm glad Microsoft is dead. Microsoft was like Nero or Commodus—only inherited power corrupts absolutely. Remember, Microsoft's monopoly position didn't start with Microsoft; it came from IBM. From the mid-1950s to 2005 (or since its inception), the software industry has lived under the shadow of monopoly. Part of the excitement around "Web 2.0" is that, consciously or unconsciously, people sense this monopoly is ending.

Of course, as a hacker, when you encounter a problem, you wonder how to solve it. Could Microsoft revive itself? In principle, yes. Consider two things: one, the amount of cash Microsoft has; two, Larry and Sergey tried to sell Google to a search engine company for a million dollars ten years ago, and everyone turned them down.

What people don't realize is that for a company as rich as Microsoft, the price of top hackers—hackers so good they're dangerous—is very low. Microsoft can't attract the smartest people anymore, but they can buy them, as many as they want. So if they want to play in the market again, here's what they should do:

One, acquire all the "Web 2.0" startups. Buy them all, probably for less than they planned to spend acquiring Facebook.

Two, put them all in a building in Silicon Valley, shield them with lead, and prevent any contact with Redmond.

It's safe to make this suggestion because they'll never do it. Microsoft's biggest constraint is that they don't know how bad they are. They still think they can write software behind closed doors. Maybe that worked in the desktop era, but that era ended several years ago.

I already know how people will react to this essay. Half the readers will point out that Microsoft still makes a lot of money, and say I shouldn't generalize from the small, insular world of "Web 2.0." The other half, the younger half, will complain that I'm talking about old news.

---

Footnote One: Software incompatibility isn't usually deliberate. You just have to stop trying hard enough to fix bugs. If you're a big company, the number of bugs is large. This situation can be compared to the writing of "literary theorists." Most of them don't deliberately try to be obscure; they just don't try hard enough to be clear. And predictably, it doesn't work well.

Footnote Two: Part of the reason is how Steve Jobs was booted out by John Sculley in a way that's uncommon in the tech world. If Apple's board hadn't made that mistake, they wouldn't have had the need to come back from the brink.