On May 24 of this year, when Zuckerberg announced he was opening Facebook to independent developers, it was clear to Jonathan Sposato that the company had done something revolutionary. He knew how to develop and successfully distribute software: In 2005, Sposato, a former group manager at Microsoft, started a company that made it easy to create software widgets, and he sold it to Google later that year. In mid-2006, he and two fellow Microsoft alumni created Picnik, a slick online photo-editing site.

But even Sposato was surprised at the response from Facebook users when Picnik was included as one of the 85 initial applications in Facebook Platform, the new development tool. Within three days, more than 100,000 users downloaded his program — about 10 times more than he'd anticipated. Because News Feed instantly and automatically notified friends whenever someone downloaded Picnik, word of the application spread exponentially. Sposato called colleagues in a desperate — and ultimately successful — hunt for extra server capacity and bandwidth to avoid outages. Currently almost 250,000 Facebook users have installed Picnik on their pages, making it the network's top photo-editing tool.

Sposato's experience shows the power of Facebook Platform as a new model for disseminating software. The plummeting costs of bandwidth, processing power, and storage had driven down the price of application development. But unless you could figure out a way onto the Google homepage, it was still tricky to tell the world what you'd created. Facebook now gave even the most modest developer the opportunity to win instant and mammoth distribution through its word-of-mouth engine. Users no longer need to search for applications that they may not even know they want; instead, the applications find them.

Since then, more than 3,200 new applications have sprung up on the site, a number that is growing by about 180 a week. Those offerings have made Facebook a fully functioning social hub, where users can keep track of one another's favorite music and videos, share and compare movie reviews, and hit one another up for contributions to pet causes. Facebook promises to become an online identity for recruiters, bosses, and colleagues looking to hire and promote; a souped-up business card for job hunters; and a dossier of people's likes and dislikes that vendors can use to provide targeted products and services. Salesforce's Benioff even imagines Facebook pages serving as universal health records.

And by turning itself into a platform for new applications, Facebook has launched a whole new branch of the software development industry, just like Bill Gates did with MS-DOS in the 1980s. By allowing developers to charge for their wares or collect the advertising revenue they generate, Zuckerberg set up a system for every programmer to get paid for their efforts. Now venture capitalists like Bay Partners are scrambling to fund almost anyone who has an idea for a Facebook application.

Skeptics may argue that we've seen this movie before — in 1999, say, when anyone with a vague concept for a Web site could get VC backing. And, they point out, nobody actually does pay for Facebook applications. Still, the startup costs for developers are extremely low, and the potential is high. For the Internet, email was the killer app — a program so useful that it transformed the platform into a massive communications tool. There's no killer app for Facebook yet. But if someone can develop one, they will be sitting on a gold mine.

For all the excitement, one sobering fact remains: Facebook has yet to prove itself as a business. The site's nearly 40 million active users generate more than a billion pageviews a day, but ad clickthrough rates are low. An estimated half of its $150 million in revenue comes from an advertising deal with Microsoft. Independent developers are drawn to Facebook because Zuckerberg lets them keep any advertising revenue their applications generate; if Facebook can't prove itself as an advertising venue, the deluge of new applications will slow to a trickle.

Nevertheless, Zuckerberg's notion of the social graph has proven so powerful that almost every other company in the Valley is trying to replicate it. Jeff Weiner, one of Yahoo's top executives, refers to users of Yahoo Mail as a Facebook-esque "dormant social network" that his company "needs to activate." And MySpace is expected to respond to Facebook's challenge; CEO Chris DeWolfe has made vague statements about the site's "evolution."

Whatever ultimately becomes of Facebook, Zuckerberg has already had an impact. A year ago, the Valley wondered if this cocky youngster had turned down his only shot at $1 billion. Now it's wondering if he has defined the future of the Internet.