Gmail user Matthew Amster-Burton had muted expectations for online applications. He'd been frustrated before by other Web mail services, but Gmail quickly hooked him -- and from there, his Web reliance snowballed.
"After realizing how much I liked Gmail, I became open to the idea of using Web applications instead of desktop apps or post-it notes," Amster-Burton says. A food writer based in Seattle, Washington, he's adopted what he jokingly calls a "Web lifestyle."
While ubiquitous Web connectivity and powerful applications can ease users' lives in previously unimagined ways, breakthroughs require lots of experimenting -- and failure. The new technologies also carry new risks. Security is an endless concern, but so are plain old outages. The more a user entrusts essential data to Web applications, the more at-risk the user is when those applications fail.
As if to underline that point, a group of Web 2.0 poster children suffered a quick succession of crashes during a one week-long stretch in mid-December. Blog hosting service TypePad went down for an extended stretch after a failed storage upgrade, while del.icio.us users endured days without their online bookmarks after a data-center power loss wreaked havoc with the site. Hosted sales software provider Salesforce.com Inc., a well-known evangelist for "on-demand" enterprise software systems, crashed for a day because of database glitches, enraging customers who were rushing to close sales deals before the end of the year. Problems persisted into the new year.
Other challenges loom as technology enables users to tap into exponentially expending data stores and networks. IBM has hundreds of scientists studying analytics. One speculative project, WebFountain, uses complex software algorithms and a grid supercomputer with 40 racks of blade servers and network-attached storage hardware to tackle clients' trickiest data-mining problems. Keeping human expertise on pace with advancing technology is an arms' race: "When we started this in 1999, I'd never heard of a blog, and now they dominate as one of the richest information sources," WebFountain Chief Scientist Dan Gruhl says.
Gruhl sees next-generation consumer technology leaping ahead of enterprise tools in innovation. "A lot of Web 1.0 was driven by people in professions that had access to information through sources like Nexis or Westlaw. We got spoiled with the ability to find information at work easily," he says. "Then it slipped. Enterprise tools drifted while consumer tools raced on. There are better tools for hobbyists than enterprises. [At IBM], we want to figure out, how can we create enterprise equivalents to things like blogging? How can we do search better?"
IBM's research minds and big iron are working on problems like developing systems that can shift through the Web and identify the latest street lingo for drugs -- a data set that would help emergency-room physicians talking with patients in crisis. Another WebFountain demo project searches blogs and college Web sites for music discussions, measuring online chatter to forecast next week's Billboard Top 40 hits.
Meanwhile, as new businesses and research projects get off the ground, the Web 2.0 hype grows. "It's one of those things that took off so fast, it loses its initial meaning," says Yahoo's Horowitz. "It's like the dot-com of this generation. Back then, if you stuck on the label, if you were Pets Dot Com, your valuation went up. Now, if you're Photos Web 2.0, your valuation goes up. Yahoo has some of best talent and minds in the world to help us find the signal in the noise, but it's getting noisy."
Still, unexpected sparks can emerge from a field of static. Amster-Burton, the blogger and food writer, says the Web application he most covets right now would be a synchronization tool to mesh files from his laptop and desktop machines. He's tried a handful of available options, none of which impress him. But experience has taught him to be optimistic.
"The form is so limited and the competition is so fierce, if you're not doing something special online, no one's going to show up to play with it," he says. "I think someone is going to come up with an awesome solution to this problem. Something that is way more clever than I can imagine."