Sometime in the mid-1990s, a group called the BandwidthConservation Society made its appearance. It tried to reduceInternet traffic in a few ways-getting Web site designers anddevelopers to create smaller pages, convincing people to dolarge-file transfers late at night, and so on. I argued in anInternet World editorial that we should encourage the opposite-thatwe should use as much bandwidth as possible to encourage thedevelopment of faster backbones and better connections. Create thenecessity to spur the invention.

That's what happened. T-1 backbones gave way to T-3 and DS3,then OC3, OC12, and other high-bandwidth connections. Cable and DSLaccess for the masses became commonplace. Streaming audio and videobecame possible for people at home. The lack of conservation gaveus Internet access that is, for all intents and purposes, too cheapto meter. (Perhaps there's a lesson here about energy policy.Instead of conserving energy, by using more we may force thedevelopment of new and better resources.

And now the next step-made possible by a lack of bandwidthconservation-is Web services. Instead of exchanging data, we willsoon be able to share applications. As standards such as SOAP arehammered out and services like Microsoft's .Net and Sun's ONE gainpopularity, companies will offer applications-or even pieces ofapplications-as objects, much the way ASPs offer services, but on amore granular scale.

Make no mistake: The advent of Web services is a major milestonein the history of the Internet and computing in general. Over thenext few years, it will cause drastic, changes in the softwarebusiness. It will make large segments of the insurance technologymarket into commodities-and if you think that's a nasty word forcarriers, wait till you see how vicious the software market becomeswhen insurers can buy line-of-business applications on contract orsubscription...and switch vendors almost at the drop of a hat.

We'll see larger tech companies scrambling to understand the newmarket, and upstarts snatching business away one application at atime. You'll read stories about reorganizations, spinoffs, newbusiness units, and partnerships. New companies will emerge to fillsmall niches in the market, and they'll compete with-and in manycases beat-the big names. Perhaps carriers will be able to rentapplications not annually or monthly, but on a daily, hourly, oreven processor-time basis.

That means potentially good times ahead for insurance companies'IT departments. Although you'll be forced to make significant,difficult changes to your software infrastructure, the long-termbenefits and savings will be outstanding. For software providers,the future is cloudy. Nimble companies with high-quality,standards-based products will prosper, while other vendors willstruggle to take advantage of the dramatic turn of events. By thistime next year, the landscape will have changed dramatically. Waitand see.