Browsing is an obvious idea, but it is not necessarily the right one. Too much of the Net's future is staked on this unchallenged notion. And the sooner we stop relying on this concept, the better. Just think: How much browsing do you do in real life, or, as John Perry Barlow would say, in "meatspace"? Most working adults don't have time to spare. Browsing is better suited to the confines of a doctor's waiting room, an airplane seat, or a rainy Sunday afternoon. Rarely does browsing suggest the serious, productive use of one's time. Rather, it suggests another era, when work, home life, and vacations were less entwined than they are today.

So what happened? Why did we suddenly elevate this faulty, serendipitous, and almost haphazard process to its current prominence - even predominance - on the Internet? The verb browse is derived from the behavior of hungry animals who, in winter when pasture is barren, forage for tender shoots and the buds of trees and bushes. This implies that there isn't a lot to choose from and that what is good needs to be actively sought out.

But browsing takes time - the one thing most of us don't have. For example, I do far less window-shopping than I did when I was young (and yes, I miss it). Undeniably, browsing can be fun and useful but, as with tourism, only so much and so often. Funny how we use the words cruising and surfing to describe our behavior on the Web. How often do we invoke the words learning or engaging when we browse?

The Web is a digital landmark, as important as the Net itself. Its inventors, Tim Berners-Lee and his colleagues, will probably never fully realize how important their contributions were, and will continue to be, because the Web can be viewed in so many different ways. For me, it's less about multimedia or hyperlinks and more about turning the Net inside out. Instead of sending email to an individual - or to a list of individuals - I can now post a message and invite people to look it over. Sure, we've always had bulletin boards, telnet, and ftp on the Net, but the Web created a new and more accessible subworld, one more like the window-shopping experience than the original message-passing rubric. And in a way, that's a shame.

Think of the change this way: the Internet is now like a city - people go places, visit communities. In fact, we even call our own pages "home." But when we arrive at a place and try to make things happen, we often end up frustrated.

Direct manipulation doesn't work
After years of work to make computers more accessible, researchers began questioning whether people really wanted ease of use in the first place. Using a computer is often just a means to an end, so wouldn't that end be better served if we could delegate tedious computer time to someone - or something - else?

As far back as 25 years ago, Alan Kay and others suggested that "delegation" was a much better metaphor than "direct manipulation" when focusing on people's productive use of computers. Later, at Apple - the font of many human-interface advances - delegation became the challenge and "interface agents" became the solution, at first in name only. Instead of constructing a computer that is easy to manipulate directly, the argument goes, why not fashion it after a well-trained English butler who knows you so well that he will do almost everything on your behalf. He will do what you ask and, in some cases, you won't even have to ask.

Most important, the idea of an interface agent entails its ability to "understand what I mean." Please just do it and don't bother me. When it comes to interaction with such an agent, less is more. Let it do the searching, surfing, and cruising. Let it browse for you and bring you the fruits of its labors.

The net population of the Net
In America today, the demographics of personal-computer use is oddly bimodal. Most kids have some access. But, surprisingly, the next largest and fastest growing group (as a percentage of the age group per capita) is those age 55 and older (more than 30 percent of whom own a personal computer). Between these two groups, we have what I call the "digital homeless": those who arrived on the planet a little too early, or not early enough, to have the time to explore the possibilities of being digital. Many people in this group feel that the online world has nothing to do with them and (according to a report in Business Week) value their hair dryer more than a personal computer.

Among those who are digital, particularly the young and the old, we find a majority of people with free or flexible time, people who can literally afford to spend their time browsing. I cannot. I need to delegate that process. And I'm not alone.

By 2000, we can expect a billion users on the Net. As recently as a year ago, this number seemed outrageously high. Today it is considered a conservative estimate. What this statistic fails to include is the huge number of machines and software programs that will use the Net on our (and their own) behalf. At the turn of the millennium, we're likely to find those billion human users joined by a much larger number of software agents, Web crawlers, and other computer programs that will do the browsing for us. The Net will be roamed mostly by programs, not people. When people do use the Net, it will be for more suitable purposes: communicating, learning, experiencing.

The idea that machines, not people, will dominate Net usage turns the model upside down, not just inside out. Suddenly "pages," if that's even an appropriate term, will need more and more computer-readable hooks so that programs can see what you or I view from the corner of our eye. When we browse, our eyes gravitate toward images - in the future, these images will need simple digital captions. This will certainly take steam out of the Net-based advertising we know today. Simply put, our eyeballs may not be there to see it.