Ignore technology at your Peril. (Editor's Note).

Facing war abroad and many pressures at home, it's difficult for CEOs to keep their eyes on technology. But now is a critical time to be doing just that. Although most companies have slashed spending on technology and capital goods, several important tech trends that began in the 90's are still playing out. I can't tell you which of them will emerge as the most important. (If I could predict that, I'd be a billionaire, not an ink-stained wretch.) But I can predict that some of you are going to get surprised.

AOL Time Warner, for example, was terribly late in recognizing the importance of broadband access from the home. It assumed it could keep millions of subscribers happily addicted to slow dial-up connections. But once an American household samples high-speed connectivity, the obvious question becomes, "Why do we need to pay $23.90 a month for AOL?"

The regional phone companies also may have misplayed broadband. The Bells have been happy with their monopolies and therefore not very enthusiastic about DSL. But cable companies, who worry about competition from satellites, have been aggressive in rolling out broadband via cable modems. If they can figure out how to run voice over their lines, they might win the game and own the relationship with customers. "Why do we need a second phone line?" people may start to ask. "Why do we need the Bells at all?"

There could be surprises in fiber optics as well. Right now, the conventional wisdom is that much of the cable laid into the ground in the 1990s was a waste. The generally accepted figure is that only about 5 percent of the fiber in long-haul, backbone networks is "lit," or being used. But someone in the business of laying fiber in metropolitan areas suggested recently that once every American city gets hooked up and millions more homes obtain broadband, we could actually end up having a shortage of fiber in the long-haul networks. The cycle could turn in a heartbeat.

Elsewhere, Sun Microsystems' Scott McNealy may have been slow to understand the significance of Linux and other open-source software. As a result, he's got a problem. Google has come out of nowhere, it seems, to dominate the Internet search business. WiFi could be big. So, too, could sensors and cameras that are linked by networks. There are likely to be some surprises as our primitive cell-phone networks make their way to third-generation technology. Likewise, there are huge stakes in how well companies manage all these networks.

CEOs can't rely on chief information officers as they used to, as a new Gartner survey shows (see page 15). Too many CIOs haven't understood the need to make a return on their investments, and too many haven't understood the business. They have sought to bend the business to fit the technology, not the other way around.

Nobody thinks CEOs should be micromanaging tech projects. But they should be forcing their organizations to understand the long-term trends and position themselves accordingly. Not to do so, even in difficult times, invites almost-certain disaster.