Everyone is hoping that broadband to the home is the next killer app. The problem is that broadband isn't an app; it's an infrastructure that enables applications. We're still waiting for the killer app, and there are plenty of reasons to be skeptical about the contenders.

Look at interactive TV. Sure, I can play along with Who Wants to Be a Millionaire?, but I can't see that kind of interaction supporting an entire industry. Online shopping? We already have home-shopping TV channels, and dialing an 800 number isn't much more onerous than clicking on a hyperlink.

How about interactive video conferencing? Businesses can get this service via ISDN, and it isn't used that much, for reasons that are still mysterious. There's something about face-to-face contact that builds trust and cements deals.

What about movies on demand? The problem is that it has to compete with video stores and pay-per-view programming, which means it really can't charge more than $5 a pop. Online video stores like NetFlix send DVDs via mail. They offer convenience at a low price, and the U.S. Postal Service has plenty of bandwidth.

There's considerable interest in multicast, a technology that enables efficient broadcast over the Internet. It's a cool technology, but I'm skeptical. Radio frequency offers broadcast without fancy technology, so why replicate packets over pipes when satellite broadcasts have a bigger footprint and proven technology?

You don't even have to use anything as exotic as satellites. The recently announced iBlast network will use excess bandwidth in the HDTV range to broadcast music, software and videogames. It plans to start service next year in 102 markets using 143 local TV stations.

You can now buy a 10GB hard drive for $60 wholesale, and with MPEG-2 compression you can store 10 hours of VCR-quality video. In a year, capacity will double. That means anyone who can afford a VCR can have enough storage to hold the top-10 videos at Blockbuster. During off-hours, satellite broadcasts can fill up local caches and consumers can choose to watch this stored material "on demand." We're already seeing first-stage devices with some of these capabilities from ReplayTV, TiVo and WebTV boxes.

Not only is it hard to find killer apps, but also the demand for broadband has been less than overwhelming. In 1999, only 4 percent of U.S. households had broadband access. Renaissance Analysis reported that less than 2 percent of households sign up for DSL, and less than 4 percent sign up for a cable modem when they are available in a customer's area.

A UC Berkeley study found that people will pay little for broadband ISDN access to the Internet. Telecommuters were most willing to pay for the service: The "always on" and low latency features appeared to be more valuable than high bandwidth. Admittedly, we didn't offer movies on demand, but we didn't see much interest in paying for broadband, even among heavy Internet users.

It looks like broadband to the home is coming. What isn't clear is whether it will be profitable.