If you believe the volumes of articles being written about wireless communications, the local Clam Shack will soon be beaming out menus to the cellphones of every Tom, Dick, and Harriet who passes by. Never mind that the person receiving the menu might find the call intrusive, or even frightening—if that restaurant knows where he is, then who else is tracking his movements? Never mind that cellphones’ small screens and tiny keys make them remarkably ill-suited for reading menus. Never mind that the recipient of the call foots the bill, paying for the privilege of being spammed.

In the conversation that follows, both Reed and Keeley say that wireless has a tremendously bright future—but a very different one than most people seem to expect.

LARRY KEELEY: When we look at wireless, we see the same exuberance we see whenever there is a new technology. People make wild predictions. What people typically miss are the deeper patterns that govern how people will actually use their new possibilities.

Most of the predictions I have seen are that the wireless Web will be the most successful medium in the history of technology. That is probably true, although it won’t develop at the rate predicted, and certainly not in the form expected. People will make wireless their own in surprising ways.

Every time I talk to users about their WAP phones ["wireless application protocol" phones were advertised as an easy way to reach the Internet from a cellular phone], they virtually describe them as dogs with fleas and mange. There are too many buttons, they say, too many things that they don’t understand. It takes them too long to get to any site, and when they get there the site isn’t worth the effort.

Companies are chasing applications that people don’t care about—especially those that foster shopping—and the devices being produced are vastly too slow and confusing.

DAVID REED: The standard pattern that emerges with new technology is that people see it merely as a form of an old technology. In this case, we’re dealing with what I call the "wireless-wire syndrome"—in other words, that wireless is everything you can do with wires but without the wires. We imagine that we’re going to take all the stuff we do at desks and do it on the beach.

That’s the wrong way to think about this. The car wasn’t just a "horseless carriage," and mobile devices won’t just be the same old devices, without wires. You have to imagine what it would be like to have your business on the customer’s body, always within arm’s reach, and then find ways to take advantage of that new immediacy.

Another issue that’s often missed is that current cellular networks are built on an architecture that creates fundamental problems.

Cellular networks are based on "circuit switching." Each phone call has a certain amount of capacity dedicated to it for the duration of the call. This is not compatible with the bursty kinds of traffic that occur with Internet applications, which make almost no use of the network for a long time and then all of a sudden require a blip of capacity. A computer needs to be able to say, I need a million bits of transmission capacity right now, and then I’m going to stop needing any capacity at all.

To create that capability would require re-engineering the whole cellular system from the bottom up, and that’s just not in the works. It isn’t planned even as part of what’s called the 3G, or third-generation, network plan. Yet, a lot of people imagine that the cellular network is compatible with the Internet.

The idea that we’re suddenly going to get Internet-style interaction and innovation on phones just isn’t supported by the plans for the networks being built underneath.

In addition, while the cellular network is increasing capacity, it’s not increasing fast enough to support the amazing demand we see. The cellular network is like a highway system operating in a clogged-up mode all the time.

Cellular system operators are clueless, by the way. They see a system full of customers and think everyone is more or less happy.

KEELEY: We have anthropologists watch cellular customers, and it’s astonishing. The anger people have about their phone service does not yet quite equal the anger they have about their cable TV companies, but it’s close.

REED: A technology called 802.11, though still in its infancy, is a good example of wireless’s ultimate promise. It allows laptops and some hand-held computers to communicate wire-free within about 100 yards in open space and 100 feet inside a building. The system is very fast, uses little bandwidth, and works well with applications like e-mail, messaging, or calling up Web pages. It feels like a computer on a great network. User reaction has been incredibly positive. It’s cheap and can be easily installed. It’s great for the "always-on" connections and bursty traffic that characterize Internet usage.

KEELEY: We have been using 802.11 at Doblin Group for more than a year. It is by far the most popular technological adaptation we’ve added in 10 years.

I’ve even installed it in my home and in a vacation home in Nantucket, Mass. Now I get teenagers sitting in the backyard using this advanced equipment just to send Instant Messages at 11 megabits a second.

It’s unbelievable the freedom this simple lack of wires gives people. In our office, we have people who can spontaneously camp anywhere and work effectively. We even set up a peer-to-peer network in an airplane en route to a client in San Francisco. We were sitting in three seats that were separated by 30 rows, but we needed to finish a presentation. Being able to work on it at high speed on an airplane was magical. We’re told this produces less radio interference than a kid playing a Game Boy.

REED: That’s just the beginning. Right now, the technology is operating below the radar. The real killer apps that go along with it haven’t yet been discovered or at least made public. Still, I think 802.11 will come out from nowhere and challenge the cellular oligopoly. Maybe it will function in the same way that the Internet did as it challenged the incumbent phone companies.

KEELEY: When people have control over how they use a product—as they will with wireless devices—growth markets spring up.

In Japan, people have adopted short-messaging services in astonishing ways. Kids split up and go to clubs across town and continuously contact each other to see where the action is. Then the group gets up en masse and moves to one agreed-upon location. It’s a simple example of how differently people behave without wires.

There are all kinds of possibilities. Imagine a bus stop with a display that communicates with buses and indicates that the next one will arrive in two minutes, 41 seconds. If the message was accurate, it would be a tremendous help to commuters.

Unfortunately, most companies don’t have a real-life environment to test ideas like that, to see how customers would use their products. So, most companies don’t know what customers would really find useful.

When we’ve studied how people use the Global Positioning System in cars, we’ve seen that auto maker General Motors Corp.’s OnStar system (www.gm.com), in particular, is way too much technology chasing way too little useful application. Many people are happy to use it for demo purposes to talk to OnStar operators via the car’s cellphone or satellite link. People show friends how the system can turn on their headlights or unlock their doors. But, in practice, OnStar isn’t used so much.

My guess is that GM is investing $19,000 per car on the goofy functions the system provides, but I can’t prove it yet.

REED: It’s fundamental for users to have some control over the way a product develops, yet all the wireless technologies I’m aware of have very fixed functions. What’s missing is flexibility—the opportunity for customers to experiment. There’s no allowance in the business for the unexpected.

For instance, the Iridium satellite network [designed to let people make phone calls from anywhere in the world] had all kinds of desirable features but was designed to do exactly one thing. When the market wanted something else, cellphone manufacturer Motorola Inc. (www.motorola.com) could do nothing but throw the whole network of satellites away.

Cellular companies are taking a huge risk by thinking about technology in terms of generations. They roll out a new generation everywhere almost simultaneously. That requires huge capital spending, so they can’t afford to make a mistake.

While someone trying to develop a market obviously needs to have a business plan that shows a return on investment, companies need to do two additional things. They need to enable the user to do something interesting with the product other than what was originally planned—like your example of how phones are used in Japan. Companies also need to pay attention to what users do and build on what they learn.

Companies are often so confident in their planning that they don’t look for the surprising things people do with their products.

KEELEY: That’s 100% right. When you look back to see the things that have been telephonic successes, they’ve been discovered and driven by users.

The way some companies plan is shockingly inept. People feel that they have to predict precisely the kinds of volume and revenue they’re going to get from something that has never existed before. I take one look at the number of nested assumptions put in predictive models that deal with one of the most dynamic markets and technologies in the history of the world, and I need to tell myself to lie down until the feeling goes away. What are they thinking? There is a grotesque need to try to prove that something will work before you try it. In fact, people are basically just guessing.

REED: I agree that successes with wireless will be driven by users’ innovations. I also think users will pay for the experiments themselves.

If you look at what really made the Internet powerful, it was that corporate users could wire themselves. Firms put in local area networks and long-distance links and then started connecting computers together. They discovered what was useful without a whole lot of planning. Consciously or not, they created the flexibility to do almost anything they wanted.

Similarly, with wireless, a lot of interesting applications will likely emerge from work groups, as clever people figure out how to adapt the technology in powerful ways. The amount of money that they spend on developing and installing equipment may collectively be even larger than what’s needed for current plans to improve the cellular network, but it doesn’t carry the same level of risk. It’s not one big capital expenditure that has to work.

KEELEY: The good news is that there is so much potential ahead of us. Wireless is roughly where the Model T Ford was, when people had to jump out of the car at every major intersection of rutted roads and fiddle with the carburetor to make the car run a few more miles.

Some breakthroughs won’t come for another five years, maybe 10. Still, I expect we’ll see a revolutionary shift in both wireless devices and applications. Phones will also become more polite, less intrusive. They’ll probably be able to make judgments about urgency or protect privacy, too.

Here is a good example: diabetes. If a diabetic’s system goes into imbalance, he shows up in a hospital with insulin shock. It can cost $50,000 to bring him back under control and send him home again. These episodes are economically disastrous for insurers and dangerous for the patient. It turns out that there is a simple solution. A wireless camera that sends occasional images from the kitchen sink can show whether a person is washing his dishes regularly. If the camera suggests that eating patterns are out of whack, that may indicate the patient is going into shock. Medical staff can call the patient and head off the problem.

REED: Doctors don’t really know what people do for their health. All they get are verbal reports, which are incredibly unreliable. Wireless devices could change all that.

The real risk with wireless, in addition to building inflexible systems, is aiming too low. What killed WAP is that it was designed as though cellphones would forever be as stupid as they are today. We know that silicon keeps getting better, so cellphones will develop into very powerful computers. Even if you can’t build a capability into a cellphone today, one year from now you’ll be able to do twice as much for the same price.

Reed can be reached at dpreed@reed.com. Keeley can be reached through www.doblin.com.