Even before computer sales slowed drastically last fall, the tech industry was buzzing about something called the post-PC era. But just what does that mean? We use terms like postmodern or post-cold war when we know where we have been, but have not yet figured out where we are or where we're going. That's not a bad description of the current state of personal technology, considering personal computers have become cheap, unremarkable elements of our home and work environments.
PC ubiquity has helped trigger a tech slowdown and stock market fuss, knocking such highfliers as Microsoft (MSFT
), Cisco (CSCO
), Dell Computer (DELL
), and Intel (INTC
) clean out of the BusinessWeek 50. For some, such as Cisco Systems Inc. (CSCO
), this period will probably be remembered as no more than a bumpy patch. For others, including Intel Corp. and Microsoft Corp., the players are being forced to seek growth in new non-PC products and services. And the changing market leaves a company such as Dell Computer Corp., which has hitched its fortune almost exclusively to the PC, to either seek a new business model or face an unglamorous future as a low-cost commodity manufacturer.
Make no mistake: The PC is here to stay. ``Hardly anything ever goes away,'' says Robert J.T. Morris, director of IBM's Almaden Research Center in San Jose, Calif. ``The mainframe and the minicomputer, too, were thought to be obsolete, but those platforms persist.'' The point, says Morris, is that while the PC, like its predecessors, will remain important, its central role in the technology industry is waning.
To figure out what a post-PC world might look like, let's start with what went wrong. As someone who test-drives technology for a living, I find the personal computer, for all its wonders, to be a bit of a disappointment. Computers have often been compared with radios, which have evolved from complex devices for enthusiasts into appliances everyone uses without thinking. But even in their third decade, personal computers seem more like shortwave radio transmitters, which can be used only with specialized training.
To some extent, the computer industry is a victim of its own success. Even in the hands of undertrained workers who use just a fraction of the available power, PCs are so cheap and flexible that they have blown away all challengers, including the much-heralded network computer. But today's PCs are overkill for most of the tasks they do in business, so faster chips and more storage don't mean much. Hardware whose power far outstrips the demands of the software has already led to much longer replacement cycles. And now the market is just about saturated, especially in the U.S. The obvious conclusion is that growth in the business market will be much slower than it has been.
The picture at home isn't much prettier. The household PC is likely to be limited more and more to the creative functions it best performs, leaving the more passive chores of browsing Web sites or playing music or online video to simpler, more specialized devices.
This creative market, broadly defined, is still huge. It covers everything from students writing papers to parents posting video of junior's first steps on a Web site for Grandma and the world to see. But most of these content creators already seem to have computers, and market penetration may be stuck at some 50% of households, where it has hovered for the past couple of years. Most recent growth in retail sales of computers has been in replacement or additional units for homes that already have one. And that demand may stagnate, because the future is likely to see homes with one full PC and an assortment of other specialized devices such as Web browsers or music players.
The main reason this hasn't happened is that the devices, or information appliances, have so far not been very good. And they remain isolated gadgets, incapable of working together. ``I don't think there has been too much progress,'' admits Donald A. Norman, whose 1999 book The Invisible Computer presaged the info-appliance revolution.
Given the rough topography of the world beyond PCs, what might be the rules of survival for the endangered species of CEOs in this industry? Here's my checklist:
Best-of-breed products. You don't need a PC if all you want to do is surf the Web and read and write e-mail. But if that's all a product is going to do, it had better do it really well, with a Web browser at least as good as Microsoft's Internet Explorer 5.5 and a mail program as good as Outlook Express. No appliance, including those developed by Microsoft and America Online, has come close. Furthermore, because Web-browsing appliances need mostly the same innards as PCs, they aren't going to be a lot cheaper; their appeal must come from quality and usability, not price. These appliances will not succeed as long as manufacturers use terrible keyboards or fuzzy displays to shave a couple dollars off the bill for materials.
Get them talking. If I schedule an appointment on the Internet appliance in my kitchen, I want it to show up automatically on the calendar on the PC in my home office. Today, the only way to accomplish this is to use slow, cranky Web-based calendars. ``Appliances have gotten a lot more intelligent, but they are all independent,'' says Norman. ``The networked home is still a lot of talk, and there seems to be a new industry consortium every week.''
The growing assortment of devices--PCs, countertop appliances, handhelds, and things we can't even imagine yet--need a network, preferably a wireless one. Ideally, each device should connect to a local network when it is available and to a wide-area network, such as the wireless-phone system, when needed. The slow development of wireless networks means this capability is several years off, but designers should keep the goal firmly in mind.
Devices on the network need a way to discover one another and learn each other's capabilities. This, for example, would allow you to send e-mail from your handheld to the nearest printer without setup or configuration headaches. Technologies such as Microsoft's Universal Plug and Play and Sun Microsystems Inc.'s (SUNW
) Jini could be a big help, but given the rancor between those companies, we are almost sure to get two rival, incompatible approaches.
While we wait for standards to develop, it would be good if consumers could just set up and run a network without having to understand things like the Dynamic Host Configuration Protocol or Network Address Translation. I doubt that self-managing networks will ever solve the problem. So there may be a new business for ISPs in running home networks for their customers. This could be a great opportunity for broadband service providers on the BW 50 list, including AOL Time Warner (AOL
), Verizon Communications (VZ
), and SBC Communications (SBC
), but none has yet risen to the challenge.
Ears and eyes. For entering large amounts of text, nothing can beat the 130-year-old keyboard. But the keyboard requires extensive training to use efficiently, demands that chunks of plastic clutter up desks, and hardly works at all for non-alphabetic languages such as Chinese.
Alternatives for handhelds are all unsatisfactory. Shorthand scripts like Palm Inc.'s (PALM
) Graffiti work fine for short messages, but accuracy and speed are elusive. Onscreen keyboards are awful, and telephone keypads are worse.
This should spell new business opportunities in the field of speech--the most natural form of human communication. But it will be years, at best, before a computer achieves the language processing skills of a five-year-old child. Speech recognition works well if you can make sure that people say only things that the computer can understand, usually by restricting topics and requiring a precise syntax. The trouble with this, says speech expert Victor Zue, associate director of Massachusetts Institute of Technology's Laboratory for Computer Science, is that ``directed dialogue means that people are still serving machines.'' More humanlike comprehension, he warns, is not immediately forthcoming.
Zue believes that computers of the future may have to combine vision with speech to gain that sort of comprehension. Mounting a camera in a handheld device is easy today, but getting a machine to comprehend what it sees is a huge challenge.
Keep Big Brother at bay. A computer that sees all and hears all is a computer that can tell all--and one you may not want in your home or office. Ironically, the price of more humane computing could be a grave threat to what's left of our privacy.
Mechanisms have to be developed to make sure that information we want kept private doesn't leak out. Technologies exist for keeping what is heard or seen in your house inside your walls. But the sociological hurdles may prove more challenging than the technical ones. The industry's cavalier attitude, typified by the widespread, quiet collection of personal data on the Web, have created a credibility gap. ``The difficulty is coming up with a sociological proof that we are protecting privacy,'' says IBM's Morris. ``It's a trust problem in the end. We have to solve the problem, or we will fail to reap the benefits of the technology.''
Ultimately, a successful transition to a post-PC era may depend as much on attitude as technology. When computers were regarded as experimental devices for the technically inclined, or as business machines maintained by a technical priesthood, users were tolerated a great deal. But one requirement that people have about gadgets, whether they are information appliances or washing machines, is that they do their intended jobs simply and reliably. The tech industry has a long way to go before it can live up to that expectation.
The Four Commandments of the Post-PC World
The computer is sure to be a fixture of your cubicle or classroom for years to come. At home, however, specialized devices will rule.
-- A device that does only one thing has to do it really, really well.
-- Appliances must link easily to share information.
-- Entering information has to get a lot easier than it is now.
-- Privacy will be paramount. A house full of devices that hear all and see all could be more threatening than helpful.