Fast Enough for You?

Telecosm: How Infinite Bandwidth Will Revolutionize Our World, by George Gilder (Free Press, 351 pp., $26)

The Internet is a topic that tends to bring out the fruitcake in people, but George Gilder is a formidable man; his books are cultural events. In Telecosm he focuses his compelling critical intelligence like a megawatt death-ray on the present and future of data networks. The result: heat and light, fire, sizzle, and flash, fireworks, explosions, and a book so good it actually rollicks. Over the last decade hundreds of books have investigated this fast-changing landscape-which underlies the future of the Internet, the web, and computers in general, of phones and TV and communication and culture; Gilder's is one of two or three that are indispensable.

Art and literary critics used to be (to speak Gilderesquely) smart interface cables that helped artists and laymen connect, so ideas could flow. They did not merely explain, popularize, or report; they advanced theories, championed causes, and moved the field. Gilder is in this sense a technology critic-one of the best. Unlike so many technology writers, however, he is not under the impression that technology is the whole world, or that the Internet will solve all of our problems. "People will continue to suffer and die, continue to sin and pay, continue to waste time and wealth," he acknowledges-the only gloomy sentence in the book, and a sign that the author maintains his perspective and dominates his material instead of allowing it (as so many technology writers do) to dominate him.

"Today," Gilder writes, "the ascendant technology is optics and the canonical abundance is bandwidth." As Gilder uses the word, "bandwidth" means data-carrying capacity, in bits per second. Data used to move fast within a computer, but between computers they shambled sloppily through crowded old-fashioned metal wires. Nowadays, glass is replacing metal; data rocket through glass fibers at the speed of light, moving faster between computers than inside them. Furthermore, new techniques make glass fibers more effective. They allow many messages to occupy one fiber simultaneously, at different frequencies-as if blue, indigo, violet, and many other color trains were all streaking down the same narrow subway tube at once. There are powerful new wireless technologies too, for the many places glass fiber can't reach. Gilder believes that the companies that are designing and building these new-style communication channels will dominate the technology future.

Telecosm's message is this: A quantitative change in data communications is leading to a qualitative change. "As more and more data sluiced across the globe by the week, the solution became more and more obvious," he writes: "an all-optical regime." There is no such thing as "infinite bandwidth," but Gilder believes that bandwidth will be so copious and cheap, it will seem infinite.

The best thing about the book is the way Gilder tells the story. It reads like a dispatch from the front: terse, strong, fast. Gilder is consistently in command of the complex facts, varied personalities, and his own prose. "Hicks ignored ray theory, easy enough to do since he hardly knew it." "Ironically, this policy of economizing on spectrum eventually led to using it all up." Sentences like these keep the narrative moving through the toughest terrain, and Gilder feeds himself into the story just enough to be interesting. His "misleading mnemonic" to help readers remember the seven layers of the official OSI network protocol is the most attractive moment in the book: a complex technical topic, Gilder admits, but it's important and "you should learn it." This seems like a small point but is in fact enormous. The author doesn't preach pompously from the technology heights; his goal is to help you up, to where you can see it all for yourself.

Even his mistakes are interesting. What does the emerging super-fast, super-capacious global network mean? That the network is now primary and the computer secondary? That your desktop computer will become a mere stepladder for reaching up into the global kitchen cupboard of the World Wide Web where all the interesting stuff is stored? That's what Gilder thinks-that "almost overnight, the CPU and its software have become peripheral; the network central." But he is wrong, as his own argument demonstrates.

In the emerging new Internet, signals move at the speed of light. With data going as fast as they possibly can, he writes, "the nemesis will be delay." In the technology world Gilder hopes for, each computer has the whole world at its fingertips; modern computer processors already spend most of their time waiting for data to show up. Say you're in New York and you need data from London: The data can't move any faster than light-speed-heartbreakingly slow from the processor's point of view. If your goal is to grab data from the web as casually as you do from local memory, the only solution (as Gilder explains) is to scatter copies of popular webpages all over the globe. Some of Gilder's favorite new companies are doing exactly that. If a copy of that London webpage has been stashed in Passaic, it can reach you that much faster.

This makes perfect sense, but Gilder forgets to notice that if we are talking about the real future and not merely the next few years, his argument only confirms the kingship of desktop computers. Gilder forecasts lots of copies of popular webpages, all over. And he notes that the cost of data storage has been falling even faster than the cost of computer power. But for once his imagination fails him. He reports on one scheme that "entails making as many as ten thousand copies" of popular webpages. Ten thousand is many? Let's be serious: If we are talking many copies, all over the world, close to their users-closer-more-closer!-more!-stop: You've reached the desktop. Ultimately the websites you care about most will be stored in your own computer. The network will be the back-office operation that keeps everything up to date.

In larger terms, Gilder has fallen for the Dorothy Mistake. The web is fabulous and inspiring; it is the greatest invention since the shopping mall-and yet there is no place like home. Ultimately people care most about their own information: their own photos, scrapbooks, e-mails, and bills, notes and files, bank balances and warranty cards, health records, diary pages, address books, voice-mail messages and love letters. Managing each user's personal information will become digital technology's most important job. And in a world where powerful computers with huge storage capacity are so cheap they will be given away with Cracker Jacks and children will be disappointed to get them ("Not another 5,000 tetrahertz computer! Wanna trade for your fire truck?"), your personal data will mainly be stored on your own personal computer. No other outcome makes sense. Personal computers are too cheap and powerful not to dominate the cyberworld of the future. The emerging "telecosm" will change them but not dethrone them.

Ultimately the web itself is an obsolete image, suggesting sticky strands full of trapped bugs or an antique rail system with computers like stations and data-trains chugging back and forth. The real future of communications is oceanic. People don't care about computers, they care about information and will deal with it directly, in the form of flashing throngs of cyberbodies moving gracefully through a cyber-continuum.

But now I am preaching my sermon, and Gilder is right to preach no man's but his own. George Gilder loves a law-he ends the book with twenty of his favorites, tendered like the dessert tray of a master pastry chef-so here is a new one in his honor: Any book about technology by George Gilder is guaranteed to be usually right, sometimes wrong, always well informed, and never (even for one paragraph) boring. My respect for the book transcends technical arguments. Telecosm is one of the best technology books I have ever read.