Bill's Bookshelf by Bill Rosenblatt

Michael Dertouzos tells us what will be

MIT pundit aims to understand the division between technology and the humanities

May  1998
[Next story]
[Table of Contents]
Sun's Site

In his book What Will Be, MIT Computer Science Lab Director Michael Dertouzos offers a wide-ranging vision of how the technology goodies coming out of labs like his will shape 21st-century society. Tempered with pragmatism and humanism, this book is more balanced than other "vision" books written by his peers. (2,200 words)

Mail this
article to
a friend
In the concluding chapter of Michael Dertouzos's What Will Be: How the New World of Information Will Change our Lives, the author describes himself thusly: "I have always been interested in pursuing the maximum breadth of knowledge that I could achieve, while striving to maintain sufficient depth to avoid dilettantism."

I greatly admire people like this. And having read the 300 pages preceding this description, I believe Dertouzos's self-evaluation to be true. But I wish that he had traded breadth for focus on a few of the interesting subjects he writes about.

As director of what is possibly the leading computer research institution in the world, Michael Dertouzos occupies a lofty position in technology. But to date he has not been as publicly visible as such luminaries as Bill Gates, Esther Dyson, or even his MIT colleague Nicholas Negroponte.

Compared to his peers, Dertouzos is somewhat late to the game in writing a "vision" book, but What Will Be is very much in the same category as such books as The Road Ahead, Release 2.0, or Being Digital. In it, Dertouzos attempts to describe the potential of the emerging world of information technology and compare it to our historic experience as humans.

(Interestingly, each of the aforementioned books was produced by one of the major mass-market publishing houses: Viking Penguin (Gates), Bantam Doubleday Dell (Dyson), Random House (Negroponte), and HarperCollins (Dertouzos). One has to ask whose idea these books really were: the author's or some acquisition editor's.)

What Will Be lives up to its portentous title by covering a sprawling array of material. The book is divided into three parts: The first describes what the author calls the information marketplace, and the new technology (most of which is being invented in the MIT Computer Science Lab) that will support it. The second part considers how the information marketplace will affect peoples' lives, from education to health to art to sex. Part three speculates about the nature of humanity: how it is being affected by a technological society, how it will never change, and ultimately, how the technological and humanistic sides of society can and must merge in generations to come.

Goodies from the lab
The first and second parts of the book are packed with idealized scenarios of the future followed by explanations of how close to realization these scenarios are. These turn out to be wonderful antidotes to the usual breathless gushery about the future.

As a veteran technologist, Dertouzos really does know what's feasible and what's not; as the director of a high-powered research lab, he knows what the industry is likely to implement next year, in 5 years, and even in 10 years. Because of this, his assessments of the future make better reading than others'.

He leavens his prognostications with several highly gratifying diatribes against various false claims and hype. One of my favorites is in the section on how technology is affecting business. He deflates the idea of the Chief Knowledge Officer position that many major corporations have sought to fill, saying that it is silly and useless. He also throws in a few "tricks" involving gee-whiz scenarios that are actually true today, such as one about an artificial-intelligence program that determines digitalis dosages to heart patients.

The scenarios he chooses are imaginative and entertaining, and almost all of them are based on things that people will want or need to do. Yet I can't help but feel that -- as Dertouzos describes it -- much of the technology we're waiting for is of marginal value. Most of the advances he talks about involve human-computer interface, things like virtual-reality goggles, haptic gloves (which respond to hand movements and can exert force on your hands), and, ultimately, "bodynets." He also predicts the inevitable sharp rise in use of digital video, and provides the customary caveat that network bandwidth will be a bottleneck for years to come.

Dertouzos hand-wavingly
claims that solving
metadata problems is
merely a function of
people getting together
and deciding on
standards. Ah, if only
life were that simple.

It's not that these technologies won't enhance many different kinds of experience. They will. It's more that they will do this in ways that don't seem all that important. For example, the effort that it would take a textbook publishing company to produce hundreds of hours of video content for a multimedia educational tool (for, say, biology or geography) probably outweighs its educational value. Of course, there are exceptions, such as a surgeon using a haptic glove to perform surgery on a patient via a robot in a remote location. But the technology to support such a precise and demanding activity is quite far off.

Other enabling technologies that Dertouzos cites could be implemented today. One example is what he calls the "e-form," a specialized sort of message that one sends over the Internet to another computerized device that knows how to interpret it and perform a corresponding action. You can, of course, do this today with e-mail -- as long as you program the device on the other end to interpret the e-form language properly. This leads to one problem with Dertouzos's predictions: His assumptions about the overall interoperability of the information marketplace.

Interoperability takes on a whole new level of meaning when applied to the Internet as opposed to your PC. It's one thing to buy an office software suite that lets you share charts and graphics between spreadsheet, word processor, and e-mail packages. It's another thing entirely to get different organizations to agree on a common language for expressing the concepts and terms that matter among them.

For example, Dertouzos gives a scenario early in the book involving a couple vacationing in remote Alaska. In the middle of the night, the man has serious respiratory problems. A local emergency medical technician hooks him up to a diagnostic device that takes measurements and communicates them to doctors and hospitals specified on the man's health identity card. As a result, of course, he gets proper emergency treatment and recovers; without it he would have died.


Medical meltdown
The medical-emergency scenario is the one chosen by just about every author of what-technology-can-do-for-us books. I'm getting tired of it -- not because it's specious or frivolous, but because it's based upon 100 percent pure wishful thinking. For something like the above scenario to work, several different entities would -- among other things -- have to agree on how to communicate and interpret sophisticated types of data.

Dertouzos calls the technology required here "middleware," a term to which many meanings have been ascribed. Originally, it meant software that does lower-level protocol conversion, like LU.6-to-TCP/IP for network communication or SQL-to-IMS for database access

I prefer the term metadata, or information about information. In the above medical example, metadata would be used to describe physical characteristics of the patient, readings taken by the diagnostic equipment, treatments suggested by the doctors and hospital labs, and so on. All parties involved would have to agree on schemata for describing all these things and more.

Dertouzos hand-wavingly claims that solving metadata problems is merely a function of people getting together and deciding on standards. Ah, if only life were that simple. Metadata models are like shared database schemata. If you have any experience as a database developer, you know how hard it is to get several people within one organization to agree on these. Now think about how hard it is to get an entire industry to agree.

This did happen in the banking industry, because a closely-knit group of companies were able to join together and move forward for the sake of their industry's survival. But in health care, there are too many parties involved (insurers, medical care providers, pharmaceutics firms, patients and their employers, regulators, politicians) with too many differing agendas, plus the information itself is far more complex than banking data.

Bill and Hillary Clinton's abortive attempt to build a streamlined national healthcare system a few years ago was, at heart, an effort to solve the metadata problem for the healthcare industry (remember the President's "health ID card"?). Great idea, impossible to implement.

The medical example is only one of many scenarios that Dertouzos poses where agreement on metadata sets is a necessary precondition of implementation. The metadata problem is as much of a barrier to techno-utopia as the bandwidth problem -- probably more so, because solving the bandwidth problem is primarily a matter of building and buying technological infrastructure, whereas solving metadata problems (and there are many of them!) involves getting people to sit down and work towards agreement on highly specific issues. There is no magic bullet for metadata problems; their solutions will not come out of the MIT Computer Science Lab, Xerox PARC, or anywhere else. (Though they might come out of Microsoft, by fiat, but that's another story.)

Techies vs. humies
The last part of What Will Be is the most interesting, as well as the most revealing about Dertouzos himself. He makes two very important points in this section. The first is that technology, despite its transforming qualities, does not alter basic human nature. Notwithstanding the small number of technogeeks who spend their lives in front of monitors surfing the Web for ersatz experiences, humans will always have physical and emotional needs that technological artifacts will never fully satisfy. All the virtual reality in the world will not replace human contact.

Notwithstanding the
small number of
technogeeks who spend
their lives in front of
monitors, humans will
always have physical and
emotional needs that
technological artifacts
will never fully satisfy.

His second idea involves the dichotomy between technology and humanism, or "techie vs. humie," as he calls it. His self-proclaimed "hobby horse" is to understand the interplay between techie and humie values and try to bridge the gaps between them. He claims that the techie-humie split is not an immutable fact of human nature; instead, it is a relatively recent phenomenon, a byproduct of the 18th century Enlightenment, when philosophers identified reason as separate from faith. He says that the advanced technological path we're on now will worsen the split between techies and humies, and he calls for efforts on both sides to unify the two -- to enable technology and the humanities to coexist peacefully, as they did before the Enlightenment.

One could say that this is nonsense -- that our standards of living in the late 20th century are so much higher than they were in the early 18th century, and that technology is largely to thank for this -- but that would be beside the point. It is also beside the point, though tempting, to compare his sentiment with that of certain conservative politicians in America, who talk of trying to bring back the good old days of the 1950s when everyone knew their place and all was supposedly right with the world.

As Dertouzos points out, both sides need to understand each other better. Techies need to recognize that they have to design things to support human activity, not just for the sake of coolness; meanwhile, humies need to understand that technology is a fundamental part of people, not just a set of tools that you use to satisfy some predetermined humanistic purposes. He says that the way to unify the two sides is to get people to believe that faith and reason can coexist on equal footing, without one taking precedence over the other. They should bolster each other, because neither is a foolproof system. He uses some very clever and intriguing analogies to support his point; you'll have to read the book to discover what they are.

This last topic ought to be the subject of a book on its own. As it is, it's a bit difficult to see Dertouzos as a neutral party in the techie vs. humie wars -- given the 200-plus pages he spends telling us about all of the cutting edge technology coming out of the MIT labs (and elsewhere) and all of the great uses to which society could put it. He is clearly a multidisciplinary thinker, one who feels compelled to reach conclusions and make predictions about every aspect of modern society. Some of these conclusions are bound to be better informed, more accurate, and more profound than others. It would have been better if he had focused more tightly on a smaller set of issues. A book by Michael Dertouzos on the techie-humie split, and how it can be reconciled, would be on my list to read any day.

Title: What Will Be : How the New World of Information Will Change Our Lives
Author: Michael L. Dertouzos
Publisher: Harper San Francisco
List price: $25.00


About the author
[Bill Rosenblatt's photo] Bill Rosenblatt is market development manager for media and publishing industries at Sun Microsystems Inc. Reach Bill at

What did you think of this article?
-Very worth reading
-Worth reading
-Not worth reading
-Too long
-Just right
-Too short
-Too technical
-Just right
-Not technical enough

[Table of Contents]
Sun's Site
[Next story]
Sun's Site

[(c) Copyright  Web Publishing Inc., and IDG Communication company]

If you have technical problems with this magazine, contact

Last modified: