This article is from the In-Depth Report The World Wide Web Turns 25

The Future of Computing (circa 1999)

M.I.T.'s Laboratory for Computer Science is developing a new infrastructure for information technologies--the Oxygen system--that promises to realize a vision long held by the lab's director: helping people do more by doing less


More In This Article

Editor's Note: This story, originally printed in the August 1999 issue of Scientific American, is being reproduced as CERN on March 13 celebrates the 20th anniversary of Tim Berners-Lee's original proposal for the World Wide Web.

Last year a few of us from the Laboratory for Computer Science at the Massachusetts Institute of Technology were flying to Taiwan. I had been trying for about three hours to make my new laptop work with one of those cards you plug in to download your calendar. But when the card software was happy, the operating system complained, and vice versa. Frustrated, I turned to Tim Berners-Lee sitting next to me, who graciously offered to assist. After an hour, though, the inventor of the Web admitted that the task was beyond his capabilities.

Next I asked Ronald Rivest, the co-inventor of RSA publickey cryptography, for his help. Exhibiting his wisdom, he politely declined. At this point, one of our youngest faculty members spoke up: “You guys are too old. Let me do it.” But he also gave up after an hour and a half. So I went back to my “expert” approach of typing random entries into the various wizards and lizards that kept popping up on the screen until by sheer accident, I made it work ... three hours later.

Such an ordeal is typical and raises an important issue: for the first 40 years of computer science, we have been preoccupied with catering our technology to what machines want. We design systems and subsystems individually and then throw them at the public, expecting people to make the different components work together. The image this approach evokes for me is that of designing a car in which the driver has to twist dozens of individual knobs to control the fuel mixture, spark advance and valve clearances, among other things—when all he wants to do is go from one place to another.

Doing More by Doing Less
We have done enough of this kind of design. It’s time we change our machine-oriented mind-set and invent the steering wheel, gas pedal and brakes for people of the Information Age. This idea brings me squarely to the goal of my vision for the near future: people should be able to use the new information technologies to do more by doing less. When I say "doing more by doing less," I mean three things. First, we must bring new technologies into our lives, not vice versa. We will not accomplish more if we leave our current lives, don goggles and bodysuits, and enter some metallic, gigabyte-infested cyberspace. When the industrial revolution came, we didn’t go to motorspace. The motors came to us as refrigerators to store our food and cars to transport us. This kind of transition is exactly what I expect will happen with computers and communications: they will come into our lives, and their identities will become synonymous with the useful tasks they perform.

Second, new technologies must increase human productivity and ease of use. Imagine if I could pull out a handheld device and say, “Take us to Athens this weekend.” My computer would connect to the EasySabre airline reservation system and begin interacting with it, using the same commands that travel agents use. The machine would know that “us” is two people and that we like business class, aisle seats and so forth. It would negotiate with the airline computer for maybe 10 minutes, until it found an acceptable flight and confirmed it. I would have spent three seconds giving my order, whereas my electronic bulldozer—the handheld’s software—would have worked for 10 minutes, or 600 seconds. The human productivity improvement in this example is 600 divided by three, which is 200, or, in business terms, 20,000 percent.

Such huge gains will not be possible everywhere, of course. But during the 21st century, I expect that we will be able to increase human productivity by 300 percent as we automate routine office activities and offload brain and eyeball work onto our electronic bulldozers. This transformation will happen in the same way that we offloaded muscle work onto bulldozers during the industrial revolution. We have not yet begun to see these gains from the information revolution. Now we click away at our browsers or e-mail screens, squinting our eyeballs and squeezing our brains. In essence, we are still “shoveling,” but we don’t notice, because we are holding diamond-studded shovels, stamped “high-tech.” So our expectations of what computers can do for us must also change if we are to have a true revolution.

To date, computer vendors have abused the phrase “ease of use.” When they call a system user-friendly, it is tantamount to dressing a chimp in scrubs and earnestly parading it around as a surgeon. When I say “ease of use,” I do not mean incorporating more colors and floating animals into our systems. I mean true ease of use, even if the interaction is only via text. It is inconceivable to me that the differences between browsers and operating systems will persist beyond a few more years. Both access information—one at a distance, one locally—and because people need to do the same things with information regardless of where it resides, ease of use demands that we have only one set of commands for both. The current state of affairs is as ridiculous as if your steering wheel turned your car on city streets but applied the brakes out in the country.

The final way in which new technologies can enable people to do more by doing less is by including everyone in the word “people.” With some 100 million machines interconnected today, we feel pretty smug. Yet that figure represents only 1.6 percent of the world’s population. We think the world is communicating widely, but we still cannot hear the voices of billions through anything other than television and government information feeds. Moreover, the information revolution, left to its own devices, will increase the gap between rich and poor, simply because the rich will use their machines to become more productive, hence richer, while the poor stand still.

We cannot let this happen—if not for the sake of altruism, then for self-preservation. Such disparities inevitably lead to bloody conflicts. And if we decide to help, the potential is immense: the rich could use the new world of information to buy services and products from the poor, as was done earlier with manufacturing. A Virtual Compassion Corps could for the first time in history match the people proffering human help to those who need it, worldwide. In fact, a small group of undergraduate students at the M.I.T. Laboratory for Computer Science have built a Web site ( to do precisely that. And help need not always travel from the developed to the developing world. Imagine a doctor in Sri Lanka who makes $20 a day administering health care to homeless people in Boston via a kiosk, equipped with a remote video and medical instrument connection and staffed by a nurse. The service might cost $5 a visit, and although not perfect, it would be superior to no health care at all.

This, then, is what I mean when I say that people should be able to do more by doing less: bring the technology into our lives, increase human productivity and ease of use, and offer these gains to all. Given this goal, let’s take a look at the computing model over which this vision extends.

The Information Marketplace
My model of the information world in the near future is the same one I’ve talked about for the past 20 years— the Information Marketplace, the full capability of which is yet to be reached. In the coming decade, half a billion humanoperated machines and countless computers—in the form of appliances, sensors, controllers and the like—will be interconnected. And these machines and their users will do three things: buy, sell and freely exchange information and information services. Some $50 billion changes hands over the Internet today. By 2030, I estimate that this flow will amount to four trillion of today’s dollars, or one quarter of the world’s industrial economy. It will come predominantly from the office sector, which accounts for half of that overall economy. Indeed, a large part of the information services of the future will involve a new type of activity—the purchase and sale of information work. Imagine 1,000 accountants from Beijing doing accounting services for General Motors at $1 per hour.

The “free exchange” part of the Information Marketplace will be just as important. It will affect our lives through its family messages; collaborative activities; knowledge-building and accessing capabilities; political, literary and social exchanges; and many new activities.

Given the goal of doing more by doing less and the model of the Information Marketplace, how do we get there in practice? To that end, at the Laboratory for Computer Science, we have just launched a major research project. We expect it to result in a radically new hardware and software system called Oxygen, which will be tailored to people and their applications and will become as pervasive—we hope—as the air we breathe. This multimillion-dollar, five-year project involves some 30 faculty members from the Laboratory for Computer Science, working in collaboration with the M.I.T. Artificial Intelligence Lab.

Designing Oxygen
At the heart of the Oxygen system is the Handy 21, which is like a cellular phone but which has additionally a visual display, a camera, infrared detectors and a computer. The Handy 21 brings the help you need to where you are. Moreover, it is all-software-configurable in that it can change at the flip of a bit (in any country) from a cell phone to a two-way radio talking to other Handy 21s, to a network node near a high-speed wireless office network, or to a plain FM radio. The articles by Anant Agarwal on page 60 and John V. Guttag on page 58 address this aspect of Oxygen.

The second key technology of Oxygen is the Enviro 21. Unlike the Handy, which follows people, this device stays attached to the environments around people. It is built into the walls of your office and your house and into the trunk of your car. The Enviro 21 bears the same relation to the Handy 21 as does a power socket to a battery. It does everything the Handy 21 does but with greater capacity and speed. Enviro 21s may also be set up to regulate all kinds of devices and appliances, including sensors, controllers, phones, fax machines, and arrays of cameras or microphones.

Oxygen interacts with the inanimate physical world in two ways—through these controllable appliances and through the infrared detectors in the Handy 21s. If a door is of interest to your machines, you paste an infrared tag on it. Thereafter, when people point their Handy 21s to that door, the machines read the identity of the door and show what is supposed to be behind it. In other words, the system provides a kind of x-ray vision, helping people relate to the physical objects of interest in their environment.

The Handy 21s and Enviro 21s will be linked by way of a novel network, Net 21. Its principal function is to create a secure “collaborative” region among Oxygen users who wish to get together, wherever they may be. The Net 21 must do so on top of the noisy and huge Internet. It must be able to handle constant change as aggregates of participating nodes rise and collapse. It must find you wherever you are. It must connect to numerous appliances. And it must connect to the world’s networks. This is no easy task. Oxygen will require a radically new approach to networking protocols that draws on self-organization and adaptation and that augments today’s Internet.

Oxygen must also involve perceptual resources, especially speech understanding, and address people’s inherent need to communicate naturally: we are not born with keyboard and mouse sockets but rather with mouths, ears and eyes. In Oxygen, speech understanding is built-in— and all parts of the system and all the applications will use speech. The systems built by Victor Zue and his group can handle narrow domains of inquiry, such as weather or airlines. We are stitching these narrow domains together—and incorporating vision and graphics where need be—to form a new quilt covering a broader front of human-machine communication.

Oxygen’s fifth technology deals with people’s need to find useful information. We are designing Oxygen so that you can first check your own knowledge stores in ways that are familiar to you. The system will allow you to say simply, “Get me the big red document that came a month ago,” forgoing reference numbers and other clues. Oxygen will also check the stores of friends and associates who agree to share their knowledge with you, in the same sense that you might ask a friend or a co-worker a question if you don’t know the answer yourself. Finally, Oxygen will search the vast information stores on the Web and “triangulate,” relating what it finds there to your and your associates’ stored knowledge bases.

Oxygen will also let people off-load routine and repetitive work onto their electronic bulldozers. It will help users write scripts for automating various jobs, as well as monitor and control the many appliances connected to the Enviro 21s. “Turn up the heat.” “Print it there.” “Every day at noon, give me the price of my portfolio and the weather in Athens.” Oxygen will take care of such instructions using a reason and control loop, which allows a person to guide the machine gently as it carries out automated tasks.

The system’s collaboration technology will help people keep track of what they do as they move forward. For instance, the system will keep a hyperlinked summary of a meeting, provided by a human secretary, with the help of speechunderstanding annotations. When you ask what was decided about, say, a new building’s glass roof, it will give you the secretary’s three-word summary—“We eliminated it”—but if you desire will also let you probe deeper into the chain of spoken and video input that led up to that conclusion.

Last, Oxygen will include customization technology that tailors information to individual needs. There will be no shrink-wrapped software. All software will be downloaded onto the Handy 21s and Enviro 21s from the Net 21 network, triggered by user requests, errors or upgrades. The customization technology will also let people adapt the machines around them to their own needs and habits throughout their use of the other Oxygen technologies.

A Claim and a Wish
Oxygen, then, is an integrated collection of eight new technologies: handhelds, wall and trunk computers, a novel net, built-in speech understanding, knowledge access, collaboration, automation and customization. The power of Oxygen lies not in any one piece but in the totality of these human- oriented technologies together. They forge a new computing metaphor that we hope will mark an important shift from the desktop and icons of today, as those innovations did from text-only systems.

I will now stake a bigger claim: I believe that the five technologies of speech (and other perceptual capabilities), knowledge access, automation, collaboration and customization are the only new kids on the block. Out of the thousands of things that we can imagine doing in the new world of information, these five are the foundations on which any new activities that help us do more by doing less will be built. For the next few decades at least, they are the steering wheel, the gas pedal and the brakes we seek—as well as the forces leading to a fullfledged Information Marketplace.

If this claim is valid, it suggests that people who want to exploit the new world of information should explore the capabilities of the new Oxygen technologies. Every individual and organization will have access to them. The ones who will truly do more by doing less will be the ones who learn how to integrate these technologies and their people into a well-oiled, humming whole. And good Oxygen applications that exploit speech, knowledge access, automation, collaboration and customization will make it easier for people to reach their full potential. Imagine a health care application built on top of Oxygen: for knowledge access, it might use Medline (a searchable, on-line database of articles from medical journals, made available by the U.S. National Library of Medicine) and the patient records of hospitals, both available by speech. It could automate routine medical and administrative tasks, help doctors collaborate with one another and much more, taking its application “personality” from the capabilities of the underlying Oxygen system.

I hope that this vision, embodied in Oxygen and other systems like it, will help us break away from our 40-year machine preoccupation to a new era of people-oriented computing. And as we focus our technologies increasingly on human needs, perhaps we can make a bigger wish for the future. The first three socioeconomic revolutions were all based on things—the plow for the agrarian revolution, the motor for the industrial revolution and the computer for the information revolution. Perhaps the time has come for the world to consider a fourth revolution, aimed no longer at objects but at understanding the most precious resource on earth—ourselves.

Rights & Permissions
Share this Article:


You must sign in or register as a member to submit a comment.
Scientific American MIND iPad

Give a Gift & Get a Gift - Free!

Give a 1 year subscription as low as $14.99

Subscribe Now >>


Email this Article