MIT'S FAR-OUT COMPUTER LAB Backed by more than 40 big corporations, the new Media Lab at the high-tech mecca on the Charles River is trying to make computers more useful for businesses and consumers. Some of its machines talk, some make music, and some create electronic newspapers.
By Brian Dumaine RESEARCH ASSOCIATE Brett Duval Fromson

(FORTUNE Magazine) – IN A DIMLY LIT ROOM crammed with piles of black boxes and tangles of colored wires, a young scientist is talking to his computer screen in a loud voice as if it were a slightly deaf friend. Into his head-set microphone he says, ''Schedule a meeting with Walter.'' The computer stays silent. The scientist repeats the command. Finally the computer blurts out in a mechanical monotone, ''When . . . do . . . you . . . want . . . to . . . meet . . . with . . . Walter?'' The scientist answers crisply, ''On Tuesday morning.'' The computer then telephones Walter's computer, which is sitting in an office down the hall, makes the appointment, and reports back, ''All right. It's . . . scheduled . . . for . . . 11 o'clock.'' The computer never schedules early-morning meetings. It knows the boss likes to sleep late. This man and his talking computer are part of the Media Laboratory, an exciting new experiment at the Massachusetts Institute of Technology. The Media Lab, which opened its doors in January, is a gleaming computer research center whose mission is to explore what computers could be doing ten or 20 years from now. Its 120 professors, graduate students, and administrators have two goals: to create new ways for people to use computers and to make them as easy to talk to as a good friend. And it has come along none too soon. The computer industry is in a serious slump. Many people feel they don't know what to do with computers, or if they do, they find them difficult if not downright hostile. Founders Nicholas Negroponte, an MIT professor of media technology, and Jerome Wiesner, president emeritus of MIT and science adviser to President Kennedy, envisioned a lab completely different from places like AT&T's legendary Bell Labs, where the emphasis is on designing faster and smarter computers. At the Media Lab, researchers buy off-the-shelf equipment and program it for new uses in creative fields such as broadcasting, publishing, motion pictures, music, and theater. ''This is an effort to enhance human creative power through computers,'' says Wiesner. ''A person whose only interest is in a deep technological problem doesn't belong here.'' WIESNER and Negroponte cashed in on their connections to sell more than 40 heavyweight corporations -- including CBS, Digital Equipment, Japan's NEC Corp., Apple Computer, and Time Inc., which publishes FORTUNE -- on their vision. Since 1978 the two have raised $40 million from these corporations for the Media Lab's building and computer equipment, and $4 million a year for operating expenses. ''We support them precisely because they are working on such far-out stuff,'' says Jerome S. Rubin, a group vice president of the Times Mirror Co., a sponsor. ''The Media Lab is on the cutting edge of the electronic future.'' Some backers like the creative atmosphere. Says Gerald M. Levin, executive vice president for strategic planning at Time Inc., ''It's worthwhile just to send our people there and have their minds expanded. There aren't many places you can do that.'' Of the dozen major projects under way now, some are new; others have been taken over from other MIT departments. It's a diverse and mind-boggling lot that ranges from talking computers to electronic newspapers, from computer programs for kids to computers that make music. The point of each is to show what computers can do, not to create finished, marketable products. Negroponte doesn't worry much about whether the average customer will soon be able to afford the lab's inventions. He assumes that the price of computing power will continue to drop dramatically, making today's expensive prototype tomorrow's bargain. To make the lab a success, the corporate sponsors have to figure out how to capitalize on its inventions. For their money they get a five-year key to the lab. None of the work is proprietary. Sponsors can wander around and ask questions about the different projects. Negroponte wants to avoid the fate of Xerox PARC, a lab that failed to find ways to communicate its inventions to its parent company, Xerox, and get them onto the market (FORTUNE, September 5, 1983). So far Negroponte has succeeded, but at a price. He says he spends half his time demonstrating projects for the sponsors, and the staff gets distracted as well. One good-natured but hassled researcher programmed his computer to flash, ''What are you staring at, Bozo?'' whenever sponsors linger too long over his shoulder. The lab's roster of researchers reads like an all-star cast of computer talent. The staff includes MIT's Marvin Minsky, widely known as the dean of artificial intelligence, the field that uses computers to try to emulate human thought; and Seymour Papert, the MIT math and education professor who developed Logo, a computer language used by schoolchildren around the world. Not the least of the stars is Negroponte, 42, who looks more like a matinee idol than a walking paradigm of the state-of-the-art technologist. An architect and a pioneer in computer-aided design technology, he prefers using electronic mail to talking on the telephone, never travels without his portable computer, and even keeps in touch with his staff by computer from his retreat on the Greek island of Patmos. ''We want to make the computer a sensory-rich experience rather than the sensory-deprived one it is right now,'' he says. ''If we can do that, the computer will become a commonplace part of people's lives.'' He intends to speed the day when exciting and rewarding creative uses for computers abound in music, art, publishing, film, theater, education, and business. Among the possibilities the Media Lab is exploring: THE OFFICE. The Media Lab believes executives will find computers a lot friendlier in the office of the future. Rather than constantly wrestling with keyboards and obscure codes, the businessman will be able to run his computer by talking to it, pointing at it, or even glancing at it. The Conversational Desktop is the Media Lab's talking computer. It can perform such secretarial tasks as making phone calls and reminding the boss of that important meeting. Christopher Schmandt, a principal research scientist and the man who was talking to the computer about scheduling a meeting with Walter, admits there are limitations to his system, which includes a $5,000 NEC voice recognizer and a $10,000 Sun desktop work station: ''If you ask it to order a pizza, it won't know what to do.'' He insists that he's not trying to put secretaries out of business. Along with his voice, the executive of the future may be able to use a glance or point a finger to communicate with his computer, says Richard A. Bolt, a principal research scientist in the Human Interface Group. As part of his splendidly named Gaze-Orchestrated Dynamic Windows project, Bolt, wearing special glasses, sits facing as many as 40 television pictures, all running simultaneously on a giant screen. Sensors on the glasses track where his eyes are looking; that information goes by cable to a central computer. If his gaze rests on a single picture, the other 39 recede and that one fills the screen. Bolt also has a wristband with a magnetic sensor that can move an image from one location on a computer screen to another. He simply points at the image with his finger, says, ''Put that . . .'' -- indicates where he wants it placed, and adds -- ''. . . there.'' As Bolt's hand moves through a magnetic field, sensors on the wristband send signals by cable to the computer, relaying where on the screen he's pointing. His eye, hand, and voice systems could allow a busy executive to glance at a screen and have it display a report while he talks on the phone. THE HOME. Finding a useful place in the home for the computer is one of the toughest challenges the Media Lab faces. A recent national survey found that only about half of the ten million computers in U.S. homes are used more than once or twice a week. People need better reasons than storing recipes or balancing checkbooks to buy a home computer. Andrew Lippman, associate professor of media technology at MIT, and Walter Bender, a research scientist at the Media Lab, think they have one: an electronic newspaper called Newspeek, meaning a peek at the news -- not to be confused with the sinister double-talk Newspeak language of George Orwell's novel Nineteen Eighty-Four. If the subscriber of the year 2000 is interested, say, in foreign affairs and sports, during the night Newspeek would scan and collect the relevant stories from publications such as the New York Times and the Washington Post. These newspapers would be stored in computers and updated daily. In the morning a front page with crystal-clear graphics and color headlines would pop onto the home computer screen with articles on the latest hijacking to Beirut and the New York Mets' latest six-game slump. To read the story on terrorism, the reader touches the headline or the first paragraph and the full text comes into view. As he reads, related articles that the computer has chosen appear to the side. What Newspeek can't yet do is show photographs and television film clips alongside the articles, but Lippman and Bender are working on that. An article on terrorism, for example, could include a TV interview with one of the hostages. Lippman says Newspeek could be printed out in the home. He adds that as mail and delivery costs continue to rise, it could soon be cheaper to transmit newspapers and magazines electronically. THE CLASSROOM. The school is yet another place where the computer has failed to fulfill its promise. There's only one computer for every 70 U.S. public school students. Computers have failed to spread more widely, Seymour Papert says, because ''they've basically done more harm than good.'' Rather than using computers to explore ideas in language, math, and art, most students are forced to learn computer programming as an end in itself -- a real turn-off. Over the past two years, Papert has been getting students high on computers at a largely black and Hispanic junior high school in New York City. They're encouraged to use Papert's Logo computer language in any way they want, for as long as they want; there's one computer for every three students, and they average two hours a day on it. If a student decides to draw a flower, he can tell the computer to sketch lines and curves. Typing ''RT 45'' and ''FD 10,'' for instance, instructs the machine to draw a line ten units long at a 45- degree angle to the right. While having fun drawing, the student learns a lot about angles. Says Tessa R. Harvey, a deputy New York City school superintendent, ''This is a terrific way to get kids excited about thinking.'' So far Papert's school has better attendance and higher math and reading scores than similar schools elsewhere in Manhattan. With an IBM grant, he will take his project into a tough elementary school in Boston this fall. THE CONCERT HALL. In another corner of the Media Lab, Barry Vercoe, a professor of music and technology, and his colleagues have programmed a computer to play the harpsichord part of a Handel trio sonata on a synthesizer and to follow the tempo set by a conductor. When the conductor slows his baton, a sonar sensor following his hands tells the computer to have the synthesizer play more slowly. The synthesizer can react as quickly as a live musician and plays in nearly perfect sync with a violinist and flutist. Yamaha, a maker of synthesizers and electronic pianos, and a potential sponsor of the Media Lab, is interested in the technology. In the world envisioned by the Media Lab many musicians and composers will use computers to play and write music more creatively. Marvin Minsky, the artificial intelligence guru, who is also a respectable pianist, wants to study what goes on in the mind when someone writes music or listens to it; he believes that understanding how people think about music will ultimately lead to smarter machines. ''After all,'' says Minsky, ''the mind is just a hundred big computers with programs.'' The computer of the future could have helped George Gershwin, for instance, who wrote brilliant melodies but had difficulty orchestrating them. THE STAGE. Marvin Denicoff, an artificial intelligence expert and an award- winning playwright, thinks a computer could help a dramatist write plays. In his vision, still very much on paper, the playwright would draft a scene and then set up a stage on his computer screen by drawing on a rich database of stock characters, sets, and costumes. He would then instruct his electronic actors to speak and move in any way he wished until he was satisfied with the scene. The playwright could also use a computer to show his finished play to potential investors. The Media Lab is already working on the technology that could make Denicoff's electronic actors a reality. Patrick Purcell, an associate professor of computer graphics, is developing a suit that emits infrared signals to be read by four light sensors connected to a computer -- which then generates a stick figure on its screen that copies every movement. With the help of computer graphics, the stick figure could be dressed up to look like anyone from William Shakespeare to Elizabeth Taylor. NHK, the Japanese counterpart of the British Broadcasting Corporation, has already picked up this technology and created a computer-animated host for a new show on the 21st century. HOW SOON any of the other Media Lab projects will become commercial realities is anyone's guess, though Alan Kay of Apple Computer, whose research helped lead to the microcomputer, observes that getting a product from the lab to the market at a price under $10,000 usually takes ten years or so. But he thinks it auspicious that dozens of America's largest corporations have done something they rarely do: pool their resources for long-range research. Typically, Kay argues, American corporations miss big opportunities to develop and perfect new technologies because they insist on tackling the projects singlehandedly. As he puts it, ''They tend to be aggressive and act like members of a hunter-gatherer society, stripping the land and killing off any strangers.'' Japanese corporations, by contrast, can be likened to members of an agricultural society who cultivate their own gardens during the week and then help a neighbor raise his barn on the weekend. ''That's the way big ideas are born and big breakthroughs happen,'' says Kay. The Media Lab, with its ''friendly rivalry and openness where everyone wins,'' he says, is more like the Japanese model -- and should greatly help the U.S. compete with Japan in one of the most important markets of the future.