Entrevistas com Michael Dertouzos
A seguir duas entrevistas com o ex-diretor do MIT, falecido em 29/8/2001:
Director of the MIT Computer Science Lab foresees a proliferation of much friendlier computers in many forms.
Tom Spring, PCWorld.com Thursday, February 08, 2001
Michael Dertouzos doesn't come across a revolutionary, but the director of the MIT Laboratory for Computer Science is hoping to finish one. In his new book The Unfinished Revolution, Dertouzos has penned a manifesto aimed at challenging the computer world's status quo.
At the heart of Dertouzos's hypothesis are bedrock principles flowing from his lab's five-year, $50 million Oxygen project, intended to make computers adapt to people, instead of the other way around. PCWorld.com caught up with the 64-year-old Greece-born Dertouzos at his Cambridge office on the eve of a whirlwind book tour to ask him how he would like to finish the revolution.
PCW: Please explain the subtitle to your book: "Human-Centered Computers and What They Can Do for Us."
Dertouzos: Machines are not only hard to use, they are sometimes impossible to use. They are really not serving us. We are serving them. And we are getting use to it. Human-centered computing is about having computers meet our needs. It's about making machines easier to use and as natural a part of our environment as the air we breathe.
Less than 1 percent of the industrial economy flows over the Internet. And less than 5 percent of the world's people are interconnected. We are in such an infantile state right now it makes me cry.
PCW: Is the best example of human-centric computing beepers, cell phones, and personal digital assistants? Are we getting closer with these devices?
Dertouzos: Those are not examples at all. Those are examples of hardware that we may someday use with human-centric computing. Sure, we are going to have handhelds, but not like today.
PCW: Is there any technology on the commercial market that impresses you today?
Dertouzos: Like everyone else I'm impressed by increased bandwidth. I'm impressed by mobility that I see in computers and cell phones getting bigger screens. I'm impressed by Bluetooth. But those things by themselves aren't enough. It's like we are trying to build a gigantic highway system and you're talking to me about asphalt. I'm impressed by good asphalt, but we need architecture and we need to put things together. And that's where the human-centric computing comes in.
PCW: What are the barriers facing the growth of information technology?
Dertouzos: The biggest barriers are ignorance and lack of good ideas, which plague all of us and hold us back. The inertia of the world is a barrier. Consider the huge investment people have made today in operating systems. We have 300 operating systems out there [on a variety of computing devices]. That represents a lot of money out there. We can't just turn on a dime.
If you've been building operating systems for 30 years, just improving them 10 percent a year it becomes kind of hard to change.
PCW: Do you think governments are failing in any attempts to control or advance information technology?
Dertouzos: In America, the government has done what it can. It has removed the right barriers and made opportunities. I would say the same for European and most Asian governments.
Look at Tim-Berners Lee and the Web. He invented that thing ten years ago in a basement at the CERN laboratory. This was just a single guy and his helpers inventing a piece of software that was to concur the world more than any other operating system ever did. Put that in your pipe and smoke it.
PCW: Will desktop PCs someday become dinosaurs, as so many people predict?
Dertouzos: If you mean by PC the temple with a screen and box to which I go and pay my respects several times a day, no. PCs will not vanish, but they will shrink in size and importance. In the future we won't have go to the temple unless we want.
Computers will vanish into walls and ceilings. You'll say something and a microphone will pick it up and a machine will do it. Let me give you an example. Open the drapes. (The drapes in the room open.) Close the drapes. (The drapes in the room close.) My office is part of the experiment we are working on with project Oxygen.
PCW: What will happen to our concept of privacy when everything we do is tracked?
Dertouzos: The assumption that everything we do is tracked and monitored is peculiar. First of all, everything we do today on computers is not tracked and monitored. I have about ten computers at home. I've got three or four here at work. And I've got computers crawling out of my pockets. But everything I do is not tracked and monitored. So I'm careful about the things I care about and I'm not about the things I don't care about.
We have all the technology in the world today to provide as much or as little privacy as we want. The problem is, as a people (in North America), we don't know how much privacy we want.
PCW: Describe the book in broad terms and whom it might appeal to.
Dertouzos: The book is really a call to arms. It's a call for a radical shift in the way that we design and use computers. It's aimed at normal people who could have a lot more if they asked for it. And it's aimed at normal designers who could do a lot more if they focused on human beings. What I'd like for this book to do is to give us enough of a drive to stop this incremental climb on the tree of progress and allow us to get on a rocket and go to the moon.
And human-centricity is the rocket we want. Sure I'm not hitting it square on the head--but I'm hoping I'll awaken people so they can do it better.
During a recent visit to his MIT office, Dertouzos, 64, demonstrated some of the friendlier devices his lab has in mind. First he asked Jupiter, the MIT lab's weather information system, for the next day's weather forecast. A slightly stilted, digitized voice promptly gave us the forecast and then graciously inquired if it could help Dertouzos with something else. Voyager, another voice-controlled system, told us the traffic pattern on Storrow Drive, one of Boston's busiest thoroughfares, up to the last 5 minutes. And a third device, aptly named Pegasus, replied to Dertouzos' request for any flights from Chicago to Boston that day, arriving around 11:00 a.m. with, "There is only one flight that meets your specifications. United Flight 504 from O'Hare landed in Boston at 10:42 a.m. local time. It is 17 minutes ahead of schedule." But even Dertouzos, who paraded these technologies with the pride of a father showing off his children, concludes that wireless is not going to radically change the world as we know it; although the new technologies may bring in their wake some less than welcome trends.
CIO: In your book, you write about the confusion and frustration growing in proportion to the gadgets that surround us. And yet you're such a strong advocate for the power of IT to better our lives. Is wireless technology making the world better or worse?
Dertouzos: Wireless' best attributes match the human ability to move and can convey information, regardless of where and when. And that's important. Wherever you are and whenever you need to be reached, information can be conveyed to you. You can access the Internet, or you can talk to your child. At the technical level, the primary disadvantages are that wireless is not very high-speed, the screens are too small, and the amount of information is limited. At the social level, there are some disruptions and changes when you start to have human beings at the tentacles of communication pipes, because now they're reachable all the time. You're out there enjoying the sunset and the damn thing rings.
Just because we have become interconnected, we have not earned the automatic right to bother anyone we want with our e-mails or phone calls. Nor have we acquired the automatic obligation to respond to messages from others. Just because you are interconnected, you shouldn't mindlessly respond to every push button. Disconnecct if you don't want to be bothered. If the sunset is taking place, turn the damn thing off.
How does this advice reconcile with the growing trend in the business world for 24/7 access to everyone?
I could see that in some corporate situations people will say, "I've got to be able to reach you anytime I want to." And then, of course, you've got to decide if that's the kind of life you want to lead. Is it time to leave the company? Or do you want to have your union fight it? If we take wireless technology and use it as an oppressive tool to keep tabs on human beings, to rein them tighter to the corporate center, I think we'll fail. If we use wireless technology to empower the individual freedoms of human beings, we will win. It's the difference between the Henry Ford era, where workers were treated as interchangeable parts, and the post-Ford modern era, in which workers count and have to be treated with tremendous respect and given the ability to improve the whole corporation by their own sort of individualism. The companies with that understanding are the best companies in the world today.
As the number of users continues to explode, how will wireless networking technology evolve to keep pace? After all, we already have a lot of wireless dropout -- even in metropolitan areas.
We think of wireless as things that move. But the ultimate place we're going with all this is to build a huge, fixed antenna system. To gain more and more high-speed communication -- something we all want -- you can either pump more power into your device, which costs a lot, or you can make the cell size [the geographic area served by a cellular tower signal], the antenna and the surrounding things that can hear the antenna, smaller. Today a cell size is [maybe] 30 miles. When one antenna handles all the cars and mobile users in a 30-mile radius, it cannot give you much speed; it's got to proportion what it has. But suppose that instead of one antenna handling a 30-mile radius, several antennas spoke to one city block worth of people and cars? They're sitting up on telephone poles or electric company poles within a few blocks of each other. Now you can have huge bandwidth and accommodate lots of people using little baby wirelesses with small batteries. You could have a conversation that lasts an hour as you move through several cells, but this is in urban and quasiurban centers. As you go to the deep countryside or to the oceans of the world, you're not going to have that. You will always rely on other media, such as a satellite, which is also wireless, although slower, but at least with that you can reach the remotest parts of the world.
Tomorrow's network is not going to be 100 percent wireless. Everybody is going to win. You're going to have wired networks up the wazoo. You're going to have glass fibers going to stationary antennas. You're going to have coaxial cables. You're going to have wiring like crazy in the urban and quasiurban areas because that gives you speed. It's cheap, and it communicates nicely from one wired location to another. Then you're going to have a lot of wireless with local range. And you're going to have satellites for the remote parts of the world.
There's already a lot of controversy about the impact of cellular towers and wireless transmissions on people's health. Do you really think communities will allow the technology to be as ubiquitous as you propose?
I'm not advocating that we should adopt that model, I'm simply observing that this is where we're headed. I don't see the community protests being higher than the greed for self-gratification and having toys to play with. Do we know that these things are dangerous? If we do, that's why we elect governments and put regulatory agencies together to protect us from such things. But after 50 years in the electronic business, we don't have a solid piece of information as to whether this wireless activity is dangerous to our health.
Currently less than 5 percent of the world uses the Internet. How will wireless technology reach the disenfranchised?
The fascinating thing about wireless is that it can carry speech. Speech can bridge the gap between the illiterates and the literates and the [technological] gap between the Chinese and the English. Think about trendy handheld PDAs that demand that you learn entire new sets of commands when you write on their little screens so that their programs can understand you. Compare that to the H21 [a handheld device developed in Dertouzos' lab] that has no keypad of any kind. You communicate with the H21 through spoken dialogue and by viewing what it shows you. Although keyboards for typing Chinese ideograms are far more complex than those for typing English [100,000 ideograms versus our 26-letter alphabet], designing experimental speech understanding systems for Mandarin Chinese is no more complex than designing those that understand English.
In addition, we can use the low Earth orbit satellites whipping around the world to provide a backbone of wireless communications to underdeveloped countries. These satellites fly over Africa and India as often as they fly over New York. Right now, when they're over Africa and Asia, they're not busy and the marginal cost to leave them on is very low. There is so much more we could do with tax credits from the rich countries to the poor, with private donations, with training, with companies donating equipment. There is a hell of a lot we can do if we have the will.
What kind of changes do you foresee in the way we work and live?
Wireless is just a medium. All it will do is give people on the go the same capabilities that they now have when they're sitting in their offices. But in 10 or 20 years' time, it should make the work environment more seamless. Microphones, for instance, will pick up your voice wherever you are: in the office, at home, in the car. Again, though, we have to make sure we develop this technology to empower people, not machines. Let me give you an example: When you enter a friendly building in the middle of the 21st century, your computer should be able to know what computer resources are around so that it can use them. You've been dictating and have documents, perhaps 30 pages of important stuff, plus some voice and video fragments. You'd like to say, "Let's print this on the nearest computer." Now, there are several ways of doing this. One is that as soon as you enter the building, you broadcast to your computer and everybody knows you're there, but you may not want others to know you've arrived. Yet the minute you broadcast your presence in the building, you're vulnerable. In our laboratory, we did it the other way. All the resources are broadcasting their existence. No human is broadcasting his existence. So when I enter a building, every resource, every printer, every fax machine, every computer in the building is broadcasting, "Here I am! Here I am!" every 20 seconds. As I walk into the building, my computer can measure the difference between signals and know how far a resource is. After I walk around, I have a complete map in my computer of all the resources in the building without anybody knowing that I'm in the building. This is a humancentric position.
Wireless may also have an impact on the way jobs are distributed in the future. Much like we transferred a lot of our manufacturing to Taiwan and Southeast Asia in the 1970s and '80s, I envision wireless technology enabling a huge transfer of office work [which comprises 70 percent of what we do in our economy] outside of the United States and England, the English-speaking countries or the rich countries of the world, to the ones who are very hungry. It's an interesting situation in that labor from a hungry nation can be proffered across national boundaries to a rich nation. It's going to be a massive redistribution of work -- everything from customer service and transcribing to mortgage application checking and insurance adjusting.
But if all this routine office work goes overseas a lot of Americans will lose their jobs. Is that a good thing?
Some U.S. office jobs will be lost to foreign office workers just like some manufacturing jobs were lost to East Asia. That's an inevitable consequence of the lower costs abroad. Short term, the impact will be positive for poor foreign nations who will see their GDP grow as spectacularly as did Taiwan's in the past two decades. The loss of U.S. jobs will have a short-term negative impact both economically and psychologically for those who are displaced here. But as in the industrial era, U.S. workers will move further up the chain, offering sophisticated services, as opposed to the more routine and clerical ones that will be shipped abroad. Long term, I believe both the developing and the developed nations will benefit from the upward economic surge that results from the increased benefits associated with wireless technology.