Computing For The People
Michael Dertouzos, 62, grew up in a poor neighborhood of Athens but has spent most of his life at Massachusetts Institute of Technology. As director of the Laboratory for Computer Science (LCS), Dertouzos presided over a symposium in April to mark the lab's 35th anniversary. The event was punctuated by Bill Gates's announcement that he will be donating $20 million to LCS for a new building, designed by architect Frank Gehry. Dertouzos also gave his new benefactor something to think about: He unveiled plans for a new computer operating system dubbed Oxygen that a team of LCS researchers will develop with a $40 million grant from the Defense Advanced Research Projects Agency (DARPA), and for burying a time capsule with a cryptographic puzzle. business week Correspondent Paul C. Judge interviewed him in Cambridge, Mass.
Q: For a long time, you've been advocating that computers need to become more useful--that machines don't understand our needs well enough. Where are we in that evolution?
A: At the infantile stage. For the first 40 years of computing, we correctly focused on making our machines do what they like to do. We were driven by technology. Most of the software today caters to the wishes of the machine. The image I get is that of a car at the turn of the century, with the user in the driver's seat with a bunch of wires tied to each finger: One controls the ignition, one controls the fuel injection, another one controls the carburetor, and another controls the brakes. And I say: "Go from Boston to New York." What we need is the gas pedal, the steering wheel. We need the fundamental human controls.
Q: Do we have a dangerous cultural obsession with glitzy technology?
A: I don't think so....I have so much faith in the ability of humans to throw away things that threaten their survival. I'm convening a group here at LCS to figure out how to deal with the evils of E-mail.
E-mail has a threshold. If you are below that threshold, you are as happy as a clam. If you are above that threshold, you are ready to kill. The threshold is approximately 30 messages a day for most human beings, though it can be pushed to 80 or 100 for real freaks.
Q: How many do you get?
A: Dare I say? I am around 80 to 100 a day, but I have help. So what I'm doing now, I'm declaring war. Since I think I understand the mechanics of E-mail, I'm going to pull together a group here at MIT and try to come up with some effective strategies for when you cross that threshold. You can declare regions of friends and colleagues to whom you respond within seconds. And you can declare regions to which you don't respond at all. I don't think as a society we are obliged to respond to everything we receive just because we are interconnected.
I see a sort of ethics evolving to deal with this. Most of my friends who are in the same position--and I've been talking to [Hewlett-Packard Co. CEO & Chairman] Lew Platt and [Microsoft Corp. CEO and Chairman] Bill Gates and [Intel Corp. Chairman] Andy Grove--they all love to spend about an hour a day on their E-mail because they get a sense of the pulse of their corporation. And Bill, I know, loves it. But when the interconnected population shifts from around 1.5% today to 30% or 40% of the world, which will happen in the next 20 or 30 years, and when the fever of electronic commerce takes over, and the $50 billion a year today goes to the $4 trillion, I predict 30 years from now, everyone in the world will be in the same situation that I am in today.
Q: Is there an engineering solution to this?
A: No, the solution is human, because a human has only so many clicks in him or her. There are some things that can be handled by automation. But a message that can be handled by automation is not really communication. It's a transaction. Real messages that require participation of the human brain and psyche for communicating in the human sense of the word will remain confined to a smaller number. Maybe the 30 can grow to 50 with the help of some tools. But there will always be a limit.
Q: What's the significance of the Microsoft antitrust case?
A: I have to personalize this a little bit because I've known Bill [Gates] for many years. And one part of the antitrust case is rubbing against a technology and business belief for me that is inviolate: Browsers and operating systems are destined to merge. No judge can keep them from merging, and I don't think any judge wants to keep them from merging.
The reason is simple: People want to deal with information the same way, whether that information is sitting on their local machine or on a distant machine somewhere on the network. The same few commands ought to do the same things about viewing information, hearing it, massaging it, displaying it, printing it, acting on it with programs.
I've had big discussions with Bill on which we disagree. He wants to go there incrementally, he wants to modify Windows, and he has to--he's got 100 million copies out there. I believe that's ultimately not as useful to people, it's still very difficult to use. And I'm not picking on Bill. Whether it's Macintosh or Unix or Windows, the systems are impossible. Let's not kid ourselves.
Q: Can technology be harnessed in some way so that the gap between rich and poor doesn't grow wider?
A: There are several things we can do. The low earth orbit satellites are interesting, because half the time they are over the developing world. If we leave them on, there is no other business they are doing at that time. With very little marginal cost we can provide backbone communications to the developing world that are very exciting.
The same argument can be made with the geostationary [satellites] for the ghettos of the developed world. Who's going to subsidize the costs? The richer countries, not just for compassion, which is good enough in itself, but for self-protection. If the differences increase dramatically, we are going to have blood, and it won't be the Bangladeshis attacking the Bostonians. It's going to be the L.A. ghetto revolting against the L.A. suburb--not for reasons of race, but for reasons of money.