* He came to the United States as a Fulbright scholar in 1954. "I ended up
in Arkansas," he said. And, in a true display of optimism, he added, "It
was a great way to enter America."
* As I have reminded him often, he has been Director longer than I've been
alive...
* But another so-called improvement that [Missy] mentioned seems to go
directly against Dertouzos' ideas. Users will be able to tell "where a file
was installed." Dertouzos' contention is that users do not even want to
know words like "file" and "install."
* "Simplicity is a bit like marital fidelity, in that while everyone is in
favor of it in principle, practice varies." This is ironic for MLD...
* His Beemer is a V-12? huh? must have removed the special badging, then.
In general, I fail to see the moral superiority of a $30m system in the
hands of a dozen instead of an inferior $30 liberating tool in the hands of
10m people. Simplicity will come from coevolution -- that is to say, the
*perception* of simplicity. There's nothing simple about today's life for
Thomas Jefferson, nor vice versa. Simplicity is a perception of habit.
=============================================
June 24, 1997
Unlikely Warrior Leads the Charge For a Simplified Personal Computer
By JAMES GORMAN
CAMBRIDGE, Mass. -- Dr. Michael Dertouzos, head of the Massachusetts
Institute of Technology's computer laboratory, is railing against personal
computers. "You want to use them as boat anchors," he says. "People should
really revolt. I think we should take up arms."
The question that springs to mind is "What do you mean, 'We'?" Dr.
Dertouzos, 60, does not buy his computers on sale at CompUSA, or sit for 20
minutes on hold waiting for tech support to help him figure out how to hook
up his new printer. He has been running the MIT laboratory, which now
employs about 500 people, for more than 20 years. Many, if not most, of the
people who created the Internet and today's computers and software came out
of that laboratory, where Dertouzos and others now run the consortium that
sets standards for the World Wide Web.
Nonetheless, a computer will not boot any faster, even for an MIT wizard,
and in a new book, "What Will Be" (HarperEdge), and in a recent interview
in his MIT office, Dertouzos detailed what is wrong with the mix of machine
and software that offers the average consumer checkbook programs, Internet
access and headaches. In essence, he says, the personal computer is just
too complicated. Among his specific complaints about the PC are these:
(a) It has too many features, too few of which are the ones the user wants
at any given moment.
(b) There is too much to learn. Dertouzos says his software manuals for the
programs on his desktop computer are equal in length to an entire
multi-volume Encyclopedia Britannica.
(c) The machines often take charge, overruling the desires of the users.
Dr. Dertouzos offers the example of being forced to accept a time-consuming
upgrade to airline reservation software over the Internet when what he
really wanted to do was use the software to make an emergency airline
reservation.
(d) Enormous programming effort and vastly increased sizes of, for
instance, word processors are devoted to trivial improvements in appearance
or presentation.
Calling today's machine "user friendly" because of its endless choice of
fonts and screen patterns, Dertouzos writes, "is tantamount to dressing a
chimpanzee in a green hospital gown and earnestly parading it as a
surgeon."
It might seem to the average computer buyer that these are fighting words,
that Dertouzos is some kind of turncoat, saying what other techies (the
world of Dertouzos is divided into techies and humies, or humanists) fear
to say. Not so.
"Today's machines are just too complex to be accessible," another MIT guru
and sometime rival of Dertouzos, Dr. Nicholas Negroponte, just complained,
acting somewhat shocked, in his column in the July issue of Wired magazine,
home of all things cool and computable. Negroponte, head of the
university's media lab and author of the recent "Being Digital," wrote: "Is
it time for a strike or a users' cartel? You bet it is."
And other experts are ready to chime in. "Personal computers have always
been abysmal devices," said Paul Saffo, a former student of Dertouzos and
head of the Institute for the Future, a nonprofit group in Menlo Park,
Calif., that specializes in management consulting. Over time, he said, they
have become much more powerful, but not any better.
Dr. Leonard Kleinrock, a member of the computer science department at the
University of California at Los Angeles who is a longtime friend of
Dertouzos, coined the term "feature shock" years ago to describe what
happens when people get new computers all filled up with things that they
never wanted in the first place.
Even Jonathan Roberts, director of product management for Windows at
Microsoft, arguably the source of some of the complications Dertouzos and
others deplore, is apologetic. The biggest complaint that Microsoft gets,
Roberts said, is that "you have something that worked and then you install
something else and what you had doesn't work."
The sour notes from Dertouzos are particularly surprising in that he is a
thoroughgoing technological optimist. "Toys, toys, more toys!" is the motto
of his laboratory, he says, and it is one he embraces at work and play. He
drives a 12-cylinder BMW. He designs equipment for his 37-foot power boat.
Dertouzos traces his optimism partly to growing up in Athens during World
War II, while his father, who was an admiral in the Greek navy, was in
Egypt. After the war, other problems seemed small. Besides, he said, "It's
easy to be a pessimist." He came to the United States as a Fulbright
scholar in 1954. "I ended up in Arkansas," he said. And, in a true display
of optimism, he added, "It was a great way to enter America."
He received his Ph.D. in electrical engineering from MIT in 1964, after
spending some time in research and development. He started teaching at the
university immediately, and quickly founded a high-tech company making
early computer terminals in 1968. He sold that company in 1974 when he
became the head of the MIT laboratory. He now frequents the national and
international halls of power in industry and government. He is currently
co-chairman of the World Economic Forum, a business-financed research
organization in Geneva.
Even when he is criticizing computers Dertouzos likes to point to the new
and wonderful things these machines do, like enable users to cruise the
World Wide Web. And when he complains of the sad state of computers now,
the critique is only a prelude to how good the future will be. In effect,
he is saying that the emperor is naked and shivering -- right now. But the
court clothiers are aware of the danger of pneumonia, so next year's suit
will really be warm and comfy.
Dertouzos looks forward, with varying degrees of enthusiasm, to a world of
intelligent kitchens that plan menus, cook meals and tell people what is in
the pantry, to smart cars, smart closets that pick out matching clothes,
even to a smart sink that reminds people to use a dental pick because of
their risk of periodontal disease. Fortunately, the sink would be difficult
to design and very expensive. Underlying all these fantasies is the notion
of computers' becoming both smarter and easier to use.
Some of his friends say that his political skills, which have made the MIT
laboratory so successful, may have tempered some of the criticisms he
levels in the book. And Dertouzos does not spend his time pointing the
finger of blame.
While Kleinrock is quick to say that "anything Microsoft does makes it
worse," Dertouzos concentrates on the fact that the industry is in its
infancy. Dertouzos compares today's personal computer to the first
airplanes, familiar from old films that show them laboring to escape the
ground and then crashing in a heap. Computers crash, and are difficult to
use, largely because despite rapid technological change, they are still
very new.
Another problem he sees is that engineers and programmers love to add a new
twist here and a new capability there. Roberts at Microsoft admits, "It's
much more difficult to make the everyday easier than it is to add a new
feature." But the new features sell, Dertouzos points out. The marketplace,
in the shape of the consumer, buys bells and whistles. "We're the culprit
in that," Dertouzos says.
Indeed, says Saffo, "we are unwitting co-conspirators" in a system in which
poorly tested products are sent to market for the buyers to test. In a
recent study done as part of the Computer Industry Policy Project at
Stanford University, Abron Barr and William Miller came up with evidence to
show what observers of the industry have said all along. To make money with
software, a company must get its product on the market first, and fix it
later. Early examples of software, like a new "Word" or "Quicken," are
called beta versions.
Saffo observed, "We ask users to pay for the privilege of being beta
testers." And the users do pay. For the moment, the economics of computers
are such, Saffo said, that "I would fire the CEO of any company that held
off its product until it was ready."
The answer, Dertouzos said, is that consumers must demand simple, usable
machines that work more like cars than science projects. Turn them on and
drive. No one would buy a minivan that had a 600-page manual on how to
drive it. Why should people do that with computers?
One way to achieve simplicity is to bury deeper what the computer does, so
that, for instance, a user can ask it questions in common speech. At MIT, a
system called Galaxy, being developed by a team under the lab's associate
director, Dr. Victor Zue, can answer questions posed to it about weather,
airline reservations and restaurants in particular locations. It has been
in the works for 20 years and so far has cost about $30 million.
Another approach suggested by Dertouzos is to allow users to tailor their
computers to themselves, allowing them, in effect, to become programmers.
Of course, then the programming has to be simplified, or else things are
made worse, not better.
Simplicity is a bit like marital fidelity, in that while everyone is in
favor of it in principle, practice varies. Microsoft, for instance, says
Roberts, is on the simplicity train. In the next version of Windows there
will be a feature called "on now," which will make a computer come on right
away, as they used to 15 years ago, instead of taking minutes, as they do
now.
But another so-called improvement that Roberts mentioned seems to go
directly against Dertouzos' ideas. Users will be able to tell "where a file
was installed." Dertouzos' contention is that users do not even want to
know words like "file" and "install."
"I'm really quite pessimistic," Kleinrock said, about improvements in
personal computers. And Saffo said that in truth the problems were never
solved. Technologies change so fast that they become obsolete before they
mature. Computers are not likely to get simpler. Something unforeseen is
likely to replace them.
For the moment, Saffo said, he is pleased to see Dertouzos criticizing his
intellectual offspring, while he paints a rosy picture of the future.
Dertouzos trained many of the people who have made the current computers
what they are, Saffo said. "It's the sign of a good teacher to criticize
the work of his students," he said. "He's playing exactly the role he
should play."