INTERVIEW:
Question 1
You have been in the technology revolution since its starting years. Can you tell us what
you have found most surprising in these 20 years, from the time when you were working at
Xerox PARC (Palo Alto Research Center) until now ? What is the biggest change that you
have seen?
Answer
Well, looking back 20 years, I guess the best way to look at it is what was
surprising and what wasn't, and there was a mixture, some things were a surprise and some
things seemed inevitable. And I guess going back 20 years, I was part of a group that saw
computers getting smaller and smaller and cheaper and cheaper, and of course this was at a
time when the computers were as big as a house and were very expensive. But we knew they
were going to get cheaper, and that happened; so that was not a surprise. To many other
people it looked like eventually computers would stop getting smaller and cheaper, but
they never have, not yet anyway. So a lot of the assumptions of research at the Xerox
Research Center and in my own work over the years has just been based on the assumption
that computers would be more numerous and cheaper and that has turned out not to be
surprising. They areBut then there have been big surprises which were revealed only by
building them. That is, it is very hard to sit around and predict reliably what an impact
is going to be. So, for example, in the early days of the Internet, the Internet was built
originally in order for people to share expensive computers that would be located at
various universities. Very quickly after the first versions of the Internet were built, we
discovered that they weren't used so much for that purpose, but in fact were used to carry
electronic mail among users. This was a huge surprise, especially to the sponsors of the
early work. And this surprise continues today. I mean, even these days people are
surprised by the impact of e-mail.There are a whole string of surprises related to the
fact that people want to communicate. They don't want to compute as much as we think. They
want to communicate and computers are more and more useful as communicating tools than as
the computing tools that we imagine them to be.
Question 2
So, inside the companies the way of working has changed dramatically and this change was
caused by networking. You invented the first step of networking with the Internet LAN. Can
you tell us the history of this ?
Answer
Well, Internet was invented on May 22nd, 1973, a long time ago. And the
story of its invention requires you to forget all that's happened over the last 23 years,
because we didn't really know what was going to happen. In 1973, a group of us had just
finished building the first generation of the Internet, the wide area network and the
purpose of that was to connect these very large computers around the world. And then when
I went to Xerox Park in 1972, I was the networking guy at Xerox Park. And the team there
was intent upon putting a computer under every desk - not on every desk, they didn't know
that they could do that - the thought was to put the computer under every desk, and so it
fell to me to build the network that would connect them. And this network had the
disadvantage that the computers were quite a bit smaller, the ones under the desk, I mean,
were quite a bit smaller than the big ones, but it had the advantage that these computers
were very close together within the same building and so they could communicate very fast
if we wanted them to.Our first motivation was that we were building a laser printer, the
first laser printer, in 1973 we were building it, and it printed a page per second, that
is 500 dots per inch of resolution, which requires a lot of bits per second, many more
than you could send over the then modem. In those days the modems ran at 300 bits per
second and 1200 bits per second. So we built the first Ethernet having in mind that we
wanted to be able to carry data fast enough to keep this big laser printer busy. We needed
it for a bunch of computers, one under every desk, to send data to the laser printer fast
enough so that it would be kept busy. And, of course, one of the early uses of LANs, local
area networks, has been printing, laser printers. Of course the laser printers these days
are quite a bit smaller and wouldn't really need a fast network, at least that
fast.Another purpose of the LAN was to extend the wide area network; by then we were using
the early version of the Internet to send e-mail and to transfer files. So one of early
uses of this Ethernet LAN was to make it look like this computer network, the Internet,
that was in the building. Of course what happened... I guess I haven't really described
how Ethernet was invented, but the ideas for the Ethernet came from two previous networks,
the early version of the Internet, which was then called the ARPE net, was a network that
carried packets among computers, so the basic technology is called packet
switching--Ethernet is a packet switching network. And the other network was a network
called the ALOHA network which is a packet radio network in 1970 that connected dumb
terminals to a mainframe host computer using radio and it used a particular method of
operating the radio. So the Ethernet was taking the packet switching of the ARPE net -
Internet, and the collision resolution method of the ALOHA network and putting them at
very much higher speed on a coaxial cable.
Question 3
This concept of package transmission can be applied also to the wireless. Is it possible
to have a LAN with wireless technology for the transmission?
Answer
Well, yes. The Internet, as I just mentioned, began with the ALOHA network, which was a
wireless packet radio network, and we chose to put it on a coaxial cable so it could run
very fast. But today, it's possible now to have finally mobile wireless computers and so
now Ethernet can go back to being wireless again after all these years.
Question 4
And what is your vision of the big networks of the future? When we speak about the
infrastructure for digital highways, do you think there will be infrastructure more on
fiber optic, more on satellite and wireless technology or more based on telephone cable?
Answer
The actual plumbing of these high speed digital networks, the answer to the question,
depends a lot on the time, what year it is. But say in the next 10 years, the backbone
networks that carry the bulk of the traffic will be optical fibers, not satellites, not
microwave, not coaxial even, but optical fibers. The much more surprising answer to this
question comes not in the backbone but in the last mile going from the telephone company
central office or the cable television headend into the home. And it is there most people
think that the information superhighway to the home is a fiber to the home. But that's
wrong. It's not going to be wireless but it's not going to be a fiber either, it's going
to be the existing telephone twisted copper wires for the next 10 years or even beyond the
next 10 years, because those wires are installed and they're cheap and everyone knows how
to deal with them. So my--and this is very controversial I might add--but my view is that
in let's say for the next 10 years we'll have optical fiber backbones and telephone
twisted pair connections from those backbones into businesses and homes.
Question 5
And which is the best technology to connect the fiber, to bring the signal from the fiber
optics to the coaxial cable to the home ? The ADSL compression system or maybe other
technology to have a solution to this problem of the last mile?
Answer
Well, this is really complicated. It's hard to say what's actually going to be installed.
I'm almost sure that, as I just mentioned, the optical fibers will not be going into the
home. So then the question is whether it's the telephone wires or the television cable
coaxial or wireless. I'm pretty sure it's not going to be wireless, but that it's going to
be twisted pairs or coaxial. The trouble with coaxial cable and the cable TV systems that
exist is that they tend to be one way. And many of us working on computer networks have a
higher ambition for networks than they be simply one way. And I think this will lead to
the telephone twisted pair, which is two way, as being the dominant one. Now as whether
it's ADSL (Asymmetrical Digital Subscriber Line) or ATM (asynchronous transfer mode) or
something else, that's very hard to say; I guess if I had to bet now, I guess I'd say ADSL
would be a short- term solution but ATM (asynchronous transfer mode) would be the
long-term solution.
Question 6
A more general question. Many people now think that we have Internet, that is, a network
that is a planetary network. Internet arrives in 4 or 5 continents and is useful also for
the multimedia transfer. So why do we need info highways? Will the info highways be
competing with Internet or will there be something integrated with Internet in the near
future?
Answer
Well, the term "information superhighway" is a concept that is
very hard to get people to agree on. In my own model there are six different conceptions
of the information superhighway. They would be corporate networks as there are now
networks within companies. There's no one conception of the information superhighway,
especially when you begin to connect companies together in their corporate networks. A
second conception is the video telephone. The telephone companies are working to upgrade
the telephone network from voice to video; that's taken 100 years to get to this point.
The third conception is interactive television. Television already has video but it's one
way, so a third conception of the information superhighway is that it would interactive or
two way. A fourth conception of the information superhighway is bulletin board systems. In
the United States, for example, there are 70,000 bulletin board systems, small personal
computer-based on-line systems, BBS with 10 to 20 million callers, who knows, but 17
million is the estimate that I've seen. And that's a different conception of the
information superhighway. This would be the grass roots information superhighway. The
fifth conception of the highway would the commercial on-line services like Compuserve,
Prodigy, America On-line and now the Microsoft Network. And then finally the sixth
conception of the information superhighway would be the Internet in my mind.And now these
different conceptions are converging, that is, bulletin boards are being put on the
Internet, the video is going onto the telephone network, which is already interactive,
which is providing on-line services, there's convergence all across all six of these, so
the information superhighway isn't competitive with the Internet; the Internet is one form
of it. In fact, I believe the Internet is the leading form of it, the Internet is the
closest thing we have to the information superhighway. So you see the other forms, the
on-line services, are going on the Internet. Corporate networks are increasingly based on
Internet protocols and use the Internet for connecting companies together. There's now an
effort to provide TCPIP services over cable TV systems, so the Internet seems to be the
one of the six which is having the most influence on the others.
Question 7
But till now there has been a tendency from the government in the USA first and also in
Europe, to want to decide the regulation of the information highways. They want to decide
the frame in which the development of information highways must arrive. Do you think this
is necessary or will it be the market to decide the best technology, the most useful
system and so on ?
Answer
The biggest problem with the phrase "information superhighway" that term, that
analogy, is that it implies a government role, because the highways, the real highways,
the transportation superhighway was built by government and needed to be built by
government. So when we think about these new digital networks as an information
superhighway we are almost led to think that there should be substantial government role
in their development. In the United States there is a strong consensus, and I certainly
think so, that the number one obstacle to innovation and low prices and good service is
the government-created monopolies in telecommunications. And so we're busily trying to
de-monopolize our telecommunications industries to make way for innovation and service and
price improvement. So that to be as direct as I can, I would love to get the government
completely out of any role in developing information superhighway. Right now the
government and in particular the monopolies our government has created, are standing in
the way of progress, so the telephone companies and the cable companies in particular
would be a lot farther along if they had a little competition.
Question 8
Do you think that a network will really change the economy and will be a big opportunity
for employment, and will create new jobs, new professions in the next 5 to 10 years?
Answer
Yes. You know we talk about the industrial revolution and compare it to the information
revolution. The industrial revolution was an opportunity for enormous economic growth and
increasing prosperity and a lot of problems... Democracy was created by the industrial
revolution. The same can be said for the information revolution. It's going to create the
next major opportunity for economic growth. Now just as the industrial revolution had its
Luddites, people who believed that new technology was bad for the world, the information
revolution has its Luddites, people who think that using computers will eliminate jobs or
hurt families. And I don't subscribe to any of that. I'm very optimistic about technology
in general and about information technology in particular.Now there will be dislocations.
There were dislocations during the industrial revolution; transients, the people who used
to make buggy whips were dislocated when the automobile was invented, and it's a very
complicated social issue what you should do with buggy whip manufacturers during the time
when the automobile was being invented. The same thing is true today. There will be many
industries eliminated by the disintermediation potential of these on-line networks; so
it's a social problem what to do with people dislocated in the interim by these changes.
But long-term I have no doubt that this new technology will raise standards of living and
improve the quality of life.
Question 9
And there may be some big risks, social risks for the have's and have not's, and the
problem of socialization because people who are working in the network, are communicating
in the network, don't meet personally, so the sense of communication and social life will
change.
Answer
Well, I've heard this idea that if people were on-line more that they'd be less sociable
and there'd be less interpersonal contact. I've seen no evidence of that. I hear often
that people are worried because we'll be spending so much time on-line surfing the
Internet and using Internet services that somehow this will dramatically diminish
interpersonal contact and create isolated, alienated people. I have seen no evidence of
that. There are extreme cases. There are obsessive hackers who spend 24 hours on-line and
of course there are obsessive people even off line; there are obsessive people who play
the violin 24 hours a day. The evidence I have is that going on-line increases one's
contact with people, because you can be in touch with many more people much further away
more often, but that doesn't tend to come out of your talking to them in person. And in
fact what little data has been collected on this suggests that the more time that people
spend on-line, the less time they spend on television. So they're watching television less
and they're using computers more, not that they're failing to meet people. I think it
enhances communication and sociability.
Question 10
Do you have statistics about the time spent on computers and time spent watching
television ? Do you remember exactly what this change is and in the last year when there
was the switch from television to computer ?
Answer
Well, there is very little information available on this but what I've heard
is, in the United States that is, that people watch television up to 7 hours a day in the
average family. Now looking at those that have personal computers at home, their usage has
been rapidly climbing because there's more and more content available on these PCs
especially connected to on-line services, so the numbers are getting into the low numbers
of hours a day: 1 hour, 2 hours, 3 hours per day. And the evidence is that those hours are
coming out of the time that was previously spent watching television. If you look at
America On-line for example, they've told me that their peak hours of use occurs at 10.45
at night, when people come home, have dinner, and go on-line and the usage increases until
they go to sleep much later in the evening. And so this is coming out of television time..
I think it's pretty easy to argue that watching less television would be good for the
world. With this show as an exception, I suppose.
Question 11
But when interactivity arrives on Internet, and we are close to this, the people will
restart to watch television on computer maybe.
Answer
I don't... It may turn out that the same transmission facilities that carry what we think
of today as the Internet will carry video and therefore television, but it will be quite a
different kind of television. It'll be television that comes on with the subject that you
want at the time that you want, and will allow you to be much more selective about what
you watch and how much of it you watch and how deeply you watch it. So it'll be a more
interactive, more convenient form of television than the kind we now have.
|
 |
