XV. OVERVIEW OF FUTURE TECHNOLOGY.
A. Computers of the Future In an article in Scientific American 62-67 (Sept. 1995), Computer Science Professor David A. Patterson (of Univ. of California-Berkeley) suggests the following:
Today's microprocessors are almost 100,000 times faster than their Neanderthal ancestors of the 1950's, and when
inflation is considered, they cost 1,000 times less . . . . Looking ahead, microprocessor performance will easily
keep doubling every 18 months through the turn of the century.
Patterson goes on to say that in the year 2020, one desktop computer will be as powerful as all of the computers in
Silicon Valley today.
B. Wireless Technology. In the same Scientific American 69-71, George Zysman, Chief Technical Officer of
AT&T Wireless Systems observes that in 1983, industry analysts predicted that fewer than 1 million American
would be using cellular technology by the year 2000, while in 1995 more than 20 million Americans do, and the
number of cellular users is increasing 50% per year in the USA, 60% per year in Western Europe, and 70% a year
in Australia. On the other hand, Zysman says that half the people alive today have never made a telephone call.
Industry analysts are now predicting that by 2001, 3/4 of the households in the USA will subscribe to a wireless
service. Zysman goes on to note that every 18 months the size of the digital chip needed to run a wireless phone or
base station shrinks by 50%. He expects that before long, your cellular phone will strap onto your wrist.
C. Roadkill on the Information Highway. On July 11, 1994, University Video Communications (P.O. Box
5129, Stanford, CA 94309), released a taped presentation by Dr. Nathan Myrvold, entitled Roadkill on the
Information Highway. Myrvold is the leading futurologist at Microsoft. Myrvold was formerly a theoretical
physicist who worked with the world-renowned relativity physicist Stephen Hawking. See
Although Roadkill on the Information Highway is dated, it contains many observations that suggest what the future
of technology may hold.
Myrvold starts the video presentation with the question: "How do computing and communication come together
and change our world?"
Myrvold explains that two technologies support the computer revolution: VLSI (silicon chip technology) and software.
Myrvold considers the price/performance ratio of computers over time. At a constant price, performance has gone
way up. At a constant level of performance, the price of computers has gone way down. Myrvold says that
computing power has increased by a factor of 1 million in the 20 years from 1970 to 1990, and that he expects it to
increase by a factor of 1 million in the following 20 years. To make the scale more real, Myrvold notes that a
factor of a million reduces one year to 30 seconds. Thus, when computing power increases by a factor of a million,
that means that the new computer can do in 30 seconds what it would have taken an older computer one year to do.
If there is a million-fold increase in the 20 years following that, then by the year 2010 a computer will be able to do
in 30 seconds what it would have taken the 1970-era computers a million years to do.
Myrvold next talks of random access memory storage (RAM). RAM density increases by a factor of four every 18
months. The price of RAM drops 30% per year. Hard disk prices drop 60% per year. Myrvold predicts that,
taking the size of your hard disk today, in 3 - 5 years you will have that much RAM. Disk storage will similarly
increase. This ignores the possibilities of breakthroughs in optical storage (holographic memory), which would
dramatically increase disk storage.
Thus, in the future we will be able to compute more, and store more.
Myrvold suggests that unimaginable performance will "blow out all stops." However, there are some problems that
can never be surmounted, no matter how powerful computers become. An example is determining the various
sequences in which 59 objects can be arranged. That is slightly more than a pack of playing cards. To compute all
possible arrangements of the 59 items would be involve 1020 arrangements, which is what physicists estimate is the
total number of baryons (protons, neutrons) in the universe. Thus, to print out such a listing would require all the
matter in the universe.
There are also bottlenecks that will not go away. The chief bottleneck will be the human nervous system, which
limits the amount of information that a person can receive. Computers will be able to completely saturate people's
ways of absorbing information, with computing capacity to spare.
On the subject of communicating, Myrvold notes that communicating has not followed the exponential growth that computing has. He likens the telephone company central switching office to mainframe computers, and pbx systems to mini-computers. They ruled the world before personal computers came along. However, mainframes and mini-computers were decimated by the microprocessor. Myrvold thinks two things will do that to the existing communications base: ATM switching and fiber optics.
Fiberoptics provides connectivity. Today we can modulate lasers very well, and can move bits in great numbers.
However, today we cannot target where those bits go, very well. ATM switching controls where those bits go.
Myrvold thinks that the ATM switch is analogous to the VSLI chip, and will make communications wide open.
Phone bills have not dropped like personal computer prices. With ATM switching and fiberoptics, Myrvold sees
competition entering, which will drive the price of communications down. Within 5 - 10 years, voice will be
packaged with other services, like video-on-demand, and voice communications will be offered for free. Today,
voice communication (which is at 9,600 baud) costs 30 - 60 cents per minute. In comparison, the raw communication cost of video is 50 cents/hour for 4 mega bytes/second. Plus, the price of new communications technology will
drop, so that new competitors will be able to get equipment that is better than existing equipment, for a fraction of
the cost of the existing equipment. The existing telephone network providers will be forced to cut prices and
Myrvold sees all this leading to a switched digital network that offers point-to-point high bandwith digital
communication. Information will become a utility.
Edison created the first electrical distribution company, but he needed a killer application ("killer app") to make it
popular. The killer app was the light bulb. Electrical systems were then created around the world, for the sole
purpose of powering light bulbs. Slowly other items were created that had electric motors that tapped into the
electrical system. Soon, electricity ceased to be just the source for light, and instead became a general purpose
utility that could be used for a host of purposes. Myrvold sees the same thing happening to information.
Myrvold sees that our two presently-existing information networks--the telephone network (point-to-point communication) and the cable network (one-to-many)--will merge into one "general information utility." Personal
computers will not supplant televisions as a way of accessing this information utility. Both TV and PC's will
survive. Even your water heater will be hooked up to the information utility. Information will be just as
fundamental a utility as electricity is today.
Myrvold expects to see two types of data: digital on-line data and digital off-line data. Things that are now on-line
will move off line. Thus, TV shows and movies will be available upon demand, and can be stored and forwarded
to be viewed when the consumer wants to view them, and not when the network programs the broadcast.
Myrvold says that each technology has characteristic economics. Economic considerations are the primary
constraint on how people use information. "Change the economics of information distribution and you change the
Before Gutenberg, a book was an object, like a sculpture. The price per copy was an enormous barrier--such a
barrier that people did not even learn to read. Mechanical printing changed that. Every time you make it
significantly easier to communicate, billions of dollars change hands.
Myrvold notes some dangers from the exponential growth of computing and communicating. If you're not fast on
your feet, you'll be out of luck. There will be lots of room to succeed for those who do adapt.
Uncertain is whether the charge for information will be metered by the bit, or based upon the value of the
information delivered. Uncertain is whether the new information utility will be driven by advertising (like radio,
TV, newspapers, and magazines) or more like books and movies, which are purchased by the consumer.
The winners will be the people who have the most open, most flexible business model that allows for the largest
number of variations to occur.
Myrvold does not see the demise of current information experiences. Plays were supplanted by movies, but plays
survive. Movies were supplanted by network television, but movies survive. Network television was supplanted by
cable TV, but network television survives. Videocassettes supplanted network and cable television, but they both
survive. In all those instances, the market just got bigger.
What is the killer app that will make the information utility take off? For cable TV, the killer app originally was
better reception, now more choices. For PC's, originally the killer apps were wordprocessors, spreadsheets and
databases. As for the information highway, Myrvold expects not one but many killer apps.
Myrvold expects the information utility to eliminate the need for warehousing--items will be manufactured upon
demand. Banks now aggregate deposits and put the aggregated sums into the world's financial markets. The
information utility will eliminate banks; you will put your cyberdollar directly into the world's financial markets.
Middlemen of all kinds will be squeezed, since they are in the information business, and people will communicate
directly using the information utility.
Information shows humans off at their best. Gutenberg changed lives for the better. In the new information
revolution, who will win and who will lose is unknown. All we can do, says Myrvold, is try to avoid being roadkill
on the information highway.