FUTURE COMPUTER TECHNOLOGY AND ITS IMPACT

FUTURE COMPUTER TECHNOLOGY AND ITS IMPACT

The digital computer has penetrated many professional
specialties and many aspects of everyday life; conputing
technology itself is entering a new and explosive phase
of development. What does this portend for the future?
Let us consider the future impact of computers on
society, business, industry, and the military. First,
however, it is desirable to call attention to a particular
feature of the computer and to one of the ways in which
it is used. Second, I want to describe the advance in
computer hardware in the 15 years of its commercial life-
time, and the anticipated changes during the next decade.
Traditionally, we think of a computer as a device to
do arithmetic. But, as Fig. I shows, it is possible to
encode other information in terms of numeric symbols. For
example, in the telephone dialing system the letters EX
are represented by the two digits 3 and 9. This principle
allows the digital computer to accept any information that
Any views expressed in this paper are those of the
author. They should not be interpreted as reflecting the
views of The RAND Corporation or the official opinion or
policy of any of its governmental or private research
sponsors. Papers are reproduced by The RAND Corporation
as a courtesy to members of its staff.
This paper was presented to the Board of Trustees
of The RAND Corporation and the Project RAND Air Force
Advisory Group in November 1965.
can be encoded in symbolic form. Moreover, by giving the
computer special kinds of operations, we can have it
manipulate such symbolic information. It becomes an
information processing machine. As Fig. 2 indicates, the
computer can parse sentences, drawing a picture of the
sentence structure. If the symbols happen to be those of
algebra, the computer can perform algebraic manipulation.
Or, it can process symbolic pictorial information and re-
construct a picture (Fig. 3). We should think of the
digital computer as a device that can accept and process
any information encoded in symbolic form.
There is a particular way of using a computer called
modeling or simulation. Consider a bottle of soda pop
(Fig. 4) in which a few chemical radicals and compounds
are present. In the mixture itself there are oxygen,
carbon dioxide, and other chemical ions. Oxygen, nitrogen,
and carbon dioxide are exchanged across the air-liquid
interface. We can mathematically describe the chemical
activity and the energy balance in this system, and com-
pute its equilibrium state. What we have done is to model
a real-life, physical system in terms of a set of mathe-
matical relations.
A model can also be a description of a biological,
physical, economic, political, military, financial.
physiological, psychological, organizational, or any
other system. A model identifies vnriables in a system
and states the relations between them. If the variables
of the model and the relations between them satisfactorily
represent the real-life situation, then the model is a
precise description of the real-life situation. The modelwill exhibit the same behavior as that part of reality
which it simulates. It can be used to perform experiments;
it can be used to explo,. bituations or cases which may be
impossible : replity or which we hope will never happen.
Now let's briefly trace two branches of computer hard-
ware development. Figure 5 shows what switching circuits
looked like in the early 1950s. Figure 6 represents one
form of contemporary circuit technology, the so-called
solid-logic construction, containing both printed and de-
posited circuits. The large elements are resistors (Fig.
7); the small squares at the right of the figure are
transistors, 40/1000 in. square. (Transistors are fabri-
cited approximately 600 to the square inch.)
We are just entering the era of integrated circuits.
The small square in the center of Fig. 8 is the circuit
proper; the rest is mechanical packaging and external
connections. The material at the bottom is a piece of
thread. Figure 9 shows two examples of an advanced form
of integrated circuit technology. A slab of pure silicon
is successively doped and processed to form a grid of basic
circuit components. The basic background grid (Fig. 10) is
a rectangul.•r area about one-tenth of an inch long and one-
sixteenth of an inch wide. Such a rectangle contains about
40 circuit elements, of which 30 or so are transistors. By
further deposition of conducting and insulating material
over the background grid, the customized fully integrated
or microcircuit is built (Fig. 11). The postage stamp
area about half an inch by slightly over five-eighths of
an inch contains a total of about 800 transistors 350 circuit components. The packing density in this ex-
ample is about 3000 circuit elements per square inch. By
the early 70s, integrated circuit technology is expected
to produce packaging densities of 200,000 circuit elements
per square inch, an improvement over present art by a
factor of about 70.
Figure 12 shows three generations of circuit packaging.
In the background is a contemporary form of printed cir-
cuit plug-in packages: in the left foreground are commer-
cially available integrated circuit packages; and, front
center is the fully integrated microcircuit. Each repre-
sents the same electronic capability.
Let us turn our attention to the storage component of
a digital computer. The backbone of storage technology
has been and may well continue to be the magnetic core art.
Figure 13 shows a large magnetic core plane from the early
50s; the small one in the foreground is contemporary. Each
represents the same storage capacity, 4096 binary digits,
but the small modern plane is faster by a factor of 15
or so. Figure 14 illustrates the steady decrease in size
of magnetic elements used in stores over the years. The
final entry on the left is hardly visible, and has been
reproduced much enlarged at the right. The tiny X-55
annulus at the top right is magnetic material, sevein
thousandths of an inch inside diameter, twelve thcusandths
of an inch outside diameter. Such cores are fabricated
into a so-called plane with several wires threading each
core (Fig. 15).
The magnetic core store also comes in large economy
sizes; Fig. 16 shows one with a capacity of 16-binary digits. Magnetic storage also comes in other forms.
Figure 17 is a disc store with a capacity of 60-million
binary digits; and Fig. 18, an even larger store with a
capacity of about 200-million binary digits.
There are also new forms of terminals enabling men to
communicate directly with machines. Figure 19 shows a
production model of a personal console which is connected
by means of a telephone circuit to a center computer. The
entire conversation between the user and the machine is
carried on via this electric typewriter. Such personal
console stations have given rise to the so-called on-
line, time-shared computing system (Fig. 20). Many users
scattered throughout a building or over a large geograph-
ical area are connected to one or more large, central
computers by a communication system. 7o each u3er the
machine appears to be solely his; but, of course, its
enormous speed enables it to circulaZe among all users,
giving attention to each in turn.
Another new terminal is the graphical input-output
(Fig. 21), which enables a user to input any kind of
graphical or pictorial information and receive the same
kind of output. Slide material may also be projected onto
the rear of the tablet surface for convenient tracing into
the machine (Fig. 22).
Looking briefly at computers over-all, we find that
the machine of the early 1950s was typified by RAND's
recently retired 13-year-old JOHNNIAC (Fig. 23). Today's
typical machine (Fig. 24) is not too much different in
external size, but is somewhat cheaper, contains much
more hardware and eight times the storage, and is by a factor of 30 to 50. Computers now can be designed
to fly in space--e.g., the GEMINI machine (Fig. 25).
We are just beginning to see the introduction of
integrated circuits into commercial computers. The arith-
metic section of one is shown in Fig. 26; Fig. 27 shows
the storage part of the same machine. The complete ma-
chine appears in the lower left corner of Fig. 28; the
objects surrounding it are various terminal devices for
coupling the computer to its environment--a display, a
console, a typewriter, and a small magnetic tape unit.
Comparing the old and new computers of Figs. 23 and
28, we find that the 1953 machine weighed about 5000 lb,
had a volume of 300 to 400 cu ft, and required about 40
kilowatts of power. The contemporary computer is a
hundredfold lighter (about 50 ib), a thousand times smaller
(about 1/3 cu ft), and requires 250 times less power (150
watts), Moreover, it has twice the storage and runs about
ten times as fast as JOHNNIAC.
We can summarize the amazing progress of computer
hardware technology in a few trend charts. Figure 29
shows the change in size. From 1955 through 1965, the
size of a central processing unit with its storage has
decreased by a factor of about ten. From 1965 through
1975, the impact of fully integrated circuits is expected
to produce a further reduction in size by a factor of
about 1000. For the two decades of 1955 through 1975,
Figures 29-32 are taken from P. Armer, "Computer
Aspects of Technological Change, Automation, and Economic
Progress," The RAND Corporation, November 1965.
there will have been a size reduction of 10,000 in the
art for building central processing units.
Figure 30 refers to the cosf: of computing power--
not the cost of a computer itself. In the first decade
of the computer's existence the cost of doing a million
operations decreased by a factor of about 300. By 1975
the cost will decrease by another factor of 300 to less
than one 200-thousandth of it3 1955 value.
Figure 31 shows how machine speed has changed. From
1955 through 1965 the internUl speed of the computer in-
creased by a factor of about 200. By 1975 it is expected
that the speed will increase by another factor of 200 or
so; so that by the mid-70s, ire can look forward to doing
computer operations at the rate of about a billion per
second.
Finally, let's look at the installed computing nower
in the United States (Fig. 32). In 1955 all installed
computers working together coild do about 500,000 addi-
tions per second. By 1965 the machine population could
do about 200 million additions per second, an increase
in capability of about 400 fold. If the same growth rate
continues through 1975, capabiLity will increase by
another factor of about 400. 'A somewhat less optimistic
projection for the coming decace still sets the expected
growth at 20 fold or so.) The number of computers in
the Air Force alone has increased from 350 in 1963 to
nearly 700 in 1965.
What does this all add up to? Beginning in the early
1970s, computers will be small, powerful, plentiful, and
inexpensive. Computing power will be available to who needs it, or wants it, or can use it. He may have it
by means of a personal console connected to some large
central computing facility, or he may own a small personal
machine.
We all know, however, that the computer is more than
a piece of hardware; it has to be programmed. Historically
computer programming has been expensive and time consuming,
but we expect the future to be different. Cheap hardware
will enable us to consume vast amounts of computational
power to make a machine convenient and attractive to a
user. Furthermore, with the current or near future state
ot computing knowledge, we can frame languages including
appropriate symbols and syntax which are completely natural
to a novice user and to a user trained in any professional
specialty. We can design a tool for a given individual
from the ground up; a tool to match his normal training
and way of thinking.
For example, most automobile drivers don't bother to
understand the details of the engine under the hood, or
even how the automatic transmission works--such knowledge
wouldn't help them to drive better. Similarly, the com-
puter user of the future will not be able to perceive the
inner details of the machine, nor would it help him if he
could. Communication with a machine is becoming that easy.
The new class of users will no more have to be programmers
of the traditional kind than an auto driver has to be a
mechanic to handle his car.
Such astounding and explosive changes in technology
and the growing ease of communication with a computer are
almost certain to have a staggering impact. What are of the possible consequences of this expected tremendous
growth in information processing?
Extrapolating into the future can sound like science
fiction, but I hope that my predictions now have a cred-
ible basis. My visual summary of the change in computer
hardware could be paralleled by a corresponding summary
of research in application of computers to new and varied
tasks. In particular, we should note the growing capa-
bility of the computer with graphical and pictorial
information.
Various projections have been made of computer
achievements in the 1970s. Let us note one such set of
expectations.
"o Computers will be readily available as a public-
domain service (but not necessarily as a regulated
monopoly).
"o Information per se will be inexpensive and readily
available.
o Large and varied data banks will exist and be
accessible to the public.
"o Computers will be used extensively in management
science and decision-making.
"o Computers will be economically feasible for firms
and activities of all sizes.
"o Computers will process language and recognize
voices.
"o Computers will be used extensively at all levels
of government.
"o Computers will increase the pace of technological
development.
P. Armer, loc. cit.
Let us assume that these expectations have definitely
materialized by the mid-1970s, and consider a period
further distant. Although some of the suggestions that
I will make might arise late in the 1970s, I want to pick
a time more remote; let us consider the mid-1980s, perhaps
the Orwellian year of 1984.
I am not making predictions, but rather I am suggest-
ing things which in my opinion computer technology can
make possible. Whether these suggestions actually materi-
alize into fact will depend, of course, on many other
things--such as political, social, and economic forces;
rate of capital investment, rate of production, etc. The
computer will contribute to our future in two ways: a)
it will make some things possible because of its capa-
bility as a research tool; b) it will be an integral and
operational part of other systems which without the com-
puter would have been impossible. Behind everything I
suggest stand the assumptions which I hope you now readily
accept: computers will be inexpensive, small, powerful,
and plentiful; modeling will be a powerful technique; the
computer as a tool will be a user-oriented device.
I must at least touch on the issue of general social
impact first. The computer is helping technology to move
so swiftly that professional skills rapidly become obsolete
and large blocks of employment openings disappear from one
industry to reappear in another. Frequent retraining and
re-education is likely to become the normal way of life.
Change, not status quo, will be everyone's lot--in civilian
as well as in military careers.
As we shall see, the computer can assist pecple in
adjusting to this new state of affairs. Certainly the
introduction of computers into industry will improve pro-
ductivity, but will dislocate jobs. It is easy to believe
that men will be without work. However, I am partial to
a more optimistic view. The computing industry per se is
creating n'w jobs, as well as moving old ones to new
places. The wants of society are not now being met, and
even with the increased output from the economy that
better productivity can bring, plenty of jobs will proba-
bly be available, f though they may not be in the "right"
places. We must face and solve the retraining and re-
education issue, a problem which will not be limited to
the labor force. Each professional specialist or admin-
istrative official faces the problem of continuous re-
education and adaptation as well.
Ey the 1980s, the use of the computer as a teaching
machine will have increased the entire pace of education.
It is an ideal device for exercising, instructing, and
examining students on a large amount of material. Sophis-
ticated training films produced by computers will give
students deep, rapid insight into physical and scientific
problems.
Students will have computational power available to
them wherever they may need it. The public school systems
and universities will have to provide each student with
For example, the Bell Telephone Laboratories has
produced a twenty-minute film--Force. Mass and Motion--
by computer, which gives great insight into gravitational
laws.

computational support, either as a personal conscle con-
nected to a centralized large computer or as a smail com-
puter of his own, perhaps the size of a small cereal box.
Parenthetically, in most homes I foresee a personal console
(or small machine) for use by the entire family--it will
be in effect another appliance.
With the increased pace in education, students are
likely to complete their formal schooling much sooner;
alternatively, they will acquire more training in the same
time. Consequently, everyone will have more productive
capability, which in turn will speed up technology and
science. There may be an even more overwhelming effect:
if it is true that youth is an important aspect of sci-
entific creativity, the increased pace of education will
result in more young productive years.
The rapid changes in technology caused by the computer
tend to make technical skills obsolete, but the computer
will help to alleviate the very situation it is causing by
making possible not only rapid re-education and refurbish-
ment of technical knowledge, but also swift acquisition
of new skills.
For example, in the Air Force, manpower skills will
be developed more rapidly; an officer should achieve a
much higher level of technical competence much earlier
in his career, and be able to refresh his abilities
readily. By means of models we can exercise him in a
variety of situations, and so sharpcn his judgmental
ability much sooner. Similarly, the computer might be
exploited as a training device in military aid programs
to improve the technological skill level in underdeveloped
countries.
By the 1980s, I expect the tempo of scientific dis-
covery to inctease. I previously granted that the pace
of technology will increase; I now refer to the acquisi-
tion of new scientific information. Computers will allow
data to be displayed in ways not otherwise possible; they
will let the scientist observe situations which he could
not otherwise examine.
The computer will be the most important tool ever
available for the conduct of research. Sophisticated
mathematical and simulation :echniques will be widely
exploited for many situations and systems which scientists
would not otherwise be able to study; e.g., solving mathe-
matical models of the atmosphere and displaying the re-
sults as motion pictures in order .o study weather patterns
in a vastly speeded-up time scale. Such a capability
would be a valuable adjunct to Air Force operations.
In the future, experimentation by computer will be
less expensive than other methods, permitting scientific
investigations otherwise impossible. In fact, I would
not be surprised to find laboratories tending to go out
of style. The nan-and-his-computer may well rank ahead
of the man-and-his-laboratory as the source of new sci-
entific knowledge. I foresee a future in which the
principal ifurcLion of the laboratory is to validate com-
puter models. The quickening pace of scientific investi-
gation will mean new capabilities, new materials, new
technologies, and new tools for the Air Force.
Because vast quantities of information will be trans-
ported from place to place, there will be an enormous
demand for communications services. The networks, however, will have become totally digital; no
longer will we transmit analog voice or video signals.
Cheap digital components will allow the use of sophis-
ticated techniques for removing redundancy from signals
before digitizing and transmitting them. When error-free
transmissions are important, controlled redundancy will
be reinserted into the messages. It is reasonable to
expect that all transmissions will be encoded and hence
private. I can foresee that all the voice, video,
facsimile, and data transmissions needed by a place of
business-or a residence--will be handled digitally cier
a broadband cable.
We can expect to know how to modify and to control
weather, although we may not have the large energy sources
required for a working system of weather control. The
computer will have made this possible, because of the
general increase of knowledge which it will have supported
and the understanding of atmospheric physics achieved
through modeling and simulation techniques. If we achieve
a working capability for weather control, the consequences
will be measured in the billions of dollars. Universal
good climate will eliminate bad real estate. The Nevada-
California deserts could become the breadbasket of the
United States. Crop failures will vanish because sunshine
will occur in the right amounts at the right tikaes.
"Weather Central" will arrange the storms and probably
even publish an annual schedule of rains and snows. The
weather schedule will be as well known as the dates of
holidays; one may well buy it at the Government Printing
Office for a nominal sum. The ability to control andmodify local weather is obviously important to Air Force
operations.
I foresee that our entire engineering design process
will be computer-based. By means of graphical terminals,
the engineer will be able to converse directly with his
machine. He will sketch only the roughest outlines of a
design and let the machine provide all details. Before
the device is built; the machine can exhaustively test
it for him by calculating and simulating its performance.
Engineering drawings will be unnecessary; the blueprint
will be replaced by (say) a roll of magnetic tape con-
taining all the details for automatic fabrication tools.
The medical and biological sciences will probably
use the computer more extensively than the physical
sciences. Hospitals will have become completely computer-
supported in medical as well as business and administrative
aspects. Because they are cheaper and more accurate, com-
puter models will handle most laboratory work. A computer
will be connected to surgical patients in order to monitor
body processes and to warn of dangerous incipient condi-
tions. Consequently, much more daring surgical and medical
procedures will be used. For example, a computer analyzing
electrical activity of the brain can regulate the position-
ing of microelements for brain surgery. Similarly, com-
puters will be used to control prosthetic devices. Post-
operative or intensive-care patients will also be monitored
by a computer.
Even more dramatically, the increased pace of sci-
entific discovery owing to the use of the computer in
research may contribute to the extension of the human lifespan. Through computer analysis and experimentation
with enzyme structures and with the basic building blocks
and codes of proteins, we may learn how to




synthesize specific enzymes, which will make possible the growth of
new body parts or organs, or the revitalization of an
entire body by causing it to resume growth. In fact, if
we learn enough about enzymes and body building blocks,
we ought to be able to grow an organism to specification.
The implications of a significant increase in human
life span are overwhelming. The population explosion
problem would be enormously more important. Industries
based on mortality tables might not survive. Other
organizations would have to find new advancement oppor-
tunities for young people, since present employees would
have a longer productive life. All such changes would
certainly affect religious and social patterns.
The possibilities in medicine and biology have con-
sequences for the Air Force. The computerization of
hospital functions also applies to field medical operations--
keeping track of patients, keeping records on medication,
monitoring patients for dangerous conditions, etc. Per-
haps medical treatment might even be computer based or
even computer controlled. Moreover, body behavior in un-
usual situations or in novel vehicles can be studied
through models.
It has been forecast that by 1975 the files of in-
formation which society needs to govern itself will be
computer based. I refer to information on real property,
credit status, legal status, financial status, licenses,
and so on. In view of population increases, will have to depend on the computer. Hopefully, this will
make the government more efficient but less expensive.
Because of the computer's ability to accept and correlate
information from the many large data banks and files which
will exist, there will be possible in intensive social and
personal surveillance by any agency that elects to do it.
The depth of surveillance may well surpass simple invasion
of privacy.
A criminal activity data bank is one such information
file that will undoubtedly exist. Even today, the computer
is beginning to contribute to the control of crime by
rapidly retrieving information from files. The future role
of the computer is bound to grow as its data banks expand
and as it becomes better at making inferences from frag-
mentary factual information. A machine will be able to
provide much more incisive indictments about crimes and
criminals. Moreover, for each case it will undoubtedly
recommend a treatment for curing rather than impounding
the criminal.
Remember that in the future computing power will be
readily available to everyone, either is a small personal
machine or as a personal console. The computer will
certainly be useful to society in combating crime. But
might it not also help the criminal plan his crime or
the large criminal organization manage its affairs?
Certain implications of future computer technology
will be peculiar to the Air Force. Without a doubt, any
officer or enlisted man who can use it can have computing
power. If necessary, we will be able to build a machine
the size of a cigarette package. As in civilian lifeall communications will have become digital, making com-
puter technology an integral part of communication equip-
ment. Everyone can have private communications by using
a small personal computer for scrambling. The individual
may be permitted to use communication satellites; his
personal computer will assemble the message, encode it,
handle the error-control problem, provide secrecy, etc.
We can provide voice, video, facsimile, and data trans-
mission to individuals as needed--in each case, with
privacy. Computers with wideband data links can provide
graphical communications, allowing widely separated Air
Force elements to discuss plans, documents, maps, etc.,
as though they were in private conference.
We are currently hearing discussions of control of
general war. It seems to me that no enemy will believe
that the United States can control a general war unless
he sees, among other things, a credible and tight command
and control system that is well-trained and regularly
exercised. In fact, in the spectrum of possible deterrent
mechanisms, I believe that a computer-based command and
control system could be a credible deterrent--just as
weapons are.
On this next suggestion I am less certain of the
time. I suspect that it's ten or twenty years further
away than anything I have touched on so far--close to the
end of the century; it is perhaps the most dramatic effect
that computer technology could have on the Air Force:
Computer technology may make obsolete traditional warfare--
warfare in which destructive energy is delivered on an
enemy by weapons. Let me suggest a plausible argument support this conjecture. Twenty or thirty years from
now we'll have all the computational power that we can
possibly use, so the computer per se will. be no problem.
By that time we should also be extraordinarily proficient
in modeling. In particular, we should be able to model
in detail any segment of the world's political or economic
situation--in particular, the status of what we would now
call the enemy. Presumably he will be able to do like-
wise against us. I foresee the possibility that warfare
will be a series of political moves and countermoves backed
up not by exchange of military attacks as in traditional
warfare, but by the manipulation of an enemy's external
environment (e.g., the economy of other parts of the
world) or even his internal environment (e.g., his weather).
One might even argue that warfare could become an exchange
between your model of the enemy and his model of you.
If something such as I have suggested were to come
true or even partly true, the role of the Air Force would
certainly change. What might be its role in such a world?
Perhaps the Air Force would become a professional arm of
the government, as the Army Corps of Engineers presently
is the professional construction arm of the government.
Perhaps the Air Force would become the professional ex-
plorer or adventurer for the government--perhaps the
explorer of space. Perhaps the Air Force will continue
to fight a traditional warfare, but in outer space or on
another planet. Perhaps the Air Force will be the exe-
cutive agent to deliver energy to the atmosphere to achieve
weather control. Whatever, the role of the Air Force
will change.

I've suggested several possible impacts of the com-
puter. I've not detailed any of them, but I hope that
I have shown each to be credible. Many of my suggestions
have important sociological implications and challenges;
I have acknowledged but not explored the interaction be-
tween technological possibility and political, economic,
and social factors. My case for believing that these
events can come about rests on these points:
"o Computer hardware will continue to increase in
speed but reduce in cost and size.
"o The computer can process all kinds of symbolic
information.
"o The technique of modeling allows the computer
to simulate and experiment with all kinds of
systems and situations.
"o We are solving the programming problem.
Dr. William Baker of the Bell Telephone Laboratories
has characterized computer technology as a question; I
know of no answer to it:
What other technology is there in which the
United States has such a commanding lead,
which will have as much effect on how we
design and do things, which will be as per-
vasive, and which will both attract and
appeal so strongly to the young mind?
This brings us to the end of our tour through Com-
puterland, I hope that I have enlarged your image of
the computer or, as we ought to call it, the information
processor. It is the most powerful and most flexible
tool ever available to man and to society. It is not a
replacement for man in any large and encompassing sense;
it will displace him in many jobs, but it also will him many riew opportunities. The computer will touch
men everywhere and in every way, almost on a minute-to-
minute basis. Every man will communicate through a
computer whatever he does. It will change and reshape
his life, modify hiq career, and force him to accept a
life of continuous change.