Notes for John G. Kemeny Man and the Computer

Key concepts: batch processing, executive system, frame, machine language, model, randomness, simulation, social analyst, symbiosis, symbiote, symbiotic evolution, time sharing.


Related theorists: Vannevar Bush, Jay Forrester, Hafner, Hayles, Janz, Licklider, Lyon, von Neumann, Ong, Ramsay, Herbert Simon.

Preface

Part One
The New Symbiosis

1
A New Species Is Born

Definition of symbiosis as two different organisms living in intimate, beneficial union; surprising to think he was not familiar with Licklider using this term as well, instead quoting Wells and Huxley.

(3) H. G. Wells and Julian Huxley in their book The Science of Life define symbiosis as “two organisms of different kinds living in intimate union and to the benefit of both.”

A Brief History
(5) VON NEUMANN PROPOSALS
1. Fully Electronic Computers
2. Binary Number System
3. Internal Memory
4. Stored Program
5. Universal Computer

Rupture with prior static symbols and semiotic systems because of machine unattended internal memory stepwise operation leading to programming, although it may become understood as another natural language, if not already assumed for digital natives (thus computer based testing in public education).

Review of Von Neumann proposals: fully electronic, binary number system, internal memory, stored program, universal computer.

(6) Von Neumann proposed that once should be able to store a set of instructions within the internal memory of the machine so that the computer could go from step to step by consulting its own memory without waiting for human interference. Such a set of instructions is now known as a “program,” and the ability to program computers has been the single major breakthrough that differentiates a modern computer from an old-fashioned business machine.
(6) Of course all the electronic components that von Neumann proposed some twenty-five years ago are now hopelessly out of date, but even the most complex modern machine is based on the principles that he outlined at that time. He was a prophet in predicting the impact of modern computers, but even he underestimated the rapid growth of electronic technology and therefore failed to anticipate the incredible increase in computing power and the impact that the computer would have within a generation.

Are Computers a Species?

Responds to biochauvanism that also emerges in Ong to answer positively and hinting at rupture in humanities when significance of intentionally leveraging symbiosis while studying it critically; this can be done while programming and everyday working code.

(10) I would like to argue that the traditional distinction between living and inanimate matter may be important to a biologist but is unimportant and possibly dangerously misleading for philosophical considerations.


2
What Each Species Contributes

The Computer
(14) The single most important feature of a modern computer is its incredible speed.
(14) Its second most important feature is its incredible memory.

Important differentiating features of computers are speed of operations and size of addressable memory; reliability assumed although earlier forms were so unreliable as to preclude increasing speed or memory extent; speed, addressability and ultimate capacity are contours of computer species alien phenomenology.

(15) Reliability of electronic components has improved even more rapidly than the size and speed of machines, so that the probability of a single error in an entire day of operations is much smaller today than it was fifteen years ago.

Man

Would relieve criticism if a future version of an otherwise exemplary work dealt with putative but hopefully harmless male bias.

(17) One can't help reaching the conclusion that it is more efficient to use a human being as the computer's partner than to spend many years trying to teach the computer a talent for which it is not well suited.
(17) We seem to be able to do it through mysterious processes of association that no one has duplicated on a computer.
(18) Men also have mysterious talents which are vaguely described as intuition and creativity.

Familiar differentiation of human tasks and machine tasks misses opportunity of computers to monitor themselves noted by Hafner and Lyon.

(18) Man should decide the best use of computers. Man should set the goals and tell the computer how to work toward them. It is best for man to monitor the work of the computer so that he may use his powers of intuition and evaluation to guide it in its work.

Working Together


3
Time Sharing
A New Relationship

Although not referencing Licklider worth checking whether Licklider believed or was cognizant of the potential of time sharing to radically transform the human computer symbiosis as he conceived it, by greatly impacting communication between humans and machines.

(21) During the 1960s a fundamental change occurred in the relationship of man to computer. This new relationship is known technically as “time sharing.” . . . It is only through this new development that a true symbiotic relationship between man and computers is possible.

Curious that inefficiency of batch processing from human point of view mirrored in relationship between teachers and students, relationships between the same species.

(22) The shortcoming of batch processing is the simple fact that most computer runs do not result in the solution to a problem, but in the detection of errors or in the printing of completely wrong answers. . . . However, if each such trial run delayed the user by twenty-four hours, it typically took two or three weeks for him to produce a correct program. . . . While batch processing is completely efficient from the point of view of keeping the machine busy at all times, it is most inefficient from the human point of view.

Perceived affordance of having individual access to a computer recapitulates benefits of private reading.

(23) What he would like is to have a computer of his own, which he can use privately at his own convenience, correcting his ten to twenty errors in one session and not quitting until he has obtained his results.
(23) In a time-sharing system a hundred individuals all use the same computer at the same time, in complete privacy, and enjoy the illusion that the computer's sole purpose is to help them. The goal is accomplished by using the tremendous differential in speed between the computer and man to allow the computer to process “simultaneously” a hundred different requests. The result is great efficiency in computer use, with a vastly different effect on human beings. An accidental but all important by-product of time sharing is man-machine communication.

The Dartmouth Time-Sharing System

Compare this discussion of the workings of an actual computer to instrument of Burks, Goldstine and von Neumann or ARPANET by Hafner and Lyon: both are more concrete then Turing beyond the interlocutor obfuscating interface; the DTSS diagram shows many user terminals sharing a communications computer, which alone (rather than each individual user, as in the IMP network design) connects to the central processor and high-speed memory, whereas input and output peripherals and bulk memory fill out the Burks, Goldstine and von Neumann model.

(24) The terminal is linked to the computation center through an ordinary telephone line and therefore its location is irrelevant.

Round robin scheduler at heart of time sharing.

(27) A given central processor can work only one problem at a time. However, it is programmed to work in a simple round robin, assigning to each active user a fraction of a second of computing time.

Affordance now taken for granted of having third party applications ready at hand so they do not have to be programmed by the user may have covered over entry to direct participation in formation of problem solving space to participation in selection of arranged options where the control programs are not directly modified by any of the library users.

(28) Most of the bulk memory, however, is available for the storage of user programs. . . . In addition, the bulk memory has available a large collection of “library programs” which any user may call upon to solve a variety of standard problems. . . . This brings a variety of mathematical techniques within the grasp of a user who would not have been able to write these programs himself.

A Language for Communication

Kemeny ranks learnability of BASIC over FORTRAN over machine language; natural machine language is the overly predetermined but more importantly unthinkable for which Ong resists, and would likely dismiss FORTRAN as well, but higher level languages like C++ may be on par with other second languages.

(29) Built into computers is a language known as “machine language.” It is a natural language for an electronic computer since each instruction corresponds to simple electronic circuitry, but it is most unnatural for a human being to learn.

On another argument is how programming languages compare to learned Latin.

FORTRAN design to be easy to learn for those familiar with English.

(29) FORTRAN combines words from ordinary English with simple mathematical symbols in a language which, although artificial, is so designed that it can be easily learned by a human being.

Machine language for few, FORTRAN for many, BASIC for everyone college educated now via user interfaces removes requirement to learn how to program in order to usefully use (the reason Turkle shifted from studying learning programming by the small population that did so to general use by orders of magnitude larger populations), rendering programming competence no longer a required component of intelligent human being.

(30) While the availability of FORTRAN extended computer usage from a handful of experts to thousands of scientific users, we at Dartmouth envisaged the possibility of millions of people writing their own computer programs. Therefore, we decided to design a new computer language that would be accessible to typical college students. This is how the language called BASIC was created. Profiting from years of experience with FORTRAN, we designed a new language that was particularly easy for the layman to learn and that facilitated communication between man and machine.

Boast that Dartmouth freshmen can begin programming after listening to a few lectures and reading a short manual.

(30) Dartmouth freshmen listen to two one-hour lectures and then read a short manual. Before the end of the first week of the course, the typical student is able to write at least one usable program.
(30-31) Each student, during a ten-week term, must write four text programs entirely on his own and work on them until they are errorless. . . . Therefore we have demonstrated that with a language as simple as BASIC computer programming is not beyond the competence of any intelligent human being.

BASIC first language designed with affordances of time-sharing in mind.

(31) The secret of BASIC is the fact that it is written on several levels.
(31-32) Since BASIC is the first language to have been written since the invention of time sharing, a means for direct communication between computer and human being was written into the language. . . . These “interactive programs” may in the long run prove to be the greatest asset of time sharing.

Privacy to make mistakes key to learning, which is why so many learned at home as my prior interview data suggests: imagine contests like achieving results of ten week university term or initial week training session applied to language of your choice.

(32) This psychological factor is of overwhelming importance. While many students could have mastered computers even without this guaranteed privacy, most faculty members would have refused to expose themselves to the embarrassment of publicly committing hundreds of mistakes before they became experts.

Some Typical Uses of Time Sharing

Amusing to think where this measure would be expressed now that the capitalist market pervades everyday computer use.

(33) It is not surprising that the departments of mathematics, natural sciences, and engineering are all heavy users of the computation center. More unexpectedly, however, the heaviest users are actually the students in business administration and some of the social sciences.

Freedom zero to run software for any reason in the program library environment considered like human library with further freedoms due to nondestructive nature, up to the point that too many resources are being used.

(33) Just as any student may go in and browse the library, or check out any book he wishes without asking for permission or explaining why he wants that particular book, he may use the computation center without asking permission or explaining why he is running a particular program.

Opportunities for big humanities as faculty quickly adapted to using computers by outsourcing implementation to underlings.

(34) A large number of undergraduate and graduate students have lucrative computer-programming assistantships in which they help faculty members on research projects.
(34) The computer library contains a wide variety of games, most of them written by students.

Fear of embarrassment of not knowing how to use absent digital natives, who play lots of locally modified games available in the library, a sort of local culture surrounding DTSS rather than discrete physical human communication spaces and places.

(35) But, more importantly, for many inexperienced users the opportunity of playing games against a computer is a major factor in removing psychological blocks that frighten the average human being away from free use of machines.

Networked human team with single computer only a dream; today realized with networks of humans and computers in LANs and global Internet.

(37) A number of educational uses of this multiteletype hookup, such as business games and small group experiments, have been considered, but so far we have only the vaguest impressions as to the full power of a team consisting of several human beings and a computer.

A Technical Description of DTSS

Scale of resources available for general use on DTSS dwarfs anything proposed by von Neumann, although even its multiteletype hookup today exceeded many orders of magnitude by Internet capabilities, yet still abysally short of UTM: the materiality of code is expansive.

(41) Nearly 20 million words are reserved for copies of all systems programs, for the library programs, for the user catalogs, and for large data bases. That leaves more than 40 million words for user programs.

Ambiguous ordering of executive system as it seems from a real time, time sharing perspective editing, listing and saving as user commands beyond ordinary execution would be more important: perhaps the other direction makes sense only in terms of gross CPU time spent doing everything; then again, it matters how many programs were typically executing, for users spent most of their time editing them; they were not constantly running along with them they way things do now in the distributed application world with which we communicate.

Executive system like supervisory control of pinball game.

(42) The executive system carried out about one command per second, most of which fell into the following categories, given in order of importance:
1. Execute the program
2. Create a new program or retrieve an old one
3. Save a program
4. Give me a listing of my program
5. Perform an editorial function.

Evaluation

The ubiquitous mobile device fulfills prediction, although what is not stated is that nontrivial cost ranges will exist nonetheless, in part because his prediction that government would spend more on programming education turned out to be wrong.

(42) But anyone can acquire a private terminal in his home, and an hour of terminal time a day, at the cost of maintaining a luxury car.
Within the next two decades
the price will undoubtedly come down to a level which will make computer terminals in the home quite common.

That we are collectively not bothered that programming skill may devolve from anticipated height in 1980s to mere use competency evidence in current absence of widespread programming instruction; instead, the deep thinker following how software works in order to use it is replaced by the manipulation of complex user interfaces distributed among countless other software systems.

(42-43) More importantly, the use of computers has been made so simple that acquiring programming skill is no harder than learning how to use a large library.


4
Division of Labor

A Computer's Eye View

Human struggle to grasp machine perspective, which Bogost calls alien phenomenology.

Procedural rhetoric example of Ramsay pataphysic production of imperative narrative describing a means of thinking about machine operations, which is also a machine way of thinking about humans, by considering how humans think about machines.

(46-47) If the reader will now substitute computers for human beings, human beings for the more intelligent creatures and reduce the time scale by a million, he will understand the computers' point of view about time sharing.


The Problem of Communication
(49) The temptation, of course, is to try to teach the computer how to evaluate its own results more intelligently.

Guidelines


5
Symbiote or Parasite?
The Popular Reaction to Computers
(56) Computers have lowered the price tag for Big Brother but have not brought about a change in principle. Questions such as whether we have the right to see what information governments or finance companies have on us, and whether we have a right to correct erroneous information, must be answered. But this is a problem to be solved by legal experts, not by computer experts.

The Deeper Problems

Makes explicit philosophical pronouncement that we can question or put on the back burner to move forward with working code rather than social critique, is socialism required to achieve utopia where most people program, also being subsumed by floss (this will toward more government spending to maximize extent of everyone working code).

(57) I would like to see a fundamental change in philosophy on the part of both government and business. Both should be willing to spend more to make life better and easier for everyone. Indeed, this will be a major theme throughout the remainder of this book. Only when such a philosophy is adopted generally can we look forward to a time when the average human being will look at the computer as a friend rather than a foe.

Critical programming potential in human beings acquainted with powers and limitations in creative design use and shared visual and audio functions.

(58-59) Until we can bring up a new generation of human beings who are thoroughly acquainted with the power and limitations of computers, who know what questions have to be asked and answered, and who are not intimidated by computer experts in a debate, we cannot hope for a fundamental change. I see great promise in the reactions of recent Dartmouth students. Now that most of them have first-hand experience with computers, they approach computer applications without fear or superstition and with considerable understanding of how computers can serve mankind.


Part Two
Symbiotic Evolution

6
A Look at the Future
More Power to the Computer

Interesting design constraint differentiating phenomena that can occur in the heart of the CPU versus at a distance, based on the real time requirements of the nanosecond process cycle.

(64) Light travels one foot in a nanosecond. Therefore a computer that is to have a basic cycle time of one nanosecond must be able to send the signal from any point in the machine to any other point without traveling more than a foot.

Today this is the basic home computer whose price is at an all time low, setting up multiplied dissemination for beyond college students envisioned by Kemeny half a century ago.

(66) I see absolutely no reason why a very reliable computer terminal could not be manufactured to see for the price of a black-and-white television set. This will be necessary if computers are to be brought into the home.

Computers and Communications
(67) The next decade is likely to see the development of huge computer networks. . . . The full impact of modern computers will be felt by most people only when large multiprocessor computer centers are built all over the United Stages and tied efficiently into the existing communications networks.

Building a Network

Prescient of present Internet although envisaged as low cost terminals fed by distant networks; wait, that is what happened.

(68) It would be desirable to design the network in such a way that most users could reach it through a local telephone call.

Evolution
(71) But the multicomputer systems and the computer networks are likely to be the most important evolutionary developments.

John Kemeny, inventor of BASIC programming language, shared the optimism of Licklider that a great future of continuous improvement was in store for both machines and humans; ironically, the outcome is human devolution and machine evolution, unless we change course.

(71) We are witnessing even now the evolution of a species in which the individual is subsumed under a group consciousness. Indeed it is a telepathic race. And I expect that computer networks will display all the marvelous traits that science fiction predicted for such strange beings.


7
Computers in Education
First Attempts

Shifted expectation that instructors would develop their own programs for teaching, perhaps echoing past assumption they would publish their own textbooks.

(74) Given such a large time-sharing system CAI [Computer-Aided Instruction] will be as good as the instructor's program can make it.

Shortcomings of CAI

That we have swallowed this dual assumption today points to the subsumption of human intellect into collective consciousness entangled with the machines: recall prior arguments about material specific advantages of spiral bound manuals and other forms of programming instruction noted by critical code studies theorists (Montfort et al).

(74) These are, first, that the computer is a very expensive substitute for a book, and, second, that it is a very poor substitute for a teacher.

An Alternative Approach

Distinguishing programming and computing, valorize programming, teaching the computer to think, which iterates as learning through teaching the computer, over mere computing, such as the rote activity of solving mathematical problems handed over to machines in programs; however, a short circuit occurs between isomorphic intentionalities in which the ease of use of one forecloses requisite knowledge acquisition to opportunity cost calculations over the other: the dumb user trying to input mathematical equations into the latest touch or speech recognition interface forgets the purpose of doing the problem in a way one who was forced to write a program for the computer to solve it would not likely be ignorant, although maybe getting lost in the details of programming languages while ignoring the goal mathematics aimed to be taught.

Learning by programming versus learning through solely human team activities (encompassing students working together and teacher student relations) depends on unique ability of computers to execute code at very high cycle rates for a comparative eternity for human cognition; at this level is distinguishable from other forms of computer-aided instruction on account of the being forced to teach the computer aspect, and is degraded surface interface interaction satisfaction; the argument needs to be made why programming is part of optimal symbiosis (comportment): despite accurate predictions about the emergence of the global Internet Kemeny was overly optimistic that programming would remain sufficiently valorized once sound, intuitive, user friendly interfaces arose in massive software projects like Microsoft Windows, Apple OS, and GNU/Linux floss.

At the other extreme from taught helplessness, anticipating terms elucidated by Ian Bogost, is taught procedural literacy, as Kemeny argues that learning through teaching the computer exemplifies symbiotic transformation; it is, of course, necessary to keep the momentum going so that programming skill becomes like handwriting, cooking, general home economics.

(79) The students learn an enormous amount by being forced to teach the computer how to solve a given problem. . . . The student must concentrate on the basic principles; he must understand the algorithm thoroughly in order to be able to explain it to a computer. On the other hand, he does not have to do any of the arithmetic or algebra. At Dartmouth we have seen hundreds of examples of spectacular success of learning through teaching the computer.

What is Needed

Interesting comparison between passive relationship for learning by phonograph and CAI that does not demand reflexive knowledge of programming; elevate to becoming stupid having lost essential practice of teaching the computer to think to solve problems formerly worked by human computers.

Danger in simplification of interface so procedural rhetoric no longer learned in the process.

(80) Most students leave Dartmouth with a thorough understanding of the nature of modern computers and with a good idea as to how they may be used in later life. Since in CAI the student plays a rather passive role, somewhat like learning a language from a phonograph record, none of these benefits accrue.

Continuing Education
(92) I see three major components in education in the home: (1) lectures over TV; (2) books; and (3) student interaction with the home computer terminal.


8
Library of the Future

A National Automated Reference Library

Instead of this library, which includes a fee, we got the commercial Internet: is this an aspect of how we have unintentionally subverted better intentions for the human computer symbiosis, as here they problems of storage, search, transmission; also consider Janz on problems using search for philosophical questions.

(87) Once information is stored in machine-readable form, and a substantial time-sharing system is made part of the automated library, the entire problem of searching for relevant information takes on a new dimension. For the first time there would be hope that the scholar who is interested in the available knowledge on a specialized topic could systematically search all the available literature and find the items that are useful for his work.

The Problem of Storage
(89) For example, one technique, known as “ultra-microfiche,” can store a volume of up to 1000 pages on a single 3-by-5 card.

Continuous film photographic storage like Bush Memex that will likely be digitized foreshadowing media convergence; at least store abstracts (metadata) in machine-readable forms in high speed memory.

(90) If the miniaturized images were stored on a medium that is physically similar to the magnetic tape used with modern computers, the 1000 volumes could then be stored on approximately 150 feet of tape, which can be positioned by existing devices to any one of the thousand volumes in just a few seconds.

The Problem of Search
(91) While I visualize the books and journals stored as photographic images, I would like to see the abstracts stored in machine-readable form within the computer's memory.

Instead of a conversational partner we got big business advertisement driven search.

(92) Considerable research will have to be devoted to the question of how such a conversational mode program can be written and how one can impart enough intelligence to the computer to enable it to be a truly efficient partner in search.

The Problem of Transmission
(93) There is no doubt that today's technology makes such photographic transmission possible. The only question is whether sufficient bandwidth can be supplied to service thousands of users simultaneously.

On demand printing exactly what has happened though many more storage locations of copies due to abundance of secondary storage and bandwidth: recall importance of designing network to accommodate new transmission patterns unpredictably arising with digital technologies but originally noticed to be bursty, high download small upload quantities.

(94-95) Instead of publishing books and articles in large editions, most copies of which are never read, one copy of an article could be filed in the national library (or one in each branch library) and additional copies printed when somebody actually expressed an interest in reading the article.

Evaluation
(95-96) The establishment of a national automated research library would add an important new dimension to the man-machine symbiosis.

Explosion of jobs for editors did not occur as predicted; instead, amateur, ad hoc content arrangement and absence of consciously crafted metadata is part of why we are getting stupider.

(96) Indeed, in a way the demand for editorial services will increase, since no item could be inserted in the automated library without a carefully written abstract and careful cross-indexing for the purposes of subject-matter search.

Better measure of readership feedback for authors.

(97) Since there would be a user fee imposed for each transmission, part of the fee could be turned back to the author in the form of royalties. (This system would be similar to that of paying royalties for phonograph records used in broadcasting.) A not insignificant benefit from this procedure would be the fact that authors would get an accurate picture of how much their work is read.

A Word of Caution

Fear of dangers inherent in single federal national library today more likely actualized by corporate codes (Lessig).

(98) Without proper safeguards, a national reference library operated by the federal government could become a dangerous weapon for the suppression of “undesirable” knowledge.


9
Computers as Management Tools
Present Use
(99) The most common applications of computers in business and industry today fall into a category that might be described as bookkeeping operations.
(101) Such faulty management decisions accompany all technological innovations and cannot be blamed on the new technology.

Management Information System
(102) While the examples I have cited illustrate the great usefulness of computers for bookkeeping functions, so far the machines have had very little impact on the planning and decision-making functions of management.

Disadvantage of costly reprogramming batch processing systems because they were not designed and originally programmed under new programming styles emerging with time-sharing systems, anticipating uses of computers performed by popular applications with diminished programming requirements nearing conversational or button pressing ease of user interfaces, focusing on symbiosis rather than default system perspective or programmer convenience (Norman DOET).

(103) But they were not designed for research purposes, and one of the greatest disadvantages of a batch-processing system is the fact that reprogramming it is so time-consuming and costly.

The discovery of this fact of new affordances advantaging time-sharing over batch, and the free open source option over proprietary, changes digital humanities research along with rest of built environment and humans multiple times.

Suggestion of writing programs to carry out research project initiates new kind of scientific practice that permeates humanities.

(104) Once the data are in convenient form, even if none of the existing programs will do the job, in a time-sharing system it is not difficult to write a new program to carry out a particular research project. . . . This conversion process will change student records from a bookkeeping system into a management information system.
(105) [Herbert]
Simon's thesis is that a good information system should provide us, not with as much information as possible, but with the least information that servers our need.
(105) The purpose of a well-designed management information system is not to provide a great volume of information. The job of the computer is to store this great amount of information and to provide summaries to management as they are requested or when the computer spots certain danger signals that the management has asked to have monitored. . . . The computer should also be able to provide summary information in any form requested, not simply in the form that some computer programmer thought would be convenient.

Collaboration between manager and programmer and desire for flexible design avoiding obsolescence and inviting future extension.

(107) This requires that the management explain to the system designer the kind of information desired and the form in which it needs to be supplied. It also requires that the designer have sufficient imagination not to be bound entirely by the needs of the present day but to devise a system that is flexible enough for future needs.

Models

Modeling by programming.

Model as theoretic description of how a phenomenon like a company or university functions using sets of formulas or computer program code (Ramsay declarative and imperative).

(108) What top management needs for long-range planning and effective decision-making is a model of the operation of the company. By a “model” I mean a theoretical description of how the company functions. This may consist of a set of formulas, or it could be in the form of a computer program.
(110) The difficulty in constructing such a model is not a shortcoming of computers, or the problem of writing a sufficiently sophisticated program, but our lack of understanding of how an institution operates.

Origins of data-driven organizational modeling replacing intuitive misconceptions (Forrester); symbiote optimal when computers provide summary information for humans to make value judgments.

(110-111) [Jay] Forrester has demonstrated conclusively that intuition is a very poor substitute for a thorough understanding of the operation of a complex social system.
(111) Therefore the value judgments must be left up to human beings, but in the future these can be made after knowing all the relevant facts and all the consequences, both short range and long range, or a proposed course of action.


10
Computers in the Home
Is It Possible?
(112) In an address at the dedication of the Kiewit Computation Center in 1965, I predicted that computer terminals in the home will be widespread by 1990.
(112) The easiest problem to solve is that of the computer memory.
(113) The communication problem presented by 30,000 simultaneously users is much more serious.

Anticipates the symbiote will mostly constitute communication, which requires less processing capacity than raw computation, although network protocols replace assumptions of slow character rate transmissions of readable text like personalized newspapers.

(114-115) The vast majority of users on DTSS will use 4 seconds or less of computing time in a 15-to-20-minute session. Even that is too high an average figure, since once there are millions of customers, the applications are likely to be heavily oriented toward communication rather than computation.
(115) But can the cost be brought down to a level at which the average home can afford to have a terminal?
(116) I therefore visualize nine regional centers initially, each with twenty processors, serving a total of three million customers.
(116) There is certainly an advantage in having more than one, since competition can improve the quality of service. On the other hand, having as many as ten national networks would confuse the average user and make each network less useful to him.

A Personalized Newspaper

Combination cultural and technical convention dividing television screen newspaper into frames becomes new basis for writing as example of intertwined technogenesis and synaptogenesis induced from habitual use (Hayles).

(119) Many terminals now have a display screen similar to a television screen. These can show roughly five paragraphs, or about one-third of a column of newsprint. Let us call such a unit a “frame.” Reporters would write their stories in frames.
(119) I estimate that all the news in the daily New York Times could be contained in five hundred frames.
(120) I would estimate that they typical reader would be interested in some twenty news stories a day and would want to look at one, two, or three frames for each one. In half an hour he would conclude a highly enjoyable and profitable personalized session with
The New York Times.

Predicted roles of advertisements and network providers half correct; instead of separate networks like television networks, global Internet more like highway and telephone systems.

(121) The computer network would charge the user for various services and could pay a royalty to the Times for each access by a user. . . . If that did not suffice, the newspaper retrieval program could be so written that between frames it presented ads. However, I would then hope that by paying an extra fee I would have the option of eliminating all advertisements.

Other Uses
(122) My sister says that she would love to see home terminals provide easy information on what is available in various stores.
(122) I am convinced that for items exceeding one dollar in price cash transactions will totally disappear. The multiplicity of credit cards is already a step in that direction.
(123) I more practical system would be for each person to receive a credit card from his own bank.
(124) Let me speculate as to how each member of the family might use the computer terminal. Father, if he brings his work home from the office, can use the terminal in place of a sizable office staff.

Stereotypical middle class family roles maintained and reinforeced while transformed by home terminals; suggests male and female roles may reverse but does not elaborate on how or why.

(124-125) Mother can do most of her shopping through a computer terminal. . . . If by 1990 the roles of man and woman have been completely reversed, the computer terminal will be equally happy to work out business problems for mother and to help father with his shopping and housework.

Predicts Turkle alone together and passive recreation; enforces orderly, lawful social activities of docile bodies.

(125) Children will find the home terminal an immeasurable asset in doing homework. . . . After he or she completes all homework assignments, the computer terminal can serve as a major resource for recreation. Not only will the computer play a wide variety of games with the user but it can monitor multiperson games with each player sitting in his own home.

Acknowledges inability to predict future uses.

(126) My list is likely to be deficient precisely because all these suggestions are already within the realm of possibility. I have made no provision for a multiplicity of new applications that will become reality within the next generation.


11
Solving the Problems of Society

Concludes with ways symbiote might improve quality of life.

(127) I should like to conclude with a consideration of some ways in which the symbiote might be used to improve the quality of human life.

Information Systems
(127) We are all terribly sensitive about possible invasions of privacy.
(128-129) I can envision a time-sharing system under which employers could promptly list avilable job opportunities and remove jobs that have been filled. Applicants for unemployment could pay periodic visits to one convenient nearby office where a clerk could type in the applicant's qualifications and his preferences as to type and location of job.
(129) There are additional benefits to be derived from a reasonable information system for welfare payments or for job matching.
(129) The one area in which a substantial amount of cooperation has taken place is law enforcement. . . . The California system for providing instantaneous information on the possibly dangerous occupants of a suspicious car spotted by a patrol car is an outstanding example of a real-time information system.

Simulation
(130) Computers can serve as laboratories for the social sciences through a technique known as “simulation.” This technique was developed in universities and has received important applications in military planning but has had little or no application in civilian government agencies.

Nascent realization of usefulness of simulation, itself a new research method, for social problems.

(131) Social problems, however, do not lend themselves either to laboratory models or to treatment by analog computers.

Controllable randomness becomes rational concept for simulation, transitioning from purely negative connotations in 10 PRINT.

(132) A great advantage of a computer simulation model is the fact that chance events can be built into it. . . . The computer model will then produce such accidents with the right frequency but at totally unpredictable random moments, just the way they actually occur.
(133-134) Some excellent examples of the use of computer models for social planning are found in the work of Jay Forrester. . . . Such problems are ideal illustrations for the main theme of the book—that man working in partnership with a computer can achieve vastly more than either can achieve on his own.

Social analyst bridges specific disciplines and technical knowledge for research design, echoed in digital humanities; compare to McGann poiesis as theory and Applen McDaniel theorist practitioner.

(134) I see the need for the development of a new type of professional, one who might perhaps be called a social analyst.

Control Systems

Examples of computerized simulation and control tasks impossible for humans to accomplish alone is strongest argument for fostering symbiosis.

(136) Due to certain unpredictable forces in such a chain reaction, the control must be done with split-second accuracy in order to sustain the chain reaction but prevent a nuclear explosion. No human being could possibly carry out this task, but a human-designed computer system handles it with great efficiency. Similarly, the control of traffic is beyond the capabilities of even a large number of police since they could never evaluate the total pattern fast enough to take effective action.

A Proposal
(137) I proposed the establishment of a national computer development agency, a federally subsidized private agency that would develop computer systems for public use. . . . I recommend the idea because while the actual operation of the computer systems should be reasonably economical, the development cost may be several times the annual operating cost.

Proposed national development agency aiming for portable, reusable solutions: what happened is the story of modern technological society; compare to development of FLOSS.

(138) The federal agency could insist that all systems developed under its auspices would be usable on the hardware of any computer manufacturer that is willing to meet a few federally set standards. The program could be written in one of several generally accepted computer languages and use simple conventions to tie in with computer memories and control devices.

Effect on the City
(138-139) Therefore I conclude that the real source of the problem is not the increasing density of population in the cities but the vast daily movement of people in and out of the city.
(139) Cities are now primarily centers of trade. But even the meaning of the phrase “trade center” has changed dramatically.

Compare this theoretical vision of an information society to Castells.

(140) It is my thesis that the traditional role of cities as centers of manufacture and trade has changed to a role in which the primary purpose of cities is as a center for the collection and exchange of information and the carrying out of paper transactions. Perhaps the most dramatic example is the stock market.

Failed prediction that ubiquitous use of videophones and transformation of employment patterns.

(141) I predict that a dramatic effect on the pattern of employment and location of offices will come about by the widespread use of video telephones. . . . We could still see each other's facial expressions, establish rapport, and share written documents or pictures, without having to leave our homes or offices.
(141) A second major reason for going to the office is the fact that that is where the files are.

Failed prediction that human assistants would continue to serve knowledge workers due to ease of use of technological systems and migration of duties.

(141-142) That leaves only the last consideration, that of being where one's secretary and assistants are. . . . But I will still find it more pleasant and more efficient to give my request to a human secretary who knows the peculiarities of the system rather than to battle with it myself.
(142) As long as that office is equipped with several video-phones and terminals giving access to one or more national computer networks, its exact location is irrelevant.
(143) It is my conviction that if the need for millions of people to rush in and out of the city every working day is removed, we would be well on our way to a solution of urban problems. What exactly would be the role of the central city? Perhaps it will become truly a center for information where the machines are located but not the human beings who use them.

Symbiotic Evolution

Compare symbiotic evolution to Hayles.

Stoked by the success of the Dartmouth implementation of BASIC programming as a core student competency, John Kemeny, who invented the language in the late 1960s, envisioned symbiotic evolution as the hoped for trajectory of human and machine species; he reiterates at the educational level the enthusiasm Herbert Simon held for anticipated overall social and economic improvements.

(144) Given the rate of human reproduction, a century is much too short a period for the usual forces of evolution and natural selection to bring about a significant change. Our best hope therefore lies in a new kind of evolutionary process which I have called “symbiotic evolution.”
(144) The existence of computer-communication networks will enable human beings at widely separated locations to function as a team. The vast capabilities of computer memories will enable use to make effective use of the explosion of human information and knowledge.
(144-145) However, this evolutionary development is only possible if man is willing to make drastic changes in his life style and in his conception of his own goals. . . . Since it is unlikely that any educational system can provide a training that will see us through a lifetime, we may have to devise a system in which learning continues throughout one's productive life.

Part rhetorical, cautionary tale that seems to have occurred in part as a result of technological progress meant to prevent it, start here and develop historical and theoretical narratives to explain why the symbiote has reached its current evolutionary state that seems worse, instead of better than Kemeny enthusiastically predicts, for not faithfully following the project he envisions.

(145-146) The best-intentioned people, if they lack the technical expertise and the tools to achieve our goals, can make the situation worse instead of better. Therefore we must look to the coming of a new man-computer partnership to provide the means which, combined with sufficient concern by men for their fellowmen and for future generations, can hopefully bring about a new golden age for mankind.



Kemeny, John G. Man and the Computer. New York: Charles Scribner's Sons, 1972. Print.