Lecture notes for July 22

General Overview. Begin with Wilson site, question of the Unity of Science. Does knowing how the Brain functions help us understand how we think, how computers can be designed? What are some general principles? Principles of systems in general? What is a system - a set of interrelated parts. There are degrees of systemness. One major impact of the Internet is to make the world more systematic, more "connected." Isolation is being cut down, there may be homogenization in some ways, e.g., the English language, common sources of information. There may also be more diversity on other than geographical grounds, people with certain beliefs are able to link together, including racists, terrorists, as well as a tremendous number of innocuous groups of people who share an interest in whatever. This is done through WEB pages, chat groups, mailing lists which function like the one in our class. A question we are raising is whether there are general principles which can be applied to the evolution of the Internet, the evolution of the Computer/High Tech industrial sector, society at large when it uses these technologies.

Some general systems principles which Ben talked about:

Evolution and Autopoiesis

Evolutionary adaptation, first of all, seems to be ubiquitous in the world, perhaps because the basic logic of natural selection is a such a simple one. All one needs to have evolution, in the simplest case, is a collection of entities that can either survive or not, depending on circumstances; and that can, assuming they've been around a while, mutate or combine to form new entities. The environment that the entities are in, partly by its own direct action and partly by mediating interactions between the entities themselves, forces a "selection" between entities -- so that some entities are more likely than others to survive long enough to spawn new entities. Over time, this process leads to the production of a population of entities that are adapted to each other and to their environment.

  Autopoiesis is a strange-sounding word -- unless you're familiar with the biological systems theory literature, you probably haven't heard of it. But yet, it is no less essential a concept than evolution or attraction. It is something so basic that anyone who wants to understand the universe should turn it over and over in their heads, applying it to different situations and to their own lives, until it has become a part of their working daily mental vocabulary. Indeed, the very obscurity of the concept of autopoiesis tells us a lot about Western science, and the depth of its ingrained reductionist bias. For autopoiesis is nothing else but a precise formulation of the living nature of complex systems. What it refers to, specifically, is the ability of complex systems to produce themselves. Autopoiesis is closely related to the more familiar concept of "self-organization," but there is a difference: self-organizing systems need only rearrange themselves, while autopoietic systems must create themselves from materials at hand.

The paradigm example of autopoiesis is the biological organism -- the body! A body consists of a collection of interconnected parts precisely designed so as to be able to support and produce each other. Another example, just as apt, is the modern economy. No individual is self-sufficient, and only a small percentage are truly parasitic; directly or indirectly most individuals rely for their lifestyle on the actions of most other individuals. Yet another example, which I talked about a lot in my book Chaotic Logic, is the belief system. No one who has argued with a fundamentalist Christian will doubt the autopoiesis of belief systems. Every point in the fundamentalist's argument is backed up by at least fourteen other points in her argument!

Autopoiesis relates with adaptation, in an obvious way. Evolution serves to adapt systems, and thus to "improve" them in various ways, but in most real-world situations it only acts in the context of autopoiesis. Evolutionary adaptations which destroy autopoiesis will not lead to viable structures. On the other hand, evolutionary adaptations which lead to stronger, more stable autopoiesis will tend to lead to structures that survive. This is clearly what we have seen in the evolution of organisms.

Hierarchy and Hetararchy

As a first approximation, one may say that perception involves primarily the passing of information up the hierarchy, action involves primarily the passing of information down the hierarchy, and memory access involves primarily exploiting the associative links, i.e. the heterarchical network.

The brain is an evolving hierarchy/hetararchy of emerging systems. And so is everything else? Hierarchy and association, vertical and horizontal (although this is not really a matter of physical direction. Traditional organizational chart versus matrix organization. Task forces, soft money, parallel organization. Hetararchy in particular is facilitated by the Internet. However, hierarchy also seems to be emerging spontaneously in the system in the form of Microsoft corporation. The government is acting against this. Government might also provide some hierarchical, control functions.

One of the key consequences of the Internet is to equalize resources around the world, e.g., researchers in remote places have equal access to data. "Knowledge is Power." This spreads power. This may frighten people, who are we spreading it to?

Attractors and Archetypes

An attractor is, roughly speaking, a certain behavior of a system which gives the impression of having "magnetic power" over other behaviors of the system. Once the system is following the attracting behavior, it is guaranteed to keep on following that behavior until the end of time. And if the system is following some other behavior which is reasonably similar to the attracting behavior, it is bound to eventually wind up following the attracting behavior, or something very, very close to it.

Strictly speaking, only deterministic systems display attractors; real systems always have a random aspect and thus display only "probabilistic attractors" which have a small but definite chance of being escaped. But the basic idea is the same. In human affairs, for example, heroin addiction is a probabilistic attractor.

And so the complexity perspective on the world is, finally, a relatively simple one. The world is a network of processes, each one interacting with other processes. The focus is on dynamics, patterns of change. Processes interacting with other processes give rise to attractors -- characteristic patterns of behavior. And attractors themselves tend to fall into certain categories, having recurrent structural similarities which may be called archetypes -- archetypes like "interesting order around the edge of chaos", like hierarchy, like associativity.

Fixed point, limit cycle and "strange" attractors.

Ben, ch1: complex systems, example of ideological systems. Look at Tawana Brawley case - religious fundamentalism. Systems can be open or closed.

Ben ch4 - Artificial intelligence. Some are rule based, essentially hierarchical in that they were programmed with rigid rules. Others seek to learn through statistical patterns, or neural networks. Rule based AI is brittle, narrow. Associative AI is creative, irrational, poetic. We saw interesting examples in class. What Ben wants to do is make the associative ones more brain-like, with more free time, more evolution. Fancy techniques such as backpropagation, where the network weights different approaches (steepest descent vs. randomization) based on how well they worked in the past.

Ben Chapter 5. On the Mind itself. Is the mind the brain or something else?

Jung's theory of archetypes. These are "attractors" which he observed in many cultures.

The Collective Unconscious. Is this just BS or is it something inherent in a complex system? Would it arise in artificial intelligence, in the Internet? Jung's theory of numbers.

Charles S. Peirce: first, second, third

Hegel: thesis, synethsis, anthesis

Goertzel? Action, reaction, interaction

Are these archetypal patterns inherent in complex systems?

Ben Chapter 8 Turing machines and universal computer, logically the simplest computer can do anything.

The mathematician Kurt Godel, in the 1920's, proved the completeness of the propositional calculus: he proved that every statement in the propositional logic can be proved either true or false using the rules of propositional logic. Propositional logic understands itself. And then, in the 1930's, he did himself one better, and proved the opposite result for predicate calculus. In predicate logic, he showed, one can formulate perfectly reasonable statements that cannot be proved either true or false within predicate logic. Predicate logic does not understand itself. Like all interesting things in the universe, it exceeds itself by nature. This is the famous Godel's Incompleteness Theorem.

The essence of Godel's proof is actually very simple. Suppose we have some particular mathematical system -- some set of rules for drawing mathematical conclusions, such as the predicate calculus. Let's call this system Fred. Assume Fred is logically consistent, not containing any self-contradictions. Then, if Fred is at least as powerful as the ordinary predicate calculus, we can use the logical operations contained in Fred to construct a model of Fred within Fred. This was perhaps Godel's biggest insight -- that any reasonably powerful mathematical system is reflexive, is able to contain a model of itself. We can define mathematics within mathematics, so long as we begin with mathematics that contains counting numbers, AND, OR, NOT, FOR ALL, and THERE EXISTS.

Ben ch 9 - what is intelligence

Intelligence is the ability to behave appropriately under unpredictable conditions.

To make "unpredictable" precise, I drew on chaos theory. Intuitively, I noted, a system is unpredictable if a lot of information about its past state tends to yield only a little information about its future state. In ordinary dynamical systems theory, this kind of unpredictability is gauged by a number called the "Liapunov exponent," which measures how quickly information about the exact numerical state of a system in the past become obsolete in predicting the exact numerical state of the system in the future. However, prediction of exact numerical values is really not very interesting in the context of intelligent systems -- because, in the real world, exactitude is almost never necessary. Rather, intelligent systems are generally concerned with observing general emergent patterns in the world around them, and predicting the general emergent patterns that will arise in the future. What distinguishes intelligent systems, then, is the ability to operate in environments that are "unpredictable" in the sense of being highly unpredictable on the level of numerical details, yet moderately predictable on the level of emergent patterns. The word I like best for this kind of environment is, simply, "complex." Neither totally random nor precisely predictable -- predictable with attention to emergent pattern. Complex. Intelligence, then, is the ability to carry out appropriate behavior in complex environments.

Having thus clarified the notion of an "unpredictable environment," what remains to complete the definition of intelligence is a better understanding of "appropriate behavior." Appropriateness, by definition, is defined relative to some standard of behavior -- it is thus a special case of goal-fulfilling behavior or "optimization." An intelligent system, then, may be thought of as one that can achieve certain types of goals in complex environments.

IQ Tests are not good measures of intelligence, how about multiple-choice tests to measure learning? Probably a correlation, a "proxy measure" as it is called in economics, people who do well on tests also do well on better measures. Essay tests really better. You can have your choice.

We will look for these system characteristics in everything we look at, also look for other attractors or characteristic patterns.

Article on the History of the Internet:

1962. Linking computers, telnet. Packet switching. 1972 the first "hot" application, electronic mail. Open-architecture networking. No central, global control. Ethernet, not intended for PC's or LAN's.

Different protocols were adopted, etc. No need to memorize these.

BITNET and USENET designed for communities of scholars. Simple ASCII text sent. Communities and substructures grew. What we see here is an autopoeisis, an emerging structure. The WWW is a recent development, a part of the Internet. This is based on HTML, a common language which allows graphics, etc.

Vannevar Bush article Anticipates the need for the Personal Computer. We are being bogged down by a surplus of information. He saw photography and microfilm as the key technologies here. A forehead mounted camera rather than a scanner. Facsimile transmission. Talking directly to the recorder. Adding machines and calculators. We need a new, positional logic. Are new computer languages providing this? A central information system, e.g., for stock in a department store. Indexing systems, the human mind goes by associations, not by hierarchical links.

Consider a future device for individual

use, which is a sort of mechanized private

file and library. It needs a name, and, to

coin one at random, "memex" will do. A

memex is a device in which an individual

stores all his books, records, and

communications, and which is mechanized

so that it may be consulted with exceeding

speed and flexibility. It is an enlarged

intimate supplement to his memory.

Associative Indexing will be a key feature of this, anything can call anything else. Anything can be linked to anything else. Encyclopedias with meshes of associative links will exist. Note that the hetararchical principle is key here.

Interface Culture.

Preface: What is culture? Is technology part of culture? Yes. Yet we think of it as objective, as out there, which it is once we create it. McLuhan and Understanding Media began thinking about this. How did radio and television change things? Electronic speed erased the effects of distance. All art depends on a technology, but in the internet medium, artists must be engineers in a greater sense.

What are our goals in an STS course? To look at the links between levels of analysis. Some people think that this is the best way to do all kinds of analysis, e.g., Edmund O. Wilson believes that biological principles can be used to explain social problems. In particular, the theory of evolution seems to many people to be a theory which transcends subjects, e.g., evolution of species, evolution of social forms, evolution of ideas within the brain, evolution of computer programs, etc.

Chapter One: Bitmapping

In 1968, Doug Engelbart presented a new product: Equivalent to Ben Franklin or Alexander Graham Bell. It was a GUI. This involved direct manipulation of objects rather than typing in commands. Different from the days of DOS. Was this a major innovation? A Killer/AP such as the spreadsheet or word processor? The mouse. This is supposedly modeled on the way we think, in terms of spatial arrangements, where we left things. How do we keep track of things on our desks? Is the Windows desktop an equivalent of this? Why is it so important, e.g., to Microsoft as a company?

Vocabulary on page 24: surfing, webs, desktops, windows, dragging, etc. etc. When this first came out, a lot of IBM types thought it was for kids, e.g., with the early MAC.

What is an INTERFACE? The device which links between the human and the computer. The MEMEX designed by Vanevar Bush was an example of an interface design.

Suppose we go from the desktop to town squares, interfaces with planets, in orbit around stars, etc. These may change how we think about our data, about the world around us. Apple's 3D file management software, now called HotSauce.

The mouse. The chording keyboard, other innovations which may or may not be successful?

Changes in the Weltanschaung. Now on TV we have metashows, shows about tv rather than about life. Shows with alternative, e.g., The Door?

What these things have in common is the need for information filters, e.g., search engines. Personal filters, e.g., personal search on NYTIMES. All based on key words.

Chapter 2: The Desktop. Xerox Parc invented it, Steve Jobs saw the potential and marketed it. Marketing is the key here. It became playful, fun. Look at the help wizard in Microsoft Word. Microsoft domination vs. Apple? Vs Netscape?

Go to The Palace as an example

Chapter 3: windows. What does the windows metaphor mean? Are they really windows? Do we see through them? Actually, they are more like sheets or pages? But we think of pages within a document. We think we can learn better with a text on paper, it seems more real. How about magazines on the WEB? Slate magazine,

NY Times.

Go to Slate and Times on the WEB and demonstrate. How is it to read a newspaper this way, or a magazine. What are the advantages and disadvantages. How will this effect magazine publishing? Anyone can publish anything cheaply, but who will select it. This is what I want you to investigate. The key phenomenon is Switching Between Modes. Windows lead to a more fragmented, disconnected view of the world. E.g. Afghanistan regime banning TV. A scrolling Window.

Frame or sub-window, windows within windows. These are used for advertising. The mainstream is taking over the WEB. If you put someone else's stuff up as a link on your site it is ok, but if you put it in a FRAME it may be stealing.

Hyperlinks will stich the world together in a new way.

Chapter 4

Is "surfing" a good term to describe what we do on the Internet? Similar to channel surfing, surfing the waves? Are we "screenagers" or TV addicts? Most WEB pages, however, are approximating TV more, using graphics, animation, grainy video feeds instead of using hypertext, the unique aspect of the WEB. The LINK is the first significant new form of punctuation to emerge in centuries. They are analogous to the way the brain works with "links of association" as Dickens called them. Dickens tried to paint a whole society, with odd connections between people as you would find in a Tolstoy novel also. [We see this now in organizational structure with "task forces" and "matrix organization" instead of simple hierarchies.] The WEB viewer is not as passive as the TV viewer who is pure consumer. He/she can post his/her own page, incorporate things into boxes on his page, talk to people, send out messages, etc. Netscape "Navigator" uses a nautical metaphor, Windows "Explorer". Bush anticipated this with his "Memex" arguing that we are using old fashioned mechanisms of sorting through material. Bush talked about "threading" and "trails" rather than "links". Trailblazers was a term he used. A literary view of the world, more poetry than prose. Not like a librarian with endless shelves to file. But are libraries different with the WEB? Maybe not.

In Memex each person would blaze their own trail, WEB surfers may follow trails of others, but they can also create bookmarks or their own pages of links.

Example is Hypertext Fiction, go to presentation.

Chapter 5 - text. Writing with a pen, typewriter, word processor. Why does one feel natural? Does writing on a word processor change how we write? My own pre-computer books read about the same, I think. It is easier on a processor. But I was always a typist. New technologies often used for unanticipated purposes. Being digital means being able to reinvent yourself, morph (147). Are text commands archaic, will we go primarily to icons? Or is text reinforced by the WEB? Counting text, establishing authorship. Can computers read a file and write a summary? Why not? Would a "mind" be required? Apple is working on this, as is Ben.

Chapter 6 and conclusions to be left for the final. Core reading assignments are as follows:

From the World Wide Brain: chapters 1, 4,5,8,9

From Interface Culture: Preface and Chapters 1 to 5 - chapter 6 and the conclusion will be left for the final.

Two readings listed on the syllabus:

History of the Internet

Vannevar Bush

One article from the Times on distributed computing on July 15 (on the WEB page).