Class today will be a special Review Session aimed at students who scored less than 70% on the midterm. Those who attend will receive special extra credit points which will be useful ONLY if they would otherwise receive less than a C in the class. NOTE: this does NOT mean you are guaranteed a "C" if you show up for the review session, it just means you will get some extra credit points which will help if you are close to the line between a C and a D.
Of course, anyone may attend this special Review Session; attendance is not required but may be used to make up a missed class if you miss more than two.
You should use the lecture outlines from previous classes as a guide in your reviewing. Here is an outline of the topics I plan to cover:
A good place to begin is the distinction between determinist and complex systems science, as explained in the What is Chaos? slide presentation (Oct 1 notes). This class has focused on applications of the complex systems approach.
This coincides socially to the distinction between "industrial" and "post-industrial" society. In the first, the machine is the dominant metaphor - in the second life is the metaphor is: the brain? evolution of life? the ecosystem? These are all complex systems which are not predictable in the way a clockwork is.
In predicting the future, determinism is likely to predict trends, possibly cyclical trends (i.e., generational cycles). But trends do not continue indefinitely, because real systems include correction mechanisms - they tend to maintain a certain kind of order, to stay within limits, even if they are not strictly predictable. We construct alternative visions of the future, write scenario, do backcasting, in an attempt to anticipate a future which is not strictly determined. This is what Fukuyama is about.
Some past visions of the future have been strikingly on-target, such as Vannevar Bush's Memex Machine and Licklider's prediction of the use of computers in communication. This is because they were grounded in technology, and because they anticipated what people would want to develop.
Many people have suggested that computers will become intelligent, that this will be one wave of the future. What would this mean? Not just doing one thing quickly and well, but being able to respond to unexpected situations. How do we get there? Approaches to programming AI include: rule based, neural net and evolutionary approaches. Each had its strengths but did not do the job of creating true intelligence. (September 17).
So we want to philosophy and psychology to get a better idea of how people think - of the fundamental logic of thought (Sept 24). We found that thinking evolves and develops, it is not mechanical. Numerical archetypes are one way of conceptualizing this on a very general level
This led us into a more modern approach, "chaos theory," which introduced the concepts of attractors, strange attractors, evolution and autopoiesis. (Oct 1)
What do we think is evolving? Perhaps a "global brain." What would that mean? (Oct 22)
How does the Brain actually
work? Oct 29. Neuronal groups and major structures which interact
with each other.
Organized both heterarchally
and hierarchally - works through both evolution and autopoiesis.
Idea structures in the brain compete with each other.
How can this be programmed? On a network of computers, each of which has a determinist, von Neumann architecture. Object-oriented programming in which large numbers of objects interact with each other, drawing on a library of patterns. The Webmind software is an example, with different kinds of links and agents that interact with each other according to defined patterns.
The economy is self-organizing, led by this technological revolution. Microsoft emerged as a hierarchal principle, bringing some order to the hetararchy, now the government is trying to take that role away...