objective of this chapter:
not just the study of particular complex systems
but: the study of the phenomenon of complexity in
its own right
economic systems
business firm
human mind
sophisticated engineering designs - specifically large software systems
social plans
after World War I:
holism and Gestalt
creative evolution
focus: "the whole transcends the sum of the
parts", anti-reductionist, emergence
after World War II:
information
feedback
cybernetics
general systems
focus: feedback and homeostasis
current (1970 and later):
catastrophe theory
chaos
adaptive systems
genetic algorithms
cellular automata
focus: mechanisms that create and sustain complexity and analytic tools for describing and analyzing complexity
cybernetics (Norbert Wiener)
combination of feedback control systems
positive and negative feedback loops
relationship to General Problem Solver
information theory - Shannon (bits and bytes) versus Chunks
stored-program computer
general systems theory
abstract from the general properties of physical,
biological amd social systems
question: do systems of such a diverse kind have
any nontrivial properties in common?
question: how does this relate to our efforts in domain-oriented design environments and the Turing Tar Pit?
catastrophe theory
two (or more) distinct steady states
moderate change of a system parameter may cause it
to shift suddenly to the other or into an unstable state
in practice: only a limited number of situations
have been found where it leads to further analysis
chaos theory
chaotic systems = deterministic dynamic systems that,
if their initial conditions are disturbed even infinitesimally,
may alter their paths radically
examples: weather
source - James Gleick: "Chaos : Making a New Science", Penguin, 1988
evolution = emergence
of complexity
genetic algorithm (from
Russell, Norvig: "AI - A Modern Approach")
function GENETIC-ALGORITHM (population, FITNESS-FN) returns an individual
inputs: population, a set of inidivuals
FITNESS-FN, a function that measures the fitness of an individual
repeat
parents <--- SELECTION (population, FITNESS-FN)
population <--- REPRODUCTION (parents)
until some individual is fit enough
return the best individual
in population, according to FITNESS-FN
cellular automata and the game of life (demonstrations of self-reproducing systems)
Our whole problem is to make our mistakes as fast
as possible John Archibald Wheeler
Karl Popper: "Conjectures and Refutations: The Growth of Scientific Knowledge", 1962
- we can learn from our mistakes
- conjectures are controlled by criticism
Henry Petroski: "To Engineer is Human - The Role of Failure in Successful Design", 1985
- design as revision
- human error <----> human nature
Leonard Lee: "The Days The Phones Stopped - How People Get Hurt When Computer Go Wrong", 1992
- unfriendly skies - airplane accidents (FAA, NTSB)
- phone systems which collapse