Building Complex Worlds for Artificial Life, Artificial Culture and Evolutionary Computation

Talking about COMPLEX SYSTEMS
Multiagent Modeling / Numerical Simulation / Agent Based Modeling / Artificial Worlds

Welcome to a new way of looking at the world. This is an introduction to a new philosophy of science, one that applies what we've learned about evolution and computation in order to explore the complexities of the natural and cultural worlds. We build models and simulations of agents interacting in environments situated in both space and time. We begin with simple representations of time, space and agency and progress in steps to representations that are more complex. We create these artificial worlds and the ways these worlds work. We specify how space and time work in these worlds. We specify how its agents sense the world around them, how they think and how they then effect the world around them. We, as individuals, are not omniscient, there are limits to our perceptions of the world. Thus we can specify that the knowledge each agent acquires is local. In the real world, we see complex global patterns of interaction that none of the local participants are aware of. This gap between local and global knowledge is reflected in the phrase "from simple local rules to complex global patterns of behavior." We too can study this phenomenon among our agents interacting in our artificial worlds. In the natural world, we often see these complexities and wonder about their causes. The causes are elusive. Sometimes, when we try to analyze a complex system by cutting it into parts, we destroy our understanding of how it works. Our project is to take another approach, to synthesize a world from the bottom-up. We ask, "what if?" We create the agents in computer code, defining each in detail. We place them in a spatial and temporal environment. Then we press "run." We turn them loose and watch how their local ineractons create dynamically new global patterns of behavior. Then we ask, "Do these results match what we seen in the real world?" If not, we modify our "what if" scenario and begin again. By trial and error, by the insight of experiment and evolution, we tweak our simulation until it is a good fit to reality. We are looking to discover the processes that create and maintain the world we live in. We want to describe, understand and explain how all these processes are interrelated.

Our knowledge of the world is indirect. It is based on our individual biological senses, our perceptions and beliefs. Over countless generations they have been modified by culture. We come to know the world by intuition, both conscious and unconscious, which allows us to make decisions quickly. We also come to know the world through natural language, which allows us to communicate, but which is relatively slow. To these we have added other ways to know our world: new technologies to describe, understand and explain the world in which we live. We think, reason and communicate not just with thoughts and language, but with artifacts as well. Art, diagrams, maps & charts, sculpture, architectural, nautical and aircraft models are literally things with which to think. The sciences have created the formal languages of mathematics, algebra and calculus. We have ported a great deal of our cognition over to so-called "intelligent" devices. In the the 1950s we began to add evolution and multiple causation into our models and simulations of complexity. This is the way of knowing that we lay before you to in this course.