Wednesday, August 30, 2006

Philosophy of Artificial Life

Towards Computationalism: Philosophy of Artificial Life
Li Jianhui
(Department of Philosophy, Beijing Normal University, Beijing 100875, China)

The computational revolution in the 20th century has produced a revolution in methodology in which computer simulations are performed as experiments. One outcome of this methodological revolution is the birth of the new frontier science, Artificial Life (Alife). Alife tries to use computers to create entities which exhibit characteristics of life in the computer or in the real world, and from the perspective of computation, it views life as a special algorithm. In January of 2004, China Book Press published my new book: Towards Computationalism: An Introduction to Philosophy of Artificial Life. This is the first book that systematically explores the philosophy of Artificial Life. Now I summarize the main contents and points of this book.

Because of the uniqueness and novelty of the goals, methods, and conceptualization of the Artificial Life, not only are many scientists attracted to do research in this area, but, in addition, many philosophers are attracted by the goal of generalizing new philosophical ideas from its concepts and theories. In Chapter One of this book, the challenges presented by Artificial Life to philosophy are illustrated.

The ideas of Alife can be traced to John von Neumann. Von Neumann proved that the logic of self-reproduction can be abstracted from life itself. After von Neumann, John Conway proved that the cellular automaton is equivalent to a Turing machine; Stephen Wolfram found that there are four kinds of cellular automatons; Chris Langton realized that the rule space of cellular automata closely relates to phase transition and computation: cellular automata which are at the edge of chaos can support complicated computation. Langton herein germinated the ideas of Alife: if we can create the conditions on the edge of chaos in some kinds of media, then we might create life in them. Chapter Two of this book mainly discusses the logic of life and the relation of life, information, and computation, elucidating the theoretical foundation of artificial life.

Chapter Three explores the constructive methodology of artificial life. Differing from a traditional top-down analytical and centralized controlling methodology, the new methodology of artificial life research is bottom-up and synthetic: it tries to simulate individual units instead of one big complex unit; to use local control instead of global control; to let the behavior emerge from the bottom up, instead of being specified from the top down. This kind of individual-based modeling has already become an important conceptual paradigm, not only for Alife, but for other sciences that study complex systems.

There are two types of artificial life: virtual artificial life and realistic artificial life. Virtual artificial life mainly uses software to create artificial life entities in a computer; realistic artificial life uses hardware to create artificial entities in the real world. The main ideas and contents of virtual artificial life and realistic artificial life are analyzed in Chapter Four and Chapter Five, respectively.

There are two claims about artificial life. Strong artificial life believes that Alife entities are or ultimately will be genuine life; while weak artificial life believes that Alife entities are merely simulations of natural life. Two claims are closely related to our definition of life. Chapter Six first probes the definition of life and then theoretically elucidates whether artificial life, especially virtual artificial life, is real life or not, based on the definition. If we understand life from an informational perspective, then we would support strong artificial life.

Chapter Seven discusses the ontological status of artificial life. If we view virtual life or digital life from the perspective of the interior processes of the computer, then their reality is indubitable. The creativity of computers has already made us believe that we can create an independent world, that there is a virtual life world in a computer, and that this virtual life world has the same ontological status as our real world. Gödel's incomplete theorem can't negate the possibility of digital life.

Artificial life views the essence of life as computation. If we enlarge our perspective, we find that there are many contemporary sciences, such as DNA computational theory, artificial intelligence, computational biology, etc., that not only view the essence of life as computation, but also view the essence of intelligence as computation. We can go even further to say that the whole world is constructed by algorithmic rules, and in some sense, the universe can be seen as a huge computational system.

Key Words: Life, Artificial Life, View of Life, Philosophy, Computation, Computationalism

Jianhui, Li. 2004. Towards Computationalism: Philosophy of Artificial Life. Beijing: China Book Press. Ⅷ+238p. $28, ISBN 7-5068-1175-8