Evolutionary Automata
Murray Shanahan · The MIT Press eBooks · 1994
The class of Evolutionary Automata is formally defined. An Evolutionary Automaton generates an evolving population of "organisms" within a computational medium. The organisms move randomly around a microworld, eat, reproduce asexually (with mutation), and die of "natural causes". The results of some preliminary experiments conducted with an instance of this class are reported, in which a parameter was varied across a small range. The paper concludes with some speculation on the possible significance of these results for the study of diversity and complexity in Artificial Life. Introduction One of the most attractive features of the manifesto of Artificial Life research is that it legitimises the study of any life-like phenomenon that can be reproduced in a computer. Artificial Life research avoids the prolific detail present in real-life biological phenomena, but at the same time enjoys the freedom to explore a larger space of possible phenomena than actually arise in nature [Langton...