From Artificial Evolution to Artificial Life
Timothy John Taylor
University of Edinburgh
This work addresses the question: What are the basic design considerations for creating a synthetic model of the evolution of living systems (i.e. an `artificial life' system)? It can also be viewed as an attempt to elucidate the logical structure (in a very general sense) of biological evolution. However, with no adequate definition of life, the experimental portion of the work concentrates on more specific issues, and primarily on the issue of open-ended evolution. An artificial evolutionary system called Cosmos, which provides a virtual operating system capable of simulating the parallel processing and evolution of a population of several thousand self-reproducing computer programs, is introduced. Cosmos is related to Ray's established Tierra system [Ray 91], but there are a number of significant differences. A wide variety of experiments with Cosmos, which were designed to investigate its evolutionary dynamics, are reported. An analysis of the results is presented, with particular attention given to the role of contingency in determining the outcome of the runs. The results of this work, and consideration of the existing literature on artificial evolutionary systems, leads to the conclusion that artificial life models such as this are lacking on a number of theoretical and methodological grounds. It is emphasised that explicit theoretical considerations should guide the design of such models, if they are to be of scientific value. An analysis of various issues relating to self-reproduction, especially in the context of evolution, is presented, including some extensions to von Neumann's analysis of self-reproduction [von Neumann 66]. This suggests ways in which the evolutionary potential of such models might be improved. In particular, a shift of focus is recommended towards a more careful consideration of the phenotypic capabilities of the reproducing individuals. Phenotypic capabilities fundamentally involve interactions with the environment (both abiotic and biotic), and it is further argued that the theoretical grounding upon which these models should be based must include consideration of the kind of environments and the kind of interactions required for open-ended evolution. A number of useful future research directions are identified. Finally, the relevance of such work to the original goal of modelling the evolution of living systems (as opposed to the more general goal of modelling open-ended evolution) is discussed. It is suggested that the study of open-ended evolution can lead us to a better understanding of the essential properties of life, but only if the questions being asked in these studies are phrased appropriately.
Many people have contributed in a variety of ways to the production of this thesis over the last three and a half years.
First of all, I would like to thank John Hallam, my supervisor. John took me on as a student, and has always been ready to offer guidance, discussion and constructive criticism of my work. He has also allowed me the freedom to follow my own academic interests in the course of my work, which I greatly appreciate. Thanks John!
The interdisciplinary nature of my work has required me to seek advice and discussion from a wide variety of people. Their generosity in giving up their valuable time to answer my questions and read drafts of some of the following chapters has consistently surprised me (especially those whom I know only by email exchanges).
Chief among my list of `extra supervisors' is Mark Bedau from Reed College in the USA. Since first meeting Mark in early 1997 at the Philosophy of Artificial Life conference in Oxford, he has provided vital encouragement, suggestions and discussion of my work. From a practical point of view, he also developed a range of analysis and visualisation techniques for evolutionary systems, which I have used extensively. I would also like to thank Emile Snyder, who implemented the analysis software and has helped me to use it. Thanks, Emile, you saved me a lot of time!
Tom Ray read through an early draft of the Cosmos design document, and provided useful comments and suggestions. The work reported in Section 6.1 was originally published in the proceedings of the Sixth International Conference on Artificial Life in 1998, and I would like to thank Chris Adami and the four anonymous reviewers who offered comments on the original paper. Towards the end of my research I have had some invaluable discussions, and comments on drafts of various chapters, from a number of other people. I especially thank Nick Barton, John Holland, Howard Pattee and Moshe Sipper for reading and providing useful comments on Chapter 7 (and Chapter 2 as well in Nick Barton's case). Thanks also to Barry McMullin for providing clarification on some issues, and to Daniel Mange for some interesting discussion.
I have had a great time at a number of conferences over the last few years, and these have been made even more enjoyable by a number of people I've met at them. In particular, thanks to Jason Noble and Alastair Channon--see you at ECAL'99?
During my research I have also benefited from some much broader discussions about artificial life, artificial intelligence, life, the universe and everything. These discussions have often been lubricated with plenty of beer, and fed with copious amounts of chicken korma and egg foo yung. I am indebted to the other members of the Mobile Robots Group in the (former) Department of Artificial Intelligence, for making it such a fun place to work. In particular (at the risk of offending the others), I'd like to say a special thank you to Simon Perkins, Richard Reeve, Arturo Espinosa-Romero, John Demiris, Nuno Chagas, Sandra Gadanho and Heba Al-Lakany. Thanks also to Judith Good for the therapeutic moaning sessions about our theses, and to Bernard Goffin for putting up with our constant droning about them during dinners at Teviot.
There are many other people at the department who have made my life a lot easier. The librarians, Olga Franks and Janice Gailani, have always been more than helpful. To name a few more, Ken Dawson, Karen Konstroffer, Margaret Rennex, David Wyse and Dougie Howie have also been very helpful in a number of ways. I'd also like to thank the organisers of the EUCS Perl course, for what must be the most useful half-day course I've ever attended.
I would not have been able to conduct this research without the financial assistance of a grant from the Engineering and Physical Sciences Research Council (grant number 95306471), for which I am very grateful.
I don't have enough room to adequately thank my parents for everything they have done, in so many ways, to enable me to be in the position of submitting this thesis. In retrospect, it was a Christmas gift from them some years ago of a paperback book--on the face of it insignificant in comparison to everything else they have done for me--that planted the seed for my growing interest in evolution and artificial life and ultimately led me to write this thesis. The book was The Blind Watchmaker by Richard Dawkins. Thanks, Mum and Dad, and thanks also to Anne, Katie, Rich and Tamsin.
Of course, it hasn't all been work over these last three and a half years, and I'd like to thank those who have kept me away from artificial life and helped to make my real life so enjoyable. Apart from various friends already mentioned, thanks to all the (ex-)Strathfillan Road gang--Claudio, Trish, Gordon, Jen, Matt and Wakako. A big thank you to Simon and Ögmundur for all the squash matches and pints in the bar afterwards--I hope we keep in touch. Thanks also to Stu for the regular and very welcome coffee breaks, and to Chris and Alan for really taking my mind off work!
But most of all, there is one person who has been with me through every hour of this time, and without whom I can't imagine having done this. Thank you Pete, for your constant love, affection, friendship and support.
I hereby declare that I composed this thesis entirely myself and that it describes my own research.
May 29, 1999