Tim Taylor | Monash University (original) (raw)

Books by Tim Taylor

Springer, 2020

Is it possible to design robots and other machines that can reproduce and evolve? And, if so, wha... more Is it possible to design robots and other machines that can reproduce and evolve? And, if so, what are the implications: for the machines, for ourselves, for our environment, and for the future of life on Earth and elsewhere?

In this book the authors provide a chronological survey and comprehensive archive of the early history of thought about machine self-reproduction and evolution. They discuss contributions from philosophy, science fiction, science and engineering, and uncover many examples that have never been discussed in the Artificial Intelligence and Artificial Life literature before now. In the final chapter they provide a synthesis of the concepts discussed, offer their views on the field’s future directions, and call for a broad community discussion about the significant implications of intelligent evolving machines.

The book will be of interest to general readers, and a valuable resource for researchers, practitioners, and historians engaged with ideas in artificial intelligence, artificial life, robotics, and evolutionary computing.

The book can be purchased from Springer https://www.springer.com/gb/book/9783030482336, Amazon and other online retailers. Further information and additional materials, including a free author-formatted PDF version of the book, can be found at https://www.tim-taylor.com/selfrepbook/.

Papers by Tim Taylor

this paper) in providing a drive for evolvability has been neglected; even if a system has the ca... more this paper) in providing a drive for evolvability has been neglected; even if a system has the capacity for high evolvability, it will not realise this capacity if the appropriate selection pressures are absent.

Who Needs Genomes? Barry McMullin Dublin City University Barry. McMullin@ rince. ie Tim Taylor Ax... more Who Needs Genomes? Barry McMullin Dublin City University Barry. McMullin@ rince. ie Tim Taylor Axel von Kamp University of Abertay Dundee Dublin City University Tim. Taylor@ abertay. ac. uk Axel. vonKamp@ rince. ie for natural selection��quasi-deterministic displacements of one lineage by another. But more importantly for our purposes here, the patterns of variation establish the potential for continuing innovation, and ultimately, continuing growth of complexity. Maynard Smith and Szathmry (1999, pp. 7��9) a make a ...

This chapter addresses the nature of open-ended evolutionary processes, and the related, but more... more This chapter addresses the nature of open-ended evolutionary processes, and the related, but more subtle, issue of how fundamental novelty (i.e. creativity) can arise in such processes. A number of existing artificial evolutionary systems, such as Tierra (Ray, 1991), are analysed in this context, but it is found that the theoretical grounding upon which they are based does not usually consider all of the relevant issues for creative evolution. The importance of considering the design of the environment, and of interactions between individuals, as well as the design of the individuals themselves, is emphasised. The properties of a hypothetical ‘proto-DNA ’ structure—a suitable seed for an open-ended, and creative, evolutionary process—are discussed. A number of open questions relating to these issues are highlighted as useful areas of future research. Finally, a paradigm for an evolutionary process described by Waddington (1969) is described. It is suggested that this might represent...

The role of contingency (random events) in an artificial evolutionary system is investigated by r... more The role of contingency (random events) in an artificial evolutionary system is investigated by running the system a number of times under exactly the same conditions except for the seed used to initialize the random number generator at the beginning of each run. Twelve different measures were used to track the course of evolution in each run, and “activity wave diagrams” were also produced (Bedau & Brown 1997). The results of 19 runs are presented and analyzed. The performance of every run was compared with each of the others using a non-parametric test (a randomization version of the paired-sample t test). When comparing absolute values of the measures between the runs, some significant differences were found. However, looking at the difference in values between adjacent sample points for a run, no run was significantly different to any other for any of the measures. This suggests that the general behaviour is the same in all runs, but the accumulation of differences results in si...

It is argued that a fruitful, and as yet unexplored, avenue for artificial life research lies in ... more It is argued that a fruitful, and as yet unexplored, avenue for artificial life research lies in modelling organisms as organisations embedded within a dynamical system environment. From this perspective, the origin and evolution of life is the progressive control of the dynamical system at a local level by constraints which are represented on an organism’s genome. Such an approach shifts the focus of artificial life models away from the design of individuals, towards the interaction of an individual with its dynamic environment. It also admits no representational distinction between organism and environment. An evolutionary cellular automata system, called EvoCA, is introduced as a tool to explore these ideas. In EvoCA, an evolved individual is a collection of constraints on the state of specific cells in the CA. Results are presented of initial experiments to investigate the interaction of evolution with the dynamics of EvoCA under various regimes (as characterised by Langton’s la...

Artificial Life, 2000

In this brief position paper I highlight a number of issues relevant to the evolvability of arti ... more In this brief position paper I highlight a number of issues relevant to the evolvability of arti cial life systems. These issues became apparent over the course of my doctoral research which involved studying the evolutionary dynamics of a Tierra-like system (Taylor 1999a). The rst two issues (multifunctional phenotypic components and semantic closure) relate to the representation and function of individuals. The other two issues (embeddedness and ecological interactions) concern how individuals are related to the shared environment and to other individuals. While much of the recent work on evolvability has focussed on issues concerning individuals (e.g. genotype-phenotype mapping), which may give a system the capacity for high evolvability, I believe that the role of the ecological aspects of the system (e.g. the nal two issues addressed in this paper) in providing a drive for evolvability has been neglected; even if a system has the capacity for high evolvability, it will not real...

Our understanding of the evolution of living organisms, and attempts to derive general theories o... more Our understanding of the evolution of living organisms, and attempts to derive general theories of evolution, are hampered by the fact that we only have one example of life to study—life on Earth. There are therefore many situations where we are uncertain whether a particular feature of life is general to any comparable system of evolving self-replicators, or whether it is specific to the biological organisms on our planet.

Perhaps one of the few features shared by most artificial life approaches is that a phenomenon ob... more Perhaps one of the few features shared by most artificial life approaches is that a phenomenon observed in biological life is studied by constructing a bottom-up model, in which a number of low-level components and interactions are explicitly encoded, and one or more higher-level phenomena are expected to emerge. While this is a perfectly valid approach, one has to be careful about how the model is constructed if it is to bring any scientific insight to bear on the phenomenon in question. Too many (but not all) studies of artificial life (my own included) have adopted a sloppy approach in the past, and this has meant that the field of artificial life has not contributed as much as it might have done to broader areas of scientific knowledge. This paper highlights some areas of general methodology which should be carefully considered when designing a bottom-up simulation for scientific experimentation, and also suggests some considerations that are specifically relevant to A-life mode...

This paper describes work I intend to pursue over the course of my PhD (although it presents a di... more This paper describes work I intend to pursue over the course of my PhD (although it presents a discussion of general ideas rather than a detailed research proposal—that will come later). My primary research interest is in the study of evolutionary processes through the use of computational models. In particular, I intend to study evolution in systems of self-replicating programs, especially those in which there is a developmental process to map genotypes to phenotypes. The reasons for this choice, and discussion of the analogy I wish to draw between the proposed digital system and existing biological systems, are explained. I believe that this work will be of interest both from the perspective of making some contribution to our understanding of biological evolution and also from the perspective of the evolution of computer programs in their own right.

The first detailed mechanistic models for genome based reproduction were developed by John von Ne... more The first detailed mechanistic models for genome based reproduction were developed by John von Neumann in the period 1948-1953 (von Neumann, 1948, 1949; Burks, 1966). While these models were extremely abstract, subsequent elaboration of the structure and function of DNA proved von Neumann’s designs to have been strikingly prescient. However, some significant questions still remain as to the specific benefits of this particular reproductive architecture. These questions are relevant both to understanding the evolutionary emergence of such systems, and their proper role in engineered or synthetic evolutionary systems. This paper will review these issues, and present some preliminary results of novel evolutionary experiments in the Tierra system (Ray, 1992), where artificial “organisms” are deliberately engineered to have an evolvable genetic architecture. The Problem Situation This paper is concerned with evolutionary systems and their evolvability. By “evolutionary system” we mean a ...

A Genetic Regulatory Network-Inspired Real-Time Controller for a Group of Underwater Robots.

Presented at the Fourth Workshop on Open-Ended Evolution (OEE4) at the 2021 Conference on Artificial Life (ALIFE 2021), 2021

In previous work I proposed a framework for thinking about open-ended evolution (Taylor, 2019). T... more In previous work I proposed a framework for thinking about open-ended evolution (Taylor, 2019). The framework characterised the basic processes required for Darwinian evolution as: (1) the generation of a phenotype from a genetic description; (2) the evaluation of that phenotype; and (3) the reproduction with variation of successful genotypephenotypes. My treatment emphasized the potential influence of the biotic and abiotic environment, and of the laws of physics/chemistry, on each of these processes. I demonstrated the conditions under which these processes can allow for ongoing exploration of a space of possible phenotypes (which I labelled exploratory open-endedness). However, these processes by themselves cannot expand the space of possible phenotypes and therefore cannot account for the more interesting and unexpected kinds of evolutionary innovation (such as those I labelled expansive and transformational open-endedness). In the previous work I looked at ways in which expansive and transformational innovations could arise. I proposed transdomain bridges and non-additive compositional systems as two mechanisms by which these kinds of innovations could arise. In the current paper I wish to generalise and expand upon these two concepts. I do this by adopting the Parameter Space–Organisation Space–Action Space (POA) perspective, as suggested at the end of (Taylor, 2019), and proposing that all evolutionary innovations can be viewed as either capturing some novel physical phenomena that had previously been unused, or as the creation of new persistent systems within the environment.

Theoretical Ecology, 2020

Heterospecific pollen transfer by insect pollinators has the potential to drive inter-species com... more Heterospecific pollen transfer by insect pollinators has the potential to drive inter-species competition between flowering plants. This phenomenon may newly arise in a region if insect pollinator or flowering plant populations change. An agent-based simulation is presented to assess the potential impact of heterospecific pollen transfer by insects on two co-flowering plant species within an environment consisting of a shared central region and species-specific refugia. Where heterospecific pollen asymmetrically suppressed the reproduction of one competitor, the pollen recipient was rapidly ousted from shared regions. If pollinators made deep, repeated, forays into and out of plant refugia, the clogged species was even unseated from its own refugium. When heterospecific pollen symmetrically suppressed plant reproduction, the same effects were observed, but with one or the other species excluded at random by the pollen clogging mediated interaction. We conclude that both symmetrical and asymmetrical heterospecific pollen transfer may be important elements of inter-species dynamics. In particular, our simulation shows pollen and pollinator visits lost to heterospecific flowers may not always be wasted from the producer’s standpoint. Instead, heterospecific pollen delivery may convey a competitive advantage even when the recipient has a refuge safe from direct invasion. This is possible because the pollen producer may use pollinators to clog a competitor’s stigmas in a refugium without entering into competition there for space, nutrients, light, pollinators, or other resources. Consequently, the evolution of plant signals to promote pollinator constancy may not be the only effective strategy in inter- species competition.

Proceedings of the Conference on Artificial Life 2020 (ALIFE 2020), 2020

A paper in the recent Artificial Life journal special issue on open-ended evolution (OEE) present... more A paper in the recent Artificial Life journal special issue on open-ended evolution (OEE) presents a simple evolving computational system that, it is claimed, satisfies all proposed requirements for OEE (Hintze, 2019). Analysis and discussion of the system are used to support the further claims that complexity and diversity are the crucial features of open-endedness, and that we should concentrate on providing proper definitions for those terms rather than engaging in "the quest for open-endedness for the sake of open-endedness" (Hintze, 2019, p. 205). While I wholeheartedly support the pursuit of precise definitions of complexity and diversity in relation to OEE research, I emphatically reject the suggestion that OEE is not a worthy research topic in its own right. In the same issue of the journal, I presented a "high-level conceptual framework to help orient the discussion and implementation of open-endedness in evolutionary systems" (Taylor, 2019). In the current brief contribution I apply my framework to Hinzte’s model to understand its limitations. In so doing, I demonstrate the importance of studying open-endedness for the sake of open-endedness.

Artificial Life, 2019

Nature's spectacular inventiveness, reflected in the enormous diversity of form and function disp... more Nature's spectacular inventiveness, reflected in the enormous diversity of form and function displayed by the biosphere, is a feature of life that distinguishes living most strongly from nonliving. It is, therefore, not surprising that this aspect of life should become a central focus of artificial life. We have known since Darwin that the diversity is produced dynamically, through the process of evolution; this has led life's creative productivity to be called Open-Ended Evolution (OEE) in the field. This article introduces the second of two special issues on current research in OEE and provides an overview of the contents of both special issues. Most of the work was presented at a workshop on open-ended evolution that was held as a part of the 2018 Conference on Artificial Life in Tokyo, and much of it had antecedents in two previous workshops on open-ended evolution at artificial life conferences in Cancun and York. We present a simplified categorization of OEE and summarize progress in the field as represented by the articles in this special issue.

Artificial Life, 2019

Nature's spectacular inventiveness, reflected in the enormous diversity of form and function disp... more Nature's spectacular inventiveness, reflected in the enormous diversity of form and function displayed by the biosphere, is a feature of life that distinguishes living most strongly from nonliving. It is, therefore, not surprising that this aspect of life should become a central focus of artificial life. We have known since Darwin that the diversity is produced dynamically, through the process of evolution; this has led life's creative productivity to be called Open-Ended Evolution (OEE) in the field. This article introduces the first of two special issues on current research on OEE and on the more general concept of open-endedness. Most of the papers presented in these special issues are elaborations of work presented at the Third Workshop on Open-Ended Evolution, held in Tokyo as part of the 2018 Conference on Artificial Life.

Artificial Life, 2019

This paper presents a high-level conceptual framework to help orient the discussion and implement... more This paper presents a high-level conceptual framework to help orient the discussion and implementation of open-endedness in evolutionary systems. Drawing upon earlier work by Banzhaf et al., three different kinds of open-endedness are identified: exploratory, expansive, and transformational. These are characterised in terms of their relationship to the search space of phenotypic behaviours. A formalism is introduced to describe three key processes required for an evolutionary process: the generation of a phenotype from a genetic description, the evaluation of that phenotype, and the reproduction with variation of individuals according to their evaluation. The formalism makes explicit various influences in each of these processes that can easily be overlooked. The distinction is made between intrinsic and extrinsic implementations of these processes. A discussion then investigates how various interactions between these processes, and their modes of implementation , can lead to open-endedness. However, an important contribution of the paper is the demonstration that these considerations relate to exploratory open-endedness only. Conditions for the implementation of the more interesting kinds of open-endedness-expansive and transformational-are also discussed, emphasizing factors such as multiple domains of behaviour, transdomain bridges, and non-additive compositional systems. In contrast to a traditional Darwinian analysis, these factors relate not to the generic evolutionary properties of individuals and populations, but rather to the nature of the building blocks out of which individual organisms are constructed, and the laws and properties of the environment in which they exist. The paper ends with suggestions of how the framework can be used to categorise and compare the open-ended evolutionary potential of different systems, how it might guide the design of systems with greater capacity for open-ended evolution, and how it might be further improved.

Springer, 2020

Artificial Life, 2000

A Genetic Regulatory Network-Inspired Real-Time Controller for a Group of Underwater Robots.

Presented at the Fourth Workshop on Open-Ended Evolution (OEE4) at the 2021 Conference on Artificial Life (ALIFE 2021), 2021

Theoretical Ecology, 2020

Proceedings of the Conference on Artificial Life 2020 (ALIFE 2020), 2020

Artificial Life, 2019

Nature's spectacular inventiveness, reflected in the enormous diversity of form and function disp... more Nature's spectacular inventiveness, reflected in the enormous diversity of form and function displayed by the biosphere, is a feature of life that distinguishes living most strongly from nonliving. It is, therefore, not surprising that this aspect of life should become a central focus of artificial life. We have known since Darwin that the diversity is produced dynamically, through the process of evolution; this has led life's creative productivity to be called Open-Ended Evolution (OEE) in the field. This article introduces the first of two special issues on current research on OEE and on the more general concept of open-endedness. Most of the papers presented in these special issues are elaborations of work presented at the Third Workshop on Open-Ended Evolution, held in Tokyo as part of the 2018 Conference on Artificial Life.

Artificial Life, 2019

To be presented at the Third Workshop on Open-Ended Evolution (OEE3) during the 2018 Conference on Artificial Life (ALIFE 2018), Tokyo, Japan, 2018

This paper presents a high-level conceptual framework to help orient the discussion and implement... more This paper presents a high-level conceptual framework to help orient the discussion and implementation of open-endedness in evolutionary systems. Drawing upon earlier work by Banzhaf et al., three different kinds of open-endedness are identified: exploratory, expansive, and transformational. These are characterised in terms of their relationship to the search space of phenotypic behaviours. A formalism is introduced to describe three key processes required for an evolutionary process: the generation of a phenotype from a genetic description, the evaluation of that phenotype, and the reproduction with variation of individuals according to their evaluation. The formalism makes explicit various influences in each of these processes that can easily be overlooked. The distinction is made between intrinsic and extrinsic implementations of these processes. A discussion then investigates how various interactions between these processes, and their modes of implementation, can lead to open-endedness. However, it is demonstrated that these considerations relate to exploratory open-endedness only. Conditions for the implementation of the more interesting kinds of open-endedness - expansive and transformational - are also discussed, emphasizing factors such as multiple domains of behaviour, transdomain bridges, and non-additive compositional systems. In contrast to a traditional population genetics analysis, these factors relate not to the generic evolutionary properties of individuals and populations, but rather to the nature of the building blocks out of which individual organisms are constructed, and the laws and properties of the environment in which they exist. The paper ends with suggestions of how the framework can be used to categorise and compare the open-ended evolutionary potential of different systems, and how it might guide the design of systems with greater capacity for open-ended evolution.

Atlantic Symposium on Computational Biology and Genome Information Systems and Technology, Mar 1, 2001

An approach to time-and space-di��erentiated pattern formation in multi-robot systems Tim Taylor... more An approach to time-and space-di��erentiated pattern formation in multi-robot systems Tim Taylor1, Peter Ottery1 and John Hallam2, 1 1 Institute of Perception, Action and Behaviour, 2 The Maersk Mc-Kinney Moller Institute, School of Informatics, University of Edinburgh, University of Southern Denmark, JCMB, The King's Buildings, May��eld Road, Campusvej 55, Edinburgh EH9 3JZ, UK DK-5230 Odense M, Denmark tim@ tim-taylor. com Abstract We consider the problem of non-trivial pattern formation in decentralised multi-robot systems, ...

A decentralised real-time controller for a group of robots is presented, the design of which is i... more A decentralised real-time controller for a group of robots is presented, the design of which is inspired by biological genetic regulatory networks. A geneticalgorithm (GA) is used to automatically evolve controllers for specific tasks. Resultsof initial experiments are presented and analysed, which demonstrate that it is possibleto successfully evolve the controllers to achieve a simple clustering task. Performanceis robust under a variety of parameter choices for the GA and controller.

The reference for this paper is: Tim Taylor," Twin Studies of Homosexuality", Undergrad... more The reference for this paper is: Tim Taylor," Twin Studies of Homosexuality", Undergraduate Dissertation, Department of Experimental Psychology, University of Cambridge, 1992. Twin Studies of Homosexuality Tim Taylor Department of Experimental Psychology Cambridge University April 1992 1. INTRODUCTION The phenomenon of homosexuality is one that has endured, to a greater or lesser extent, throughout all societies from the ancient Greeks and beyond to the present day (eg Bullough 1976). In 1948, Kinsey et al. reported that 37% of ...

Nidus Design Document Tim Taylor Department of Arti cial Intelligence University of Edinburgh 5 F... more Nidus Design Document Tim Taylor Department of Arti cial Intelligence University of Edinburgh 5 Forrest Hill, Edinburgh EH1 2QL, UK 14th June 1998 timt@ dai. ed. ac. uk http://www. dai. ed. ac. uk/daidb/people/homes/timt/1 Introduction This document describes Nidus1, a model (actually, a family of models) designed to investigate the essential components and interactions required to support the origin and evolution of living organizations. Sections 2 {3 describe the motivations for developing the model, and ...

Artificial Life, 2001

Karl Sims' work [25, 26] on evolving body shapes and controllers for three-dimensional, physi... more Karl Sims' work [25, 26] on evolving body shapes and controllers for three-dimensional, physically simulated creatures generated wide interest on its publication in 1994. The purpose of this article is threefold: (a) to highlight a spate of recent work by a number of researchers in replicating, and in some cases extending, Sims' results using standard PCs (Sims' original work was done on a Connection Machine CM-5 parallel computer). In particular, a re-implementation of Sims' work by the authors will be described and discussed; (b) to illustrate how off-the-shelf physics engines can be used in this sort of work, and also to highlight some deficiencies of these engines and pitfalls when using them; and (c) to indicate how these recent studies stand in respect to Sims' original work.

Creative Evolutionary Systems, 2002

Proceedings of the Sixth International Conference on Artificial Life, Aug 28, 1998