Seminar #71: Histology for Robot Engineers
Mon. 19th, Sept. 2010 Nick Hockings presented a Life and Mind seminar on:
Histology for Robot Engineers
Mon. 19th, Sept. 2010
Fulton Bldg. 109
An analysis of why robots constructed of rigid materials fail as biomimetic systems and as embodiment for lifelike intelligence.
This seminar: Examines histology as materials science underpinning biomechanics. Introduces the concept of ‘material embodiment’. Demonstrates that the body is not an assembly of parts, but a single fibro-elastic continuum, that uses hyper-elastic fiber-gels to achieve low friction dynamics and energy conservation. Presents a series of engineering proposals for how to build soft bodied robots with near human tactile sensation and dynamics.
Nick will be presenting this seminar at the Italian Institute of Technology, Genova on Wednesday, so all critique and feedback will be greatly appreciated.
Compliant Paws – the video clip showing robot & animal gait pathology.
Watch the cats toes and remember they do not have voluntary control of pronation/supination or individual digit flexion.
Video of the first 50mins (sorry about the auto-focus hunting back and forth)
Seminar #67: Attractor Based Evolution
Chris Gordon-Smith Friday, 18th June, 2010, http://www.simsoup.info/
- Summarises key issues for the origin of life
- Identifies relevant conceptual background for attractor based evolution of chemical networks
- Sets out a network oriented view of chemistry
- Outlines what is envisaged in an attractor based evolutionary scenario
- Spells out explicitely a mechanism for attractor based inheritannce
- Shows an artificial chemistry that could support evolutionary exploration of a large number of alternative ‘phenotypes’
- Illustrates the relationship between molecular structure and chemical network structure with a four attractor artificial chemistry
- Provides an update on related progress with the SimSoup project
The origin of life is a multi-disciplinary subject; the material is accordingly intended for a multi-disciplinary audience.
Seminar #63: On the origins of genetic systems
University of Compiègne, France
Wed., March 24, 2010
Pasteur showed that contemporary organisms always derive from other already existing organisms. Nowadays, there is no spontaneous generation, not even for the simplest extant organisms such as bacteria. However, if life does always come from life, and if we go back step by step in time, there is a logically inescapable conclusion: we must arrive in the end at the very first organisms of all, and these by definition must have arisen by spontaneous generation. So, how is this possible?
Seminar #62: Skills, Springs and Contexts
Department of Psychology
MIC, Limerick, Ireland
Fri., March 19, 2010
The concept of skill plays a pivotal role in much enactive thinking, particularly when the approach is applied to questions and research topics typically the preserve of Cognitive Psychology (the so-called “higher cognitive functions”). Drawing on standard findings in traditional cognitive research as well as extant uses of the concept and its cognates in the enactive and dynamical sensorimotor literature I will attempt to layout some of possible accounts of skill available to enactive theorists, and some of the constraints we must obey in developing such accounts. Beginning with some very basic observations about biological autonomy I will examine the continuity between basic forms of adaptivity and the more complex concept of skill, putting forward one generic model that may provide for further empirical and modelling research. I will also discuss the relationship between skill and perspective as it is described by enactivists.
Seminar #57: Failed intentions, or why adaptive behavior is not sufficient for cognition
Department of Logic and Philosophy of Science
University of the Basque Country, San Sebastián
Mon., Dec. 7, 2009
First, I will analyse the “Cognition = Behaviour” and show how it is insufficient to account for cognition as a specific phenomenon different to any dynamical systems (call it “agent”) coupled to another one (call it “environment”). Then I will introduce the idea, defended by many, that the biological grounding of sensorimotor interactions can solve the problems posited by a purely behaviouristic account of cognition. Next, I will try to show that the existing formulations of the necessity and sufficiency of biology for cognition have not been formulated with precision and that additional requirements need to be added. Thus, I will expand the “Cognition = Adaptive Behaviour” into “Cognition = Adaptive Behaviour + X + Y + Z” where X, Y and Z represent those additional requirements that bring adaptive agency closer to intentional agency. I will try to argue that, despite the adding up of new conditions, the approach remains essentially problematic. To account for these problems I shall rely on a benchmark test: the skyline of our own subjective experience of intentionality-in-action (or intentional agency). Finally I will end up proposing some in-principle reasons for this failure and how we can still safe part of the basic intuition by moving the baby to a different bathwater: from biological autonomy to sensorimotor autonomy, from the self-maintenance of metabolic organization to the self-maintenance of the behavioural organization.
Seminar #54: Shallow and deep embodiment: Reasons for embracing enactivism
Ezequiel Di Paolo
4:30 p.m. Wednesday, 4th February, 2009 – Arundel 401
Over the last two decades the computational, skull-bound view of the mind has been challenged from various different angles (from autonomous robotics to sensorimotor theories of perception to cognitive linguistics). These challenges are often described as embodied because they show the non-trivial (constitutive and not merely informative) role played by bodily structures and their situatedness in the world.
In so far as such challenges concentrate on providing alternative working mechanisms for cognition, they remain susceptible to being re-interpreted in functionalist terms and so they are bound to inherit the blind-spots of functionalism, among others: a lack of a definition or grounding for terms like identity, autonomy, agency, value, meaning, cognition, and even sociality.
In contrast to this shallow sense of embodiment, a deeper sense is provided by the enactive approach to life and mind. For this approach, the starting point is a re-thinking of the problems of cognitive science in terms of continuities between life and mind. The approach embraces the traditional “how does it work?” question but puts the foundational emphasis on a different one: “what is it to be a mindful system?”
Seminar #47: Extending Autopoiesis to Incorporate Behavior
4:30 p.m. Wednesday, 4th February, 2009 – Arundel 401
“Bare” autopoiesis fails to fully address behavior and explains only direct self-maintenance (Di Paolo, 2005). As a case in point, up until recently, most computational models of autopoiesis fail to demonstrate any behavior beyond direct self-maintenance. One of the first models of autopoiesis to demonstrate behavior is that by Suzuki and Ikegami (2004,2006) which extends Varela, Maturana and Uribe’s original 1974 model with a few extra rules, resulting in a chemotactic autopoietic unity. The way that Suzuki and Ikegami extended the original model resulted in a highly entangled organization of behavior and autopoiesis — that is to say, the mechanism of behavior and of autopoiesis are one and the same. I will describe some limitations of this kind of organization and outline some alternative more “decoupled” organizations. I will then describe our new model of autopoiesis that incorporates decoupled behavioral mechanisms that we have developed to explore some of these ideas. Finally I will report on some ideas that have been clarified through our explorations of our model: “Dynamical Operational Closure” and “Autonomous Behavioral Mechanisms”.
Seminar #36: The Complex Language of Living Processes
The genome projects have shown us that reading an organism’s genetic information is not sufficient to explain how organisms make themselves as coherent, functional, adaptive wholes. This opens biology to new ways of understanding evolution and development. I shall explore this from a perspective that understands molecular networks as languages that express the meaning of an organism’s history and context in the embodied form of the organism, allowing us to interpret molecular data from a new perspective. Organisms then become agents whose coherent form and behaviour reveal emergent qualities that we can recognise directly and reliably, providing a basis for a science of qualities.
Seminar #32: How Enactivism relates to Reflexive Monism
Abstract: Dualists believe that experiences have neither location nor extension, while reductive and ‘non-reductive’ physicalists (biological naturalists) believe that experiences are really in the brain, producing an apparent impasse in current theories of mind. Enactive and reflexive models of perception try to resolve this impasse with a form of “externalism” that challenges the assumption that experiences must either be nowhere or in the brain. However, they are externalist in very different ways. Insofar as they locate experiences anywhere, enactive models locate conscious phenomenology in the dynamic interaction of organisms with the external world, and in some versions, they reduce conscious phenomenology to such interactions, in the hope that this will resolve the “hard” problems of consciousness. The reflexive model accepts that experiences of the world result from dynamic organism-environment interactions, but argues that such interactions are preconscious. While the resulting phenomenal world is a consequence of such interactions, it cannot be reduced to them. The reflexive model is externalist in its claim that this external phenomenal world, which we normally think of as the “physical world,” is literally outside the brain. Furthermore, there are no added conscious experiences of the external world inside the brain. In closing the gap between the phenomenal world and what we normally think of as the physical world, the reflexive model of perception and the broader reflexive monism it supports provide a different approach to the hard problems of consciousness. Conversely, while enactive models have useful things to say about percept formation and representation, they fail to address the hard problems of consciousness.
Seminar #31: Non-representationalism in Action
Abstract: I examine the following question: Do actions require representations? Recent work by Mark Rowlands, Michael Wheeler, and Andy Clark suggests that actions may require a minimal form of representation. I argue that the various concepts of minimal representation on offer do not apply to action per se and that a non-representationalist account that focuses on dynamic systems of self-organizing continuous reciprocal causation at the subpersonal level is superior. I further recommend a scientific pragmatism regarding the concept of representation.
Seminar #28: The Enactive Torch
The enactive approach to perception is generating an extensive amount of interest and debate in the cognitive sciences. One particularly contentious issue has been how best to characterize the perceptual experiences reported by subjects who have mastered the skillful use of a perceptual supplementation (PS) device. This paper argues that this issue cannot be resolved with the use of third-person methodologies alone, but that it requires the development of a phenomenological pragmatics. In particular, it is necessary that the experimenters become skillful in the use of PS devices themselves. The ‘Enactive Torch’ is proposed as an experimental platform which is cheap, non-intrusive and easy to replicate, so as to enable researchers to corroborate reported experiences with their own phenomenology more easily.
Seminar #27: Life and its close relatives
Nathaniel Virgo and Simon McGregor
IT’S ALIVE!!! (Sort of.) We propose that dissipative structures and the production of entropy should be part of the A-Life research
programme, explaining why lifelike properties like self-maintenance are common in the inanimate world and deserve our attention. Examples such as hurricanes, candle flames and photocopiers are even closer to living organisms than they appear on first inspection, and they are easier to study. Inspired by the theory of autopoiesis we’ll suggest that studying these kinds of physical phenomena can give us insight into the dynamics of self-maintenance in living organisms and help us explore its relationship to cognition. We’ll display some cool simulations exploring the behaviour of some simple dissipative systems and showing just how lifelike they can be.
Seminar #26: Exploring the continuity between life and mind
Ezequiel Di Paolo
Enactivism (Varela et al 1991, Thompson, 2007) articulates several major shifts towards embodied cognition in terms of its biological and experiential roots. In order to prosper, enactivism must move beyond its basis on sensorimotor skills into the problems of higher mental function. But before we can solve this problem, we should provide a proper context to deal with it in a naturalistic setting, as opposed to the apparent naturalism (but actual in-depth mysticism) of cognitivism/computationalism in its classic and embodied versions.
I propose that such a context is given by exploring the connections between life and mind. Several key aspects of mind are already present in minimal forms of life, often with such clarity that their study enables us to approach higher forms of cognition in a much more rigorous and disciplined way (in particular, by providing operational definitions and explanations of terms such as values, norms, identity, meaning, autonomy, behaviour, agency, and social interaction, all grounded in dynamical and systemic categories).
This does not mean that there are no jumps and specializations in the history (evolution and development) of cognitive systems. In fact, the enactive paradigm is very much concerned with explaining those jumps (which only make sense as such once we understand the fundamental continuities that remain unaffected by them). What is common to all these transitions is that they involve, through some specific embodied activity, the emergence of novel forms of agency and sense-making. The result is always a new form of cognitive engagement involving increased mediacy between the agent and its cognitive target.
In this talk I will give some rough overview of these ideas, concentrating on the notion of autonomy as a guiding thread.
Seminar #20: From Causality to Autonomy: A Quantitative Approach
I introduce a quantitative measure of autonomy based on a time series analysis adapted from `Granger causality’. A system is
considered autonomous if prediction of its future evolution is enhanced by considering its own past states, as compared to
predictions based on past states of a set of external variables. The proposed measure, G-autonomy, amplifies the notion of autonomy as `self-determination’. I illustrate G-autonomy by application to example time series data and to an agent-based model of predator-prey behaviour. Analysis of the predator-prey model shows that evolutionary adaptation can enhance G-autonomy.
Seminar #18: Is autopoiesis necessary for consciousness?
A living system can be distinguished in two non-intersecting phenomenal domains, namely 1) as a collection of components and their relationships (i.e. its physiology) in the constitutive domain, and 2) as a unity and its relationships to other unities (i.e. its behavior) in the behavioral domain.
There is general agreement in the autopoietic tradition that consciousness is a relational phenomenon that pertains to the behavioral domain. However, this creates a potential tension with the claim that autopoiesis in the constitutive domain (i.e. metabolic self-production) is a necessary condition for consciousness.
I propose that one way to resolve this issue is by distinguishing between phenomenal and behavioral consciousness, even though this opens up the theoretical possibility of philosophical “zombies”.
Seminar #2: The role of AI in the paradigm shift towards an embodied-embedded cognitive science
In this talk I will look at the role of AI and robotics in the ongoing paradigm shift of the cognitive sciences. I will start by analyzing the shift from cognitivism towards a more embodied-embedded cognitive science, and conclude that this move was mostly due to empirical developments in AI which allowed purely philosophical arguments to be resolved experimentally. More recently, there has been a lot of discussion of a further shift towards enactivism. I will try to give a brief overview of the central aspects of an enactive theory of cognition. In particlar, I will highlight enactivism’s central concern with the notion of subjectivity and then analyze AI and robotic’s role (if any) in facilitating this second shift. It might turn out that this shift in understanding needs something qualitatively different from the resolution of disputes in either the philosophical or empirical domain of the cognitive sciences.
Seminar #1: Play as re-creation: an enactive route to human sense-making
Ezequiel Di Paolo