Keynote talk abstracts
Minoru Asada, Osaka University, Japan
Can "Synthetic Methodology" cause a paradigm shift?
Abstract. Recent technologies have enabled us to realize artificial systems close to us in various different aspects by modeling the development process of biological/behavioral/social aspects of humans.
In this talk, we discuss how "Synthetic Methodology" can make a paradigm shift in science and technology. Synthetic approaches are revisited as a candidate for the paradigm shift, and cognitive
developmental robotics is reviewed from this viewpoint. A trans-disciplinary approach is inevitable, and the central issues should be shared beyond the differences in disciplines. An overview of
our project entitled "Constructive developmental science based on understanding the process from neuro-dynamics to social interaction" is given as one of such challenges which consists of computational
modeling, functional imaging, psychological/behavioral experiments, and robot-platform development groups, and future issues of "Synthetic Methodology" are argued.
Angelo Cangelosi, University of Plymouth, UK
Embodied Language Learning: From Sensorimotor Intelligence to Symbols
Abstract. Growing theoretical and experimental research on action and language processing and on number learning and space representation clearly demonstrates the role of embodiment in cognition and language processing. In psychology and neuroscience this evidence constitutes the basis of embodied cognition, also known as grounded cognition (Pezzulo et al. 2012). In robotics, these studies have important implications for the design of linguistic capabilities in cognitive agents and robots for human-robot communication, and have led to the new interdisciplinary approach of Developmental Robotics (Cangelosi & Schlesinger 2014). During the talk we will present example of developmental robotics models and results from iCub experiments on the embodiment biases in early word acquisition studies, on word order cues for lexical development and number and space interaction effects. The presentation will also discuss the implications for the “symbol grounding problem” (Cangelosi, 2012) and how embodied robots can help addressing the issue of embodied cognition and the grounding of symbol manipulation use on sensorimotor intelligence.
1) Cangelosi A. (2012). Solutions and open challenges for the symbol grounding problem. International Journal of Signs and Semiotic Systems, 1(1), 49-54 (with commentaries)
2) Cangelosi A, Schlesinger M (to appear, 2014). Developmental Robotics: From Babies to Robots. Cambridge, MA: MIT Press.
3) Pezzulo G., Barsalou L.W., Cangelosi A., Fischer M.H., McRae K, Spivey M.J. (2011). The mechanics of embodiment: a dialog on embodiment and computational modelling. Frontiers in Psychology, 2(5), 1-21
Luciano Fadiga, University of Ferrara and Italian Institute of Technology, Italy
Robots for Brains
While considering the new cultural context of the Neuro-Robotics, the dominant view is that the principal role of neuroscience is to provide roboticists with information and models that could be useful to create intelligent/interactive artifacts. Only rarely one thinks that robotics could be useful for the advancement of neuroscientific knowledge. In my talk I will discuss this case showing when and how robotics could provide neuroscientists with fruitful tools and stimuli which may concretely improve our understanding of how the brain develops and works.
Stuart Kauffman, The Institute for Systems Biology, Seattle, USA
Answering Descartes: Beyond Turing
Abstract. A central conviction underlies most of our current ideas about brain and mind in current neurobiology. Contemporary neurobiology, based on classical physics thinking, seeks “consciousness neurons”. This view face the central philosophy of mind problem posed by Descartes then Newton. Descartes famously proposed a dualism, Res cogitans, mind “stuff”, and Res extensa, matter in motion. Res extensa set the stage for Newton’s formulation of classical physics with his invention of differential and integral calculus, three laws of motion, universal gravitation, and initial and boundary conditions. Given the initial positions and momenta of particles in a box, and the box as boundary conditions, and the laws of motion in differential equation form giving the forces between the particles, integration yields the DEDUCED trajectories of the particles. But deduction is “entailment”. So the current state of the particles, plus boundary conditions suffice to entail the next positions and momenta of the particles.
If the above is true, and the classical physics brain is a deterministic classical physical system, noisy or not, its current state, plus deterministic noise, entirely determines the next state of the brain. Then there is “Nothing for Mind, Res cogitans, to do.” Worse, “There is no way for mind to ‘do it’. This has been our problem since Descartes. Mind can have no role, for on the view from classical physics above, mind is forced to act “causally” on brain and cannot do so.
Strong Artificial Intelligence asserts that consciousness emerges in sufficiently complex computational systems, but faces the same problem as does classical physics neurobiology. Computers are discrete state, discrete time subsets of classical physics. Were mind to arise, it could not “do anything”. It could not change the behavior of the computer. Again there would be nothing for computer “conscious mind” to do, nor a way for mind to do it. How would the computer “use” its hoped for consciousness? To “do what”? (I remark that if consciousness did not “do anything”, why did it evolve?”)
A second famous problem arises for Strong Artificial Intelligence in which consciousness emerges. The issue is the “frame problem”, never solved in computer science since Turing. Here it is: “Name all the uses of a screw driver”. The number of uses is both INDETERMINATE AND UNORDERABLE. Thus, no effective procedure, or algorithm, can list all the uses of a screw driver or, in general, find new ones. If true, the frame problem cannot be solved by algorithmic computers. But we humans solve the frame problem all the time, as does evolution, another topic. If humans solve the frame problem and computers, syntactic machines, cannot, the mind is not a computer.
My alternatives lie in known and new aspects of quantum mechanics. First decoherence occurs in open quantum systems by losing phase information to an environment and thus approach classically infinitely closely, for all practical purposes, FAPP”. Second, decoherence is entirely ACAUUSAL. This means that if mind can be “quantum” in some respect, decoherence allows mind to have consequences for brain, but without ACTING CAUSALLY on brain. We begin to answer Descartes. To do so more than once can be achieved by “recoherence” then further decohrence. Recoherence is assured by Shor’s theorem and very probably shown experimentally in light harvesting molecules. My colleagues and I call this reversible hovering between “quantum” and “classical FAPP” worlds, the Poised Realm. Increasing evidence supports its reality.
Using the Poised Realm, realizable in chemical network systems which are fundamentally based on quantum mechanics and where the Poised Realm is now demonstrated, I will propose decoherence and recoherence as one answer to Descartes. In addition I will propose the testable hypothesis that conscious experience is associated with quantum measurement and that the unity of consciousness is underpinned by quantum entanglement and non-locality. Acausal quantum measurements creating molecular structures that are stable due to the Pauli Exclusion Principle may also be answers to Descartes. Because this view of mind-brain is not classical physics, it is also not algorithmic. Finally I will suggest, based on the Conway Kochen “Strong Free Will theorem”, that conscious experience and “doing” may be found at the level of electrons and elementary particles, as this theorem asserts. If so, we find the subjective pole, lost since Res extensa dominated science.
Pier Luigi Luisi, University of Rome Three
Contingency in molecular evolution
Abstract. An old dichotomy in the field of the origin of life and evolution in general, is between determinism and contingency. According to the first scenario, in the extreme formulation, the pathways of evolution are mostly pre-determined (although not by an intelligent designer) in a linear causality of events, where one event determines the next one and is caused by the previous one. This brings, for example according to Christian de Duve, to the inevitability of the origin of life, and eventually of mankind. By contrast, the view of contingency (Jay Stephen Gould, Meyer) asserts that there is no preordered pathway, what happens is a series of events determined by contingency (something that in the older literature, e.g. Monod, was called chance), so that one process may go in one direction or the other depending upon environmental conditions, and therefore something may happen or not. Also life on earth, and also mankind, might have not come about.
This last view, to which the majority of biological science today adheres in one way or another, poses the interesting question, why things in nature are in one way and not in another one. Why does DNA have ribose, and not glucose? Why twenty amino acids and not fifteen, or twenty seven? And, going up scale, why monkeys and hominids, instead of simple bacteria?
The point made in this lecture is that modern synthetic biology has the means to partly tackle this kind of questions. Examples from literature will be given (work by Eschenmoser, and by Doi and Yanagawa, Benner, et al.) although in a very quick review.
Then, as another extension of synthetic biology, the project of the never born proteins will be presented, where the main question is: why do we have in nature this particular set of proteins, which represents an infinitesimal part of the possibilities, and not another one? A few of these never born proteins will be presented and the above main question discussed. The procedure involves molecular biology and the so-called phage display method; but also a procedure based on much simpler bio-organic chemistry will be shown.
Synthetic biology is well known for the attempts to create new forms of life, by genetic manipulation of extant bacteria (see the work of Craig Venter and others). Can synthetic biology also make living minimal cell from scratch, with only basic chemistry techniques and without genetic manipulation? To tackle this question, the scenario that comes to mind is the autopoietic unit as defined by Maturana and Varela. In principle, this can be based on liposomes containing the simplest possible form of metabolism. This however implies that the liposome acts as a continuous bio-reactor, permitting the selective input of some component and the selective elimination of byproducts. This is still very difficult, and an alternative simpler solution will be presented, where the liposomes work as batch reactors. The notion of minimal cell will be discussed in these terms, and experimental data will be presented, which also bring to the unexpected proposal for the origin of cellular metabolism. Finally, it will be discussed how we could go from this to the original autopoietic unit as the minimal living entity.
Giorgio Metta, Italian Institute of Technology, Italy
Developing intelligence in humanoid robots
Abstract. I will present the iCub humanoid, a robotic platform designed for research in embodied cognition. At 104 cm tall, the iCub has the size of a three and half years old child. It can crawl on all fours and sit up to manipulate objects. Its hands have been designed to support sophisticate manipulation skills. The iCub is distributed as Open Source following the GPL/FDL licenses and can now count on a worldwide community of enthusiastic developers. The entire design is available for download from the project homepage and repository (http://www.iCub.org). About 25 robots have been built so far which are available in laboratories in Europe, US, and soon in Japan. It is one of the few platforms in the world with a sensitive full-body skin to deal with the physical interaction with the environment including possible people.
Scientific approach. The iCub stance on cognition posits that manipulation plays a fundamental role in the development of cognitive capability [1-4]. As many of these basic skills are not ready-made at birth, but developed during ontogenesis , we aim at testing and developing this paradigm through the creation of a child-like humanoid robot: i.e. the iCub. This “baby” robot is meant to act in cognitive scenarios, performing tasks useful for learning while interacting with the environment and humans. The small (104cm tall), compact size (approximately 22kg and fitting within the volume of a child) and high number (53) of degrees of freedom combined with the Open Source approach distinguish RobotCub from other humanoid robotics projects developed worldwide.
 L. Fadiga, L. Craighero, and E. Olivier, "Human motor cortex excitability during the perception of others' action," Current Biology, vol. 14 pp. 331-333, 2005.
 L. Fadiga, L. Craighero, G. Buccino, and G. Rizzolatti, "Speech listening specifically modulates the excitability of tongue muscles: a TMS study," European Journal of Neuroscience, vol. 15, pp. 399-402, 2002.
 G. Rizzolatti and L. Fadiga, "Grasping objects and grasping action meanings: the dual role of monkey rostroventral premotor cortex (area F5)," in Sensory Guidance of Movement, Novartis Foundation Symposium, G. R. Bock and J. A. Goode, Eds. Chichester: John Wiley and Sons, 1998, pp. 81-103.
 D. Vernon, G. Metta, and G. Sandini, "A Survey of Cognition and Cognitive Architectures: Implications for the Autonomous Development of Mental Capabilities in Computational Systems," IEEE Transactions on Evolutionary Computation, special issue on AMD, vol. 11, 2007.
 C. von Hofsten, "On the development of perception and action," in Handbook of Developmental Psychology, J. Valsiner and K. J. Connolly, Eds. London: Sage, 2003, pp. 114-140.
Giulio Sandini, Italian Institute of Technology, Italy
Brain for Robots
(The abstract will be provided soon.)
Ricard Solé, Pompeu Fabra University, Spain
Synthetic life: cells, machines and the boundaries of evolution
Abstract. Computation is an attribute of all forms of life. Every single living entity performs, at different scales, information processing involving both the external environment and the internal state. What defines life is actually deeply tied with the existence of such computational processes. To a large extent, evolution has been shaping computational levels on multiple scales and we can see that ants, cells and brains are capable of different levels of information processing and decision making. At the cell and tissue levels, the observed forms of computation remind us Turing machines or standard electronic systems. Are there other forms of computation beyond the standard engineering metaphors? Synthetic biology offers a unique possibility of exploring this question and even designing forms of computation that do not exist in nature. Here we consider the potential landscape of biocomputations and what might be beyond it, including other forms of collective intelligence and new challenges to our view of biological complexity.