Surfing Uncertainty: Prediction, Action, and the Embodied Mind by Andy Clark | (PDF) Free Download

Description

How is it that thoroughly physical material beings such as ourselves can think, dream, feel, create and understand ideas, theories and concepts? How does mere matter give rise to all these non-material mental states, including consciousness itself? An answer to this central question of our existence is emerging at the busy intersection of neuroscience, psychology, artificial intelligence, and robotics.In this groundbreaking work, philosopher and cognitive scientist Andy Clark explores exciting new theories from these fields that reveal minds like ours to be prediction machines – devices that have evolved to anticipate the incoming streams of sensory stimulation before they arrive. These predictions then initiate actions that structure our worlds and alter the very things we need to engage and predict. Clark takes us on a journey in discovering the circular causal flows and the self-structuring of the environment that define “the predictive brain.” What emerges is a bold, new, cutting-edge vision that reveals the brain as our driving force in the daily surf through the waves of sensory stimulation.

User’s Reviews

Editorial Reviews: Review “Surfing Uncertainty will be a much discussed and seminal work in the field of the philosophy of cognitive science.” — David D. Hutto, Australasian Journal of Philosophy”A stimulating read for anyone interested in the intersection of neuroscience and philosophy of mind from a scientific perspective.” –Library Journal”A wonderful book…Clark’s Surfing Uncertainty will become an essential point of departure for philosophers and cognitive scientists trying to come to grips with the apparatus of predictive processing.” — Metascience “This is a truly important book. It is evocatively written and reflects a truly gargantuan amount of work. It sets the stage for future debates not only about the empirical merits of Bayesian characterizations of human cognition, but also the broader philosophical picture in which such Bayesian characterizations are embedded. I predict that many of us will be reading, discussing, and analysing this book in the months and years to come.” –British Journal for the Philosophy of Science About the Author Andy Clark is Professor of Logic and Metaphysics in the School of Philosophy, Psychology and Language Sciences, at Edinburgh University in Scotland. He is the author of Being There (1997), Mindware (2001), Natural-Born Cyborgs (2003), and Supersizing the Mind (2008). His interests include artificial intelligence, embodied cognition, robotics, and the predictive mind. In 2018 he was profiled in The New Yorker.

Reviews from Amazon users which were colected at the time this book was published on the website:

⭐One has to admire and appreciate Clark. This work represents a tremendous effort at bringing together the many developments in, or related to, the emerging Predictive Processing framework (PP). Though familiar with some of this literature, I was surprised at the scope, depth and breadth of the research efforts in this framework that have been occurring over the years. The PP framework argues that there is, in the brain, both an up-flow of information from the external world, and most significantly, a down-flow of predictive information, with the predictive information flow actually having priority (suppressing the up-flow), though this priority is constrained by the degree of error in the match between the two flows. The predictions in this neural network/computational framework are in essence probabilistic Bayesian priors, and in one of the significant developments Clark describes, networks/algorithms are being created that learn these priors from the incoming information, thus solving an important bootstrap problem as to how this predictive process gets going. The book goes into the many perceptual phenomena – numerous illusions – that PP can be extended to explain, and the deep ins and outs of this story. Clark devotes a great deal to developing the crucial role of bodily action in this picture – how it is integrally part of PP flows. Throughout, Clark is continually relating things to the many other theorists in or related to this subject – a welcome point of tie-in. Of interest to me were his attempts to tie/reconcile all this to the “enactive” theorists such as Varela, as well as to Gibson’s theory of affordances and the structure of perceptual information. Indeed, Clark describes how PP can be seen/read as supporting the creation of multiple simultaneous “priors” relating to possible affordances (actions) re any given perceptual scene. Clark is well aware that this framework is no solution to the problem of perception, i.e., what I like to term the origin of the image of the external world. We have two neural or chemical-neural flows – up and down – neither of which look anything like Clark’s example coffee cup on the kitchen table, i.e., like our actual experience of the external world. The down-flowing “prior” for a coffee cup is, nevertheless, only neural firings. He chides PP theorist Jacob Hohwy for arguing that this framework can indeed explain the origin of our experience, as Hohwy invokes the notion that from these flows emerges a “virtual image” of the coffee cup and table. If Clark tends to sin, it is that he slips continually into over-enthusiastic language, e.g., PP is, “…a compelling ‘cognitive package deal’ in which perception, imagination, understanding, reasoning, and action are co-emergent…” If one has no theory of the origin of our image of the external world, one wonders how one can claim a theory of perception? And given no theory of perception, then a theory of memory – of how this perceived experience is stored – is ungrounded (indeed, there is no accepted theory), and given this, how can we be saying we are actually explaining imagination – something which relies on invoking this experience? Yet one finds this type of language throughout the book. Just to give an idea of where the limitations might lie with which PP and Clark struggle, let me place PP and Clark within another perspective for a moment. PP is an impressive computational framework, but mere, abstract computations cannot account for consciousness – that image of the coffee cup, “out there,” on the table, with spoon stirring away. For this, a real, concrete dynamics is needed. As an example of such, consider Bergson’s concept, described in his “Matter and Memory,” 1896. Bergson, to the detriment of his contemporaries understanding him, had presciently anticipated the essence of Gabor’s discovery of holography by fifty years, as well as Bohm’s later generalization of the holographic to the universal field (The Implicate Order, 1980). But contrary to anyone who later tried to employ the holographic framework, e.g., Pribram, what Bergson had argued, in essence, is that the brain is creating (forms globally, across the totality of its processing) a modulated reconstructive wave passing through the external, holographic field and specific to, or specifying, a source of information in the field, and by this process now an image of a subset of the field – the coffee cup and spoon. We are explaining then how perception is limited, not how it arises. The selection of information from the “hologram” is based on its relation to action – to the body’s action capability – and thus, per Bergson, perception is “virtual action.” This is where the “multiple potential affordances as priors” would fit, but this is a very concrete dynamics supported/created by the brain – a concrete wave – as concrete as an AC motor generating a field of electric force. It is within this concrete dynamical structure that the entire PP model and its computations would somehow have to be incorporated as a part – certainly not the whole – of the story. And as perception, in this framework, is not occurring solely within the brain (the specification is to events/objects within the external field, right, for example, where the coffee cup indicates it is), experience cannot be solely stored there (perhaps the source of our theoretical difficulty on storing experience?). A new model of memory retrieval would be required. Nevertheless, exception above noted, I consider Clark’s work an invaluable and to-be-appreciated source for getting up to speed on the current state of the art and theory in cognitive science and robotics.

⭐This book is a revelation. Well written and researched, Clark explores the deep implications of the limits of human knowledge on how we move through the world. Probabilistic thinking and computations are powerful, ubiquitous tools that guide us through a world in which split-second decisions are necessary for survival. From perception to action, emotions to moral life, we can’t escape uncertainty. Enjoy!

⭐Basically, Clark spends the whole book giving a vast overview of the probabilistic turn in neuroscience and cognitive science, emphasizing everywhere how the brain does not merely estimate probability densities of its observables, but also estimates the expected precisions of its actual observations. This helps Clark and the probabilistic mind-scientists to unify a truly astounding amount of experimental data and theoretical machinery.Clark emphasizes the work of the Friston and Hohwy labs in neuroscience, and occasionally the Tenenbaum and Gershman labs in computational cognitive science. Clark also relentless emphasizes the embodied mind: how based on estimated precisions, the brain learns and recruits chiefly “action-oriented” probability models (or in other words, models which infer an action directly from sensory data, without necessarily encoding a causal reconstruction of the world from sensory data that accounts for all causal structure). In less technical language, the brain uses simpler representations when it expects its experiences to be noisier.Once you start to understand the general principle, though, you end up realizing the dive into the work is slightly shallow. While too many equations would have bogged down the book, I don’t feel like I even remotely understand how precision-weighting can actually work based on the total lack of equations (and graphs, etc) given. Please just show me an equation and a graph to give some intuition for how nonparametric (arbitrarily complicated) probabilistic inference can include estimates of precision. It’s easy to assume a normal curve and then infer the precision as a parameter, but what about all the real-world quantities the brain handles that aren’t normal curves?I did, however, especially appreciate the discussion of how precision-weighted probabilistic inference plays into the current view of several major mental illnesses. As the family and friend of some sufferers, I’m pleasantly surprised to find out computational psychiatry is even a thing!For the deeper look at how specific cognition problems get solved in this framework, I’ve ended up purchasing “The Predictive Mind” by Jacob Hohwy. We’ll see how that goes.

⭐A beautifully written scientific account of how brain and body ‘think’.

⭐As a non-scientist with a deep and ongoing interest in philosophy of mind and cognitive science, this seems to me to be a very significant work that transcends the usual divisions and polarities. It promises to reconcile embodied cognition with elements of connectionism to produce something like an account of ’embodied computation’. It convincingly argues that viewing the brain as a prediction engine allows us to account for the way that perception is integrated with action, while also accounting for the possibility of delusion and error. It is one thing to say that the mind is not ‘in’ the brain (Noe), and another thing to show how the brain nonetheless specifically enables cognition. Clark seems to have done exactly that. Putting Clark’s ‘surfing’ together with Noe’s ‘dancing’ may be the future for cognitive science.

⭐I think the main hypotheses are true, but this book didn’t prove it to me, and it’s probably quite difficult to get further without massive simulations.

⭐Great concept, but very heavy reading and quite repetitive. The ideas could have been conveyed in about half the actual text.

⭐A nice and very innovative account of the human mind seen as a Bayesian learning machine comprising an endless ladder of hiearchies. Exciting and inspiring read.

⭐Clark meticulously lays out the case for a predictive processing model of cognition in the embodied brain (and potentially for AI) that anyone with patience and an interest should be able to follow. Better still he maps out paths for future exploration. This must be one of the best single volume summaries of a field of intellectual endeavour I’ve seen. Highly recommended.

⭐Questo libro mi ha veramente convinto che riusciremo a capire come funziona la nostra mente; l’idea che introduce e’ relativamente semplice, documentata da esperimenti che vanno indietro di 20 anni, capace di fare previsioni sul comportamento della nostra mente che possiamo verificare con piccoli e semplici esperimenti.il difetto del libro e’ uno stile di scrittura un po’ pesante, con a volte termini insoliti, periodi lunghi e diverse ripetizioni degli stessi concetti. Detto questo, l’idea principale viene illustrata all’inizio del libro e quindi non siete costretti a leggerlo fino in fondo per apprezzare il libro.

Keywords

Free Download Surfing Uncertainty: Prediction, Action, and the Embodied Mind in PDF format
Surfing Uncertainty: Prediction, Action, and the Embodied Mind PDF Free Download
Download Surfing Uncertainty: Prediction, Action, and the Embodied Mind 2015 PDF Free
Surfing Uncertainty: Prediction, Action, and the Embodied Mind 2015 PDF Free Download
Download Surfing Uncertainty: Prediction, Action, and the Embodied Mind PDF
Free Download Ebook Surfing Uncertainty: Prediction, Action, and the Embodied Mind