5XÉçÇøÊÓƵ

Oct 5

How signals from the heart influence emotion and memory
Prof. Sarah Garfinkel
UCL

Abstract: Cognitive and emotional processes are shaped by the dynamic integration of brain and body. A major channel of interoceptive information comes from the heart, where phasic signals are conveyed to the brain to indicate how fast and strong the heart is beating. This talk will discuss how interoceptive processes operate across conscious and unconscious levels. Empirical research will demonstrate how cardiac afferent signals can influence emotion and memory. The interoceptive channel is disrupted in distinct ways in individuals with autism and anxiety. Selective interoceptive disturbance is related to symptomatology including dissociation and the transdiagnostic expression of anxiety. New work on interoceptive training will demonstrate a reduction of anxiety with enhanced interoceptive precision following targeted feedback. The discrete cardiac effects on emotion and cognition have broad relevance to clinical neuroscience, with implications for peripheral treatment targets and behavioural interventions.

Fulton Building FUL-203

Oct 19

Common Sense Physics and Structured Representation in the Era of Deep Learning
Prof. Murray Shanahan
Imperial

Abstract: The challenge of endowing computers with common sense remains one of the major obstacles to achieving the sort of general artificial intelligence envisioned by the field’s founders. A large part of human common sense pertains to the physics of the everyday world, and rests on a foundational understanding of such concepts as objects, motion, obstruction, containers, portals, support, and so on. In this talk I will discuss the challenge of common sense physics in the context of contemporary progress in deep reinforcement learning, and the question of how deep neural networks can learn representations at the required level of abstraction.

Shawcross - AS02

Nov 2

Close encounters of the mechanical kind: when clever machines meet stupid theories
Dr Yasemin J. Erden
Twente

Abstract: Some of us who research artificial intelligence (AI) and machine learning (ML) aren’t too worried about the rise of the machines or of the singularity, and we rarely lose sleep about a robot uprising. We’re not so worried about what future intelligent machines might want to do, and instead we worry about what existing stupid machines already do. Programming mistakes, poor datasets, and flawed design can lead to some unpredictable errors and consequences, and many errors that could (and should) have been predicted. When we think about the role that AI and ML can have in psychiatry, I suggest we need to consider not only the risks that are caused by stupid machines, but also those that arise when clever machines rely on outdated, biased, harmful, and frankly stupid theories. This includes popular but flawed theories of minds and brains, and of illness and disorder, as found in disparate fields that include philosophy of mind, psychiatry, and neuroscience. Once we have this in mind, then we can next consider the kinds of positive roles that AI and ML might have as tools in psychiatry. In this talk we’ll therefore examine what AI and ML can bring to the practice of psychiatry, what they likely can’t, and what they definitely shouldn’t.

online

Nov 16

Emergence & downward causation in the brain: the role of Aristotelian causation.
Prof. George Ellis
5XÉçÇøÊÓƵ of Cape Town

Abstract: Strong emergence in the brain is possible because of downward causation, which is central to much brain function (perception is a key example) even though many reductionist physicists and even some neuroscientists deny it is possible. They claim, implicitly or explicitly, that mental functioning is epiphenomenal and agency cannot occur because the underlying physics controls all in a deterministic way. I counter by claiming that they are neglecting to take into account three of the four kinds of Aristotelian causation that demonstrably occur in biology: formal, material, and final causation all take place in addition to the efficient causation they focus on, with stochasticity at the micro level allowing higher level functions to select lower level outcomes so as to allow this to occur, and also allowing the abstract causation which is the foundation of rational actions and technology. In molecular terms, on the one hand gene regulatory networks and metabolic networks alter the structural components out of which higher level functions arise, and on the other hand, macromolecules allow branching dynamics by change of conformation which at the physics level alters constraints in the Hamiltonian. In philosophical terms, this underlies the key distinction between synchronic and diachronic supervenience, and breaks the alleged causal completeness of physics underlying Kim's arguments based in supervenience. Brains are open systems and have developed predictive processing mechanisms to successfully handle the unexpected by co-opting the underlying physics in these ways. For those uncomfortable with the notion of causation and its defintiion, one can use the idea of necessary and sufficient conditions instead.

online

Nov 23

Making music with machine learning and dynamical systems
Dr Chris Kiefer
Sussex

Abstract: Machine learning tools are becoming a common part of the computer-musician's creative toolbox, for varied applications in composition, mapping, sound design and production. Many of these tools use pre-trained models and non-realtime inference; there are opportunities to open up the process of designing and training machine learning models to musicians for deeper creative engagement. This requires fresh approaches compared to current commonly-used deep learning techniques, with models that are lightweight and open to realtime manipulation. This presentation will demonstrate such approaches, showing new reservoir computing techniques for sound synthesis, pattern generation and pattern recognition, with conceptor-controlled recurrent neural networks and FPGA-based boolean networks, exploring the creative opportunities these might give to musicians. This work is part of the AHRC project 'Musically Intelligent Machines Interacting Creatively'.

Shawcross - AS02

Nov 30

The Hidden Spring: A Journey to the Source of Consciousness
Dr Mark Solms
5XÉçÇøÊÓƵ of Cape Town

Abstract:I will start by rehearsing the critique by Chalmers and others of the functionalist approach to consciousness. Then I will argue that although this critique may hold for cognitive functions like memory and perception – which are not intrinsically conscious processes – it doesn’t hold for the affective function of feeling. How can you explain the function of feeling without accounting for its subjective phenomenology? Then I will argue that feeling is generated not by the cortex but by the brainstem structures that modulate cortical processes and thereby render them conscious. In other words, I will argue that cortical (cognitive) consciousness is contingent upon brainstem (affective) arousal. In this sense, cortical consciousness is secondary, a derivative of brainstem processes. I will close with some reflections on the implications of this conclusion for our understanding of the fundamental nature of consciousness.

online

Passcode: 8093

Dec 7

Unconscious Mental Imagery Requires Unconscious Mental Qualities
Dr Sam Coleman
Hertforshire

Abstract:It is widely agreed that conscious mental imagery features phenomenology, or conscious mental qualities, as I will say. Moreover, conscious imagery is accorded an important role in various sorts of action guidance. Unconscious mental imagery is also widely posited, and is held to share an important neurophysiological basis with conscious imagery (especially in the visual case I focus on). And unconscious imagery is accorded a very similar role in action guidance. But it is almost universally denied that unconscious imagery features mental qualities. I argue that unless we ascribe unconscious mental qualities to unconscious imagery, the behavioural contribution of conscious mental imagery is threatened, indeed, that conscious imagery is rendered epiphenomenal.

Shawcross - AS02

Passcode:

8093