42047 - Minds, Brains and Machines (MCM) [UB]

How should intelligence be modelled? There does seem to be a general agreement within the Cognitive Sciences (Psychology, Neuroscience, Artificial Intelligence) that intelligence is computation. However this agreement, these disciplines differ on the adequate level of explanation in which computation should be characterized. Computational Neuroscience, for example, attempts to understand how brains "compute" but it emphasizes descriptions of biologically realistic neurons and their physiology. But is this an adequate level of explanation?

Cognitivists criticise connectionism for being too low level (Fodor & Pylyshyn, 1988), while neurobiologists complain that connectionism abstracts too far from real neural processes (Crick, 1989) ...Purely computer-based simulations are criticised by advocates of sub-threshold transistor technology (Mead, 1989) and by supporters of ‘real-world' robotic implementations (Brooks, 1986). Some worry about oversimplification (Segev, 1992) while others deplore overcomplexity (Maynard Smith, 1974; Koch, 1999). Some set out minimum criteria for ‘good' models in their area (e.g. Pfeifer, 1996; Selverston, 1993); others suggest there are fundamental trade-offs between desirable model qualities (Levins, 1966).

The aims of the course are to discus these issues and to briefly introduce AI students in the fields of computational neuroscience, neuroscience and psychology to see how these disciplines can enrich each other.

Topics include sensation, motion, emotion, perception, neural codes and information representation, embeddedness, semantics, attention, language, reasoning, decision making, developmental humanoid robotics, consciousness and free will.

See detailed current content here.