Dr Murray Shanahan
, Reader in Computational Intelligence, Department of Computing, Imperial College London
Murray Shanahan graduated from Imperial College London in 1984 and gained a PhD in computer science from Cambridge University in 1988.
His early 1990s work on logic-based knowledge representation culminated in "Solving the Frame Problem" (MIT Press, 1997), a book whose citations have considerably outnumbered its sales. Since the mid-1990s he has worked on cognitive robotics, initially drawing on his earlier work to develop a theoretical account of robot perception, but more recently exploring the potential of brain-inspired cognitive architectures.
Cognitive Robotics: A Tale of Two Methodologies
Against the backdrop of the conflict between classical and biologically-inspired artificial intelligence, this talk will present two approaches to cognitive robotics. The first approach attempts to revive the ideals of the Shakey project, and to endow robots with high-level cognitive skills through the use of traditional AI concepts such as logic-based representation and symbolic reasoning. The second approach appeals for help to empirical disciplines such as neuroscience, cognitive psychology, and the study of animal behaviour, and favours neural nets, brain-inspired architectures, and the language of dynamical systems. An attempt will be made to to assess the prospects for each methodology.