Mathematical Models: Dynamical AI
Hamid R. Ekbia · Cambridge University Press eBooks · 2008
Here is then the only expedient … to leave the tedious lingering method, which we have hitherto followed, and instead of taking now and then a castle or village on the frontier, to march up directly to the capital or center of these sciences, to human nature itself; which once being masters of, we may every where else hope for an easy victory. – David Hume (1739): A Treatise of Human Nature Complex systems and processes evade simple analysis. Cognition, for example, is a complex process that emerges from the joint activity of the brain, the body, and the environment. Each of these is, in turn, a complex system with millions and billions of components – neurons, cells, individuals – which, to make things even more complex, are also diverse in form and capability. Would it not be ideal if we could treat all of these complex systems, subsystems, and sub-subsystems using a single analytical tool? Advocates of the dynamical approach in AI and cognitive science take this question seriously. Their approach is strongly motivated by the search for a universal tool – or, as Scott Kelso, a leading light in the dynamical school, has put it, for “a common vocabulary and theoretical framework within which to couch mental, brain, and behavioral events” (Kelso 1995). Cognition, furthermore, is a temporal event, so it should be studied as such – that is, as a process that unfolds in time. This is a key intuition behind the dynamical approach.