Abstract:
This talk will range as far afield as examining how people make predictions, in particular are led into making false predictions; whether the Turing Test was a good idea; the Top 10 reasons AI is so far behind schedule; and, of course, what part I've played in all this (good, bad, and ugly -- but hopefully not temporally in that order) and hope to play the next few years. Sure, that includes talking some about machine learning (Why AM and Eurisko Appear to Work), expert systems, and Cyc.