On the thresholds of knowledge

摘要: We articulate the three major findings of AI to date: (1) The Knowledge Principle: if a program is perform complex task well, it must know great deal about world in which operates. (2) A plausible extension that principle, called Breadth Hypothesis: there are two additional abilities necessary for intelligent behavior unexpected situations: falling back on increasingly general knowledge, and analogizing specific but farflung knowledge. (3) as Empirical Inquiry: we test our ideas experimentally, large problems. Each these hypotheses proposes particular threshold cross, leads qualitative change emergent intelligence. Together, they determine direction future research.