| The last seven years have seen the field of artificial intelligence (AI) transformed. This transformation is not simple, nor has it yet run its course. The transformation has been generated by the emergence of expert systems. Whatever exactly these are or turn out to be, they first arose during the 1970s, with a triple claim: to be AI systems that used large bodies of heuristic knowledge, to be AI systems that could be applied, and to be the wave of the future. The exact status of these claims (or even whether my statement of them is anywhere close to the mark) is not important. The thrust of these systems was strong enough and the surface evidence impressive enough to initiate the transformation. This transformation has at least two components. One comes from the resulting societal interest in AI, expressed in the widespread entrepreneurial efforts to capitalize on AI research and in the Japanese Fifth-Generation plans with their subsequent worldwide ripples. The other component comes from the need to redraw the intellectual map of AI to assimilate this new class of systems-- to declare it a coherent subarea, or to fragment it into intellectual subparts that fit the existing map, or whatever. |
|
|
|