Al and ES: Possible Solutions for the Mineral Industry Brain Drain

WHAT IS AI? It's very hard to say. Most researchers and practitioners disagree with nearly any attempted definition, but most would agree that A1 does deal with the development of computer-related methods o f problem-solving. Artificial intelligence is a discipline which seeks to emulate human anatomy and/or thought processes by applying computer science, electronics, and mechanical engineering to the development of natural language processing, machine vision, robotics, and knowledge-based expert systems. Integration of these four fields will someday produce the robots of today's science fiction. A1 addresses problems of structuring large amounts of knowledge and representing relationships with rules and semantic frameworks rather than with mathematical equations. A1 is convergent rather than divergent - - conventional conputer applications use relatively small quantities of input to generate large output, but A1 generally reduces large amounts of information to establish a few major conclusions. The development of A1 has prompted greatly increased efforts on the part of scientists and engineers to understand just how humans think, organize information, approach problem-solving, and form conclusions. Some of the more exciting and possibly far-reaching advances in A1 are coming in this more-or-less soft-science area. Thirty years of A1 research (dating from a 1956 conference at Dartmouth College at which the term A1 was coined) has already spawned many widespread, almost ordinary, products: window graphics, hand-held controllers (mouse). function menus, etc., all in an attempt to make computers and people more compatible. WHAT IS AN ES? A1 can be used to develop EXPERT SYSTEMS, which are computer programs or systems which contain knowledge and rules by which problems can be solved. or at least by which possible solutions can be suggested for very complex problems. A1 can be used to better define problems so that solutions can be found. An ES is designed not to process numbers, but to process information according to rules and logic derived from actual practice and experience. WHY AI? WHY ES? Companies are increasingly aware of the value of accumulated knowledge and experience in designing, building, operating, and maintaining the increasingly complex or expensive physical plants in any business. As knowledge branches out and becomes more specialized, the accumulated learning of a veteran operator or manager or engineer becomes a scarce resource well worth preserving. EXPERT SYSTEMS offer a means of doing that, reinforcing, not replacing, human capabilities.