decision and investigative process of the scientist, they can also wri te the paragra}ils of scientific resul ts. Once a canputer research activity progresses from intuitive guesses to a systematic study, the canputer program can free the scientist to pursue new original ideas. ' his concept is illustrated in the ptper by Hut. Sussnan says "Knowledge" is knowing what not to canpute. His main message is that we must change from trying to maximize the efficiency of the canputer to maximizing the effectiveness of the scientist. The canputer is relatively cheap, it is the trained scientist who is...