Man-Computer Cooperation in Decisions Requiring Common Sense

Men and computers could cooperate more efficiently in real-time systems¿and perhaps in long-range planning too¿ if a man could tell the computers how he wanted decisions made, and then let the machine make the decisions for him. In the next few years there will probably be considerable pressure on system designers to adopt such arrangements. The problem of enabling a man to convey his decision rules to a machine will in many cases prove less formidable than it might at first appear. Three methods are discussed. As experience with man-machine cooperation of this type accumulates, problems for research will be generated. An attempt is made to foresee what some of them will be.