Adaptive decision aids: using fallible algorithms to support decision making

It is argued that the principal difference between decision aids and most other types of information systems is the greater reliance of decision aids on fallible algorithms, which sometimes generate incorrect advice. Decision aids based on fallible algorithms can easily result in aided performance which is poorer than unaided performance. It is suggested that overcoming the cost of fallibility requires some form of adaptation, where the user and decision aid adjust their combined problem-solving activity in a problem-specific way