抄録
In an age of diversification and changes, team decision making is required to solve growing complexity and to respond to growing diversification. As team members come from a wide variety of areas, a decision cannot be made one-time at an early stage as it used to be and it must be reached by trials and errors step by step. Such dynamic decision making has to convince members at each step by providing different perspectives for each member to understand the line of reasoning, and it must admit lazy evaluation, because some members cannot understand what pieces of knowledge and experience are called for until later step, when clearer perspective is available to them. Steps of decision proceed to the next step by satisfying at least one member. If it fails, then it backtracks to the previous step until it satisfies one more member. This process is repeated until all members are satisfied enough. Artificial Intelligence allows such trial and error decision making to make all members feel satisfied enough. The usefulness of this approach is demonstrated by applying it to the development of WPS production support tool.
