A review of [Computer Power and Human] [Reason], by Joseph Weizenbaum. The book's purpose is to show that there are certain activities that computers ought not to engage in and there are certain kinds of research that ought not to be done. For instance, Weizenbaum would forbid research aimed at development of speech recognition systems, and also efforts to couple an animal's visual system and brain to computers. The reviewer feels that the author has failed in his arguments to provide a logical structure to be used to decide what activities are allowed and which are not allowed. If the book is viewed as an exposition, not about computers and human reason, but about Weizenbaum--a description of the thoughts and feelings of one who has thought and felt long and hard about computers, then it is a superb book. It is full of insight, charm, and wisdom. 6 pp.