REVIEW OF D. MICHIE: MACHINE INTELLIGENCE AND RELATED TOPICS (1985)


Computing Reviews 26(2): 101-104, 1985. Publication of the book under review: New York: Gordon and Breach, 1982.


Conclusion

As an account of some imaginative computer projects, I have found this book informative and interesting. However, as a contribution to clarifying the ideas underlying so-called Machine Intelligence, I have found a serious lack of in-depth discussion. As explained by the author, both the identity of the field and its achievements are unclear and seem to depend on basic assumptions concerning human beings that are never questioned, although highly problematic. This is compounded with uncritical, confusing use of humanoid terms, such as intelligence, thought, and learning. As a whole, the book leaves this reviewer with the impression that a thorough, self-critical revision by the AI people of their field in matters of goals, concepts, and terminology, is long overdue.