This is more of a computing text than the title suggests. The book explores the progress of artificial intelligence (AI) at modeling human intelligence. The author has a long history working within AI.
The book’s scope includes areas that may not immediately seem part of “cleverness,” such as everyday communication, the computational modeling of mind or knowledge, and creative language use. It is really more about computational models of mental capabilities rather than what constitutes cleverness. The actual chosen title shows (to paraphrase Partridge on p. 393): “it is obvious that ... [there is] emphasis on capturing an audience as well as presenting the science.”
In the preface and glossary, Partridge outlines ten useful maxims surrounding our attempts to understand intelligence, for example, “intelligence admits no simple tests” (maxim 2) and “failure drives science forward” (maxim 7). The maxims form the framework for this book. I’d question maxim 3, on the difficulty of modeling the salient parts of a phenomenon such as intelligence with “understanding based on a single example” (that is, humans). Early on, Partridge discusses the variability between different people’s brain activity patterns. If using brain activity to understand intelligence (see chapters 2 through 5), we have several examples to learn from. Also, it is by no means a given that humans are the only intelligent species. What about whales or dolphins? Or the social intelligence demonstrated by ants?
The book dissects several attempts to model intelligence computationally, such as reverse engineering, holistic approaches, machine learning, and semantics. The book is clearly written, and the flow of arguments from chapter to chapter is clearly signposted. Partridge’s writing style is opinionated, and his views are clearly affected by writers such as Fodor and Kurzweil, but this doesn’t negatively affect the readability of the text. Technical terms are used and carefully explained (often at painstaking length, which may be tedious for the average Computing Reviews reader). Thoughtfully incisive comments are made on learning and adaptation as key parts of intelligence and the role of maintaining inconsistent beliefs within the scope of our knowledge. There is exciting discussion on the potential of ultra AI, that is, AI that surpasses human intelligence and better utilizes computational capabilities such as pattern processing and comparatively larger working memory.
Given Partridge’s experience with AI research, I had hoped to read more about recent interesting advances in AI that received less attention than headline-grabbing systems like Deep Blue, Watson, and MYCIN. Instead, Partridge restricts examples to the more prominent systems and is extremely critical of the (lack of) achievements of AI. Partridge often employs a technique of presenting dubious theories as fact, and then using them as the basis for constructing arguments before discrediting the theories (sometimes several chapters later). Examples include the adoption and discarding of Fodor’s language of thought as an underlying model of how thoughts are structured in the mind, or the positing of the existence of a brain module solely for calculating triangle hypotenuses, later acknowledged as fiction.
The book is for AI novices with the patience to sit through longer examples (often over several chapters), who can be easily guided through the process of making and questioning assumptions at Partridge’s preferred pace. For someone with more knowledge or someone who questions as they read, this book can be difficult, although Partridge stresses (p. 115): “If you can’t just take my word for it (and when playing the scientist, you shouldn’t), then you could test it.” You must read the full text; recommending a standalone chapter to a student could be problematic, as Partridge may not reconcile tenuous assumptions until later chapters.
Overall, the assume-investigate-discredit style employed in this book leads me to question its attacking approach by again quoting Partridge (p. 137): “After all our efforts have come to naught, are we any wiser?”