BKAFFCMP.RVW 980412 "Affective Computing", Rosalind Picard, 1997, 0-262-16170-2, U$27.50 %A Rosalind Picard %C 55 Hayward Street, Cambridge, MA 02142-1399 %D 1997 %G 0-262-16170-2 %I MIT Press %O U$27.50 800-356-0343 manak@mit.edu www-mitpress.mit.edu %P 292 p. %T "Affective Computing" There will be, I dare say, a "religious" debate over this book. While isolated visionaries have idly speculated about emotion in computers, the vast majority of the computer using, and non computer using, populace sees technology as cold, mathematical, and ultimately objective (if occasionally in error). The fact that this assessment is an emotional one gets conveniently forgotten. One of the possible divisions in the study of artificial intelligence is in the approach taken. The brute coding approach simply strives to make programs more and more intelligent, the definition of "intelligent" being left as a problem to be dealt with once we have something that is at least marginally useful. This strategy has been demonstrably successful in producing entities like Deep Blue and techniques such as expert systems. The alternative route is to observe that we already have at least one agreed upon model of intelligence, and to seek to apply what we know of the human mind to some form of programming. While that course suggests interesting tactics like neural networks, spectacular triumphs have not been forthcoming. Still, it is in pursuing this modelling approach that Picard has divined a potentially revolutionary concept in computing. So radical is this idea, in fact, that even those who praise Picard and the book tend to see affective computing as only a means to a superior user interface, and miss the proposal that affect is key to intelligence itself. The postulate that emotion is important to intellect is not new. Picard acknowledges this, and, as any good scholar would, builds on the work done by others. Part one of the book provides an overview of emotion theory, as well as general questions about emotion and its relevance to the development of computing. Part two looks more specifically at the technical aspects of affective applications. Chapter one outlines basic human emotion studies, concentrating first on the physical aspects of emotion, and then on the cognitive. It is quite easy, in this section, to see why so many readers see the book as dealing merely with the user interface, since much of this primary material has immediate implications in allowing computers to assess the emotional state of the user, and to tailor presentation appropriately. The ramifications for computing are more subtle, but do exist, for example in the determination of urgent matters or the drive to learn from errors. An initial examination of affective components in computing is the emphasis for chapter two. In addition to the ability to recognize and express emotions, Picard lays out an interesting framework for deciding whether or not computers "have" emotions. While perhaps not final, it is a very useful aid to starting work on affective computing without diverting too much energy to more philosophical questions of defining affect. Applications that might benefit from affective computing, in chapter three, include coaching, communications, counseling, education, consumer studies, entertainment, motivation, search agents, and environmental agents. (As one indication of the importance of the work Picard proposes, consider how far we are from being able to give the command that she throws away as the title of one section:" Fast Forward to the Important Part.") Concerns, in chapter four, involve trust, fraud, poor application, privacy, accuracy, tradeoffs in objectivity, centralized manipulation, negative emotions, design, and computer rights. The author states outright that many of these considerations will only be problems in outlandish circumstances, or at far future dates. Chapter five starts to look at the technical issues involved in programming emotion, starting with the processing of affective signals in a way that mimics what we know of human emotional response. (My own immediate reaction was to the lack of a coding mechanism for measurement, storage, and communication, but I am probably getting ahead of the work.) "Recognizing and Expressing Affect," in chapter six, reviews current work in dealing with facial expressions, vocal intonation, and other factors. Further research is reported in chapter seven's look at the generation of emotion, primarily via cognitive mechanisms. Chapter eight discusses in more depth the concept of "wearable" computers, mostly in terms of the ultimate "personal" computer (or network). It has been proposed that the AI goal of reproducing human intelligence is a chimera and a false trail. Machine intelligence, so the thesis suggests, is different in kind from human intelligence, and the attempt to make one copy the other would be better directed to finding the differences between them and assigning work appropriately. If this latter hypothesis is true then Picard's recommended line of enquiry would be futile in terms of producing better machine intellect--but would still be valuable in determining the dividing line. In the Preface, the author demonstrates that she is aware that there will be--how shall we put this?--an emotional response to the central tenet of the text, and that, our culture being what it is, this reaction will be additionally weighted by the fact of her gender. Picard is walking a fine line between presenting a scientific thesis to a scientific community and proposing an exciting, but radical, new field of study. I suspect that she has deliberately chosen to display an objective and understated exposition, and that she was correct in doing so. However, the manner and structure of the book do readily allow readers to consider only the mundane, if valuable, human-machine interaction questions, and miss the more elusive and controversial points. I strongly recommend this book for all audiences. For general readers, there are many interesting discussion points, both technical and social. For developers, there is advice and direction to follow in terms of the user interface. For the zealot, there may be the key to the future. copyright Robert M. Slade, 1998 BKAFFCMP.RVW 980412