Elaine Dewar's Smarts–one journalist's search for the various meanings and mechanisms of natural and artificial intelligence

Abstract

In Smarts, Elaine Dewar has created a fascinating and informative account of her search for the latest scientific meanings of that elusive concept, intelligence. Starting out with her personal encounter as a child with IQ tests, this well-written and highly journalistic book meanders like a winding river through her encounters with scientists dealing in many different ways with human, animal, plant, microbial and in silico computer cognition. (To make full disclosure, I happened to be one of her interviewees, mostly for my work on bacterial multicellularity.) As a professional journalist, Dewar prefaces each encounter with a different researcher by relating her own relevant experience and revealing private expectations and misgivings as she travels to an interview or conference. This subjective treatment is unusual, but I found her sceptical approach a refreshing contrast to the over-simplifications and similes of the common popular science book. For the serious student or scholar, Smarts also contains ample references to key discoveries and the scientific literature. These proved useful to me, for example, in tracking down specific citations documenting our realization that plants sense predators and communicate information about attack to their conspecifics. Smarts is also enlivened by journalistic connections to contemporary issues. “Intelligence” as a government activity is not ignored. Edward Snowden makes an appearance as the book considers the consequences of how far electronics has invaded our lives in the form of traceable communications by “smart” phones and email. Dewar has little patience for anthropocentric illusions that intelligence is a specifically human capacity. She spends considerable space in the first half of the book discussing research on cognitive behavior, social interactions, tool usage, innovation and communication in different primates (chimps, apes, bonobos and orang-utans). As a journalist, she also delves unflinchingly into the politics and complex moral issues facing the celebrity scientists who study primate intelligence. From primates Dewar moves on to other higher mammals, like dolphins, before she enters into biological realms usually excluded from cognitive discussions. These include plants, as I indicated above, and the book devotes considerable space discussing the technical and political issues involved with the “plant neurobiology” movement. That account makes this review particularly relevant to this journal. A constant refrain in Smarts is mention of the work based on the discovery that Physarum slime molds find the shortest path through a maze to a food source. Although this is a clear example of sophisticated chemotaxis by a so-called “lower organism,” I could not help wondering if this was one of those cases where the physics and chemistry adequately explain the observations. The thought that a diffusion gradient of attractant would necessarily mark out the shortest path through the maze kept buzzing in my head. Segueing by way of Physarum-inspired computer modeling, most of the final section of Smarts is concerned with information scientists trying to uncover the basic algorithms of intelligence through in silico experiments. The major role of biomimetics in this research is covered implicitly. For example, Dewar devotes major attention to neural networks and evolutionary computation in robotic control. But I would have preferred to see the reciprocal influences of biological and information sciences on each other discussed more systematically. In particular, analysis of the way the emergence of artificial intelligence (AI) has revolutionized the basic issues in the mechanism-vitalism debate would have provided highly appropriate historical perspective to the narrative. At the end of the book, Dewar confesses her own confusion in coming to terms with some of the most ambitious attempts to model human brain function in a computer program. She points out the irony of representing analog systems on a digital platform. Her candor in admitting her own difficulties is part of what makes this book such a refreshing read. Coupled with the extensive documentation of published cognitive research, Dewar's gripping journalistic approach makes Smarts into a particularly rewarding book for those of us who ponder the mysteries of how living organisms can be so complex and get so much right so often.