"Working with AI: Real Stories of Human-Machine Collaboration." Davenport, T. H. & Miller, S. M., 2022, MIT Press

Abstract

Davenport and Miller’s book “Working with AI: Real stories of human-machine collaboration” (MIT Press, 2022) is focused on showing and analyzing how AI is currently implemented in various organizations across the globe. This by itself makes it an interesting contribution to current scholarship, since so much of what is written about emerging technologies either focuses on technologies that have not yet been commercially deployed, or mixes present and future, making it at times hard to discern where the line between what exists in the present ends and what may come to exist in the future begins. Davenport and Miller’s focus on the present allows for a much more grounded debate about the social implications of AI technologies on humans, since instead of projecting either utopian or dystopian schemes on the future, the book deals with processes that are occurring today, that pose ethical challenges today, and that are having impact on humans today. Another important feature that sets this book apart is the richness of cases that the two authors bring to the table. The book offers no less than twenty-nine case studies, from different economic sectors, with different application types, and from different corners of the world (specifically from North America and Asia). Each case study includes a concise, yet very informative, depiction of an application of an AI technology (or sometimes a combination of a few AI technologies) in a certain organization. The authors skillfully offer sufficient description to make the ways in which the AI is used in each case clear, yet without going into too many details which might render the text tedious. All in all, this richness of case studies culminates in quite an informative text. Thus, if you are interested in how AI is currently deployed in a specific field, you will, most probably, find a relevant case study in this book. Moreover, within the mix of AI applications discussed in the book, you can also find some of the more ethically challenged applications, such as in the fields of healthcare and policing, which may appeal specifically to scholars who focus on risks within AI. Unfortunately, the book’s rigor with regards to depicting the current applications of AI by various organizations in a variety of settings, is not matched by a high level of analysis of each case, or of the general trends that emerge from them. Its problematic research method, its apparent lack of interdisciplinary outlook, and its adoption of the business-world narrative regarding AI, severely handicap it, and its ability to get a good read of the social implications and ethical challenges of AI technologies. Therefore, while I found the depictions of each case quite interesting, I found the debates that followed and the conclusions that the authors asked to draw from each case somewhat limited and flawed. With regards to methods, the initial idea of the two authors seems rather solid: to study the application of these AI technologies from the “frontline”; that is from the perspective of Citation: Duke, Shaul. 2022. Title.