Abstract
There is no need to describe the huge amount of research in the past several decades in the field of communications. However, many communication scholars are also probably aware, if only from afar, of the growing flood of scholarship in artificial intelligence (AI) research. With a few exceptions, however, these two academic fields have not found the path to combining their respective research—and certainly there is no published textbook with that nexus as its focus. Gunkel’s An Introduction to Communication and Artificial Intelligence is the first to attempt that bridge.
It should be noted at the outset that the book is heavier on AI than it is on Communications, precisely because it “is designed to provide students in the field of Communication Studies with the basic knowledge, insight, and skill to make sense of emerging technology, specifically AI, robots, and algorithms” (p. ix). To be sure, it does devote some space to how communications theory and research can advance AI studies and vice versa; nevertheless, this textbook is more a primer on AI than it is on Communications (the latter is not something missing from the publishing world).
Perhaps unavoidably, such a book emerges as a hybrid affair. On the other hand, its format and style are classical textbook. It starts with three introduction/orientation chapters (concepts, terms, operational principles of AI), then moves on to various AI applications, and concludes with AI’s social ramifications and problematic issues. As a coda to the book, Gunkel offers several “maker exercises” that are designed to offer the student some “hands-on experience” in coding (with step-by-step instructions). Moreover, each chapter commences with a short “Key Aims & Objectives” and ultimately concludes with a “summary of key points,” a useful reiteration of that chapter’s basic ideas that unfortunately turns out to be quite necessary, given the highly associative, non-linear, and variegated exposition in most chapters. They combine historical background, abstruse theory, technical aspects, story vignettes, and thought-provoking speculations or conclusions. This substantive hybridity is a function of Gunkel’s attempt to address all student audiences—from first-year B.A. to advanced M.A./PhD students. For instance, spread around the book are somewhat abstruse equations—not at all undergraduate material. Conversely, the vignettes, and to a limited extent the historical background, are useful for freshmen/sophomores who need some historical perspective in general and certainly on the topic of AI.
However, even here, Gunkel overdoes it on occasion. One example among several: the chapter about machine translation offers a sub-chapter entitled “historical context,” where the student has to plow through: the biblical Tower of Babel story (almost two pages long); the New Testament’s “Pentecost” event; Douglas Adams’ “Babel fish” in the Hitchiker’s Guide to the Galaxy; Star Trek; a 1657 book mentioning the “Universal Character” (and another four books from the 17th century); Leibniz’s universal symbolism; the Warren Weaver memo (1949; at least here we are in the modern era); and several other “asides,” such as “Translation = Cryptography.” Similar meandering occurs in the section on Natural Language Processing (pp. 137-139). Graduate students might find these enlightening (faculty too!); undergraduates will certainly perceive them as tedious or at least overly pedantic.
From an AI standpoint, this book is quite comprehensive. In addition to basic concepts (Chapter 3) and Machine Translation (Chapter 4), it also delves into Natural Language Processing (Chapter 5), Computational Creativity (Chapter 6), and Social Robots (Chapter 7) and then concludes with AI-related Social Issues (Chapter 8) and Social Responsibility/Ethics (Chapter 9). Overall, the author’s arguments and points are convincing, with an infrequent “lacuna” or exaggeration. Here are two examples. First, Gunkel writes that “[c]ommunication has been central to the definition and demonstration of machine intelligence from the very beginning” (p. 44). Given the intended audience of communication scholars, this sounds attractive, but misses two other “central” aspects of AI: “creativity” (art, music, math proofs, etc.) and “problem-solving” (e.g. data analysis, pattern recognition). Indeed, as just noted, an entire chapter is devoted to “Computational Creativity” that threatens future mass unemployment of journalists, among other things. Even if AI would never be able to communicate in quasi-human fashion, it would be hugely useful to humanity—which is not to negate the importance of AI communication (right, Alexa?).
A second lacuna I find comes when Gunkel argues (correctly) that communication theory has to go beyond its previous approach of viewing the computer (or AI) as merely a medium (McLuhan’s “extension of man”) for human-to-human message transmission and start studying it as an object with which we can directly communicate. However, his theoretical basis for this is Kuhn’s concept of “normal [paradigm] science” (p. 49), an incorrect (or least, misleading) use of the term. “Normal science” involves theorizing about objective reality (e.g. atoms) that behave in quite predictable and highly stable fashion (quantum mechanics, notwithstanding); for example, if you get the movement of planets correctly, they won’t change much (if at all) for millennia or “forever,” so that a theoretical construct can be derived for such planetary action. Technology, however, is constantly changing in unpredictable fashion; thus, it cannot be considered to involve “normal science.” This does not mean that studying technologies (AI among them) is a useless or even necessarily a Sisyphean affair; it does mean that such a scholarly endeavor must work quite differently from “normal science.”
All in all, this is a very useful textbook on an undertaught subject. Its major strength is also its central drawback (more stylistic than substantive): a hybrid mixture of lofty theorizing and interesting philosophical ruminations, alongside down-to-earth “how-to” explanations of AI’s nitty-gritty workings, and its potential consequences. For professors who wish to broaden their students’ horizons regarding a field with growing influence on communication studies, this will entail homework to decide which pages are relevant for the level of their respective students.
In short, David Gunkel has done an excellent job here for a first-time textbook on this complex subject; hopefully, his next endeavor along these lines will produce two textbooks: one for undergraduates and another for MA/PhD students and communication scholars unfamiliar with this increasingly important field.
