This is a must-read for understanding the Turing test, its origins, how it has evolved over the years, and its linkage to the broader field of artificial intelligence (AI). It is a nontechnical book for undergraduates and postgraduate students of various disciplines, including computer science (CS), philosophy, psychology, and engineering. Due to its descriptive nature, it is also an ideal companion for those interested in the ideas of machine thinking and intelligence.

The volume starts with a brief introduction to Alan Turing as a man, his early life, and the years he spent as a researcher, motivating why he is considered the father of AI. Through a brief overview of his life, his importance at Bletchley Park during the Second World War, and his private life at the Ratio Club, the book is then divided into two parts.

The first part begins with a description of the imitation game, its origins, and its linkage with machine intelligence and thinking. The Turing test is then formalized and contextualized in the broader field of AI. Concepts such as rationality, cognition, and natural language processing are briefly introduced and the reader is provided with notions related to classical and modern AI.

It then moves to a presentation of the controversies surrounding the imitation game, including cultural objections and thinking about thinking.

The book then focuses on the history of conversation systems, from the earlier Eliza approach to the modern Eugene Goostman approach. Through a review of these systems, the authors provide readers with important hints and suggestions on how to design artificial conversations.

Eventually, the first part concludes with a description of the matters arising from early Turing tests, stressing the increasing importance of designing and deploying conversation systems for counterattacking recent conversational malwares and cyber crimes aimed at defrauding people. These matters also give food for thought on what is actually being measured by the Turing test, whether intelligence or a type of human intelligence.

The second part of the book is devoted to the presentation of a set of experiments conducted by the authors to assess whether a set of candidates’ conversation systems were able to pass the Turing test. These experiments, conducted respectively at the Reading University (2008), Bletchley Park (2012), and the Royal Society in London (2014), involved a large number of human participants filling the roles of both interrogator and hidden human foil. The chapters in Part 2 include transcripts from the conversations that occurred in the designed experiments between humans and machines. These conversations were important to demonstrate in practice the difficulties surrounding current conversational approaches to mimic humans. These difficulties, and the solutions designed to tackle them, were also made explicit by interviewing the software developers of those conversation systems that took part in the experiments. The software developers not only shared important tips on how to design conversational solutions, but also their points of view on the Turing test, its utility to the field of AI, and whether machines can think. The book concludes with a provocation on how the Turing test could be extended and how this extension could be regarded as a test for the future of humanity.

More reviews about this item: Goodreads