Tzonis attempts to introduce some of the concepts of artificial intelligence through the medium of a story. The story follows the hero Hermes Steganos through a net of international murder and intrigue during which he attempts to discover the location of an ancient artifact, possibly one of the first computers invented, and at the same time clear his name of his uncle’s murder.

The book in itself should only be used as either a light introduction to the concepts of AI or an introduction to its possible uses. It should not be taken as a text for someone who must become heavily involved in the subject. If the book is taken in this light, the storytelling approach to introducing the basic concepts of AI is both useful and interesting enough to ensure that the reader is not likely to fall asleep in the middle. The book also raises some interesting points about the connections between the seemingly different areas of AI, archaeology, and cryptology, which might give the reader some pause for thought.

As stated before, the storyline follows the exploits of Hermes as he tries, both through the use of reasoning and through the use of some knowledge mapping techniques, to determine who stole the “Golden Thinking Machine” and murdered his uncle. The story is set in Greece and in itself is a witty, intelligent detective story in which the modern-day detective is a professor who has only his knowledge of AI techniques to clear his name and discover the culprit who framed him for the murder. Steganos becomes involved more deeply in the world of international secrets and dark rooms as one interested party after another comes to light as a possible suspect, making the actual ending of the book all the more unexpected.

As the plot unfolds, the reader follows with Hermes the processes and logic needed to set up an AI system. The reader is introduced to the problems inherent in trying to capture knowledge and the near-impossibility of trying to map out the seemingly easy process of logical reasoning. The reader is taken through models that can help in the process of capturing this knowledge, such as semantic nets and knowledge bases, but never far away is the idea that these are only biased techniques, which can only help to capture segments of information but never encompass the whole reasoning process. By the end of the book, the reader is left in no doubt that no magical theory or knowledge mapping technique can capture all the reasoning processes and knowledge involved in a simple logical human deduction. It is arguable that this may be the most valuable lesson for the newcomer to AI.

This interesting and funny detective story helps to introduce the reader to both the concepts of AI and the many pitfalls that can be encountered in trying to implement it. In this vein, the book provides a glimpse into the methods and techniques used and aims the reader in the right direction if he or she seriously intends to research the subject further.