Computing Reviews
Today's Issue Hot Topics Search Browse Recommended My Account Log In
Review Help
Search
The book of chatbots: from ELIZA to ChatGPT
Ciesla R., Springer International Publishing, Cham, Switzerland, 2024. 159 pp. Type: Book (9783031510038)
Date Reviewed: May 13 2024

The relatively sudden emergence of chatbots and artificial intelligence (AI) tools accessible to the general public has caused no little interest in their variety, how they work, how they may benefit people, and the potential risks in using them. This relatively short book attempts to answer these questions with different degrees of success.

There are eight chapters in this book, with a preface preceding the table of contents. The topic of the first chapter is nominally the Turing test--the classic test of whether a computer system can appear indistinguishable from a flesh-and-blood human being. However, this chapter brings in a bit more, including a synopsis of Alan Turing’s work and comments on Kurt Gödel’s theorems. Chapter 2 delves into the technical bases of AI, including neural networks, natural language processing (NLP), Markov models, and linguistics. Since many (if not most) chatbots and AI products are large language models (LLMs), the importance of linguistics and textual analysis cannot be overestimated. The third chapter covers the “classic” chatbots of decades ago, that is, ELIZA, PARRY, and Jabberwacky. These first attempts at creating a virtual entity with which a person could communicate incorporated many of the features of contemporary chatbots that are now much more elaborately implemented.

The fourth chapter is the longest in the book. It has a long catalog of chatbots now in use other than those with uses in specific application areas. Included among them are familiar ones like ChatGPT and Bard, as well as everyday devices like Siri on Apple products and Alexa on the Amazon Echo Dots. They all are intended to interact with the user using a human voice and some personality. The fifth chapter deals with medical applications. Healthcare has special constraints arising from privacy and security restrictions, liability, safety, and the triage of patients. Chapter 6 discusses commercial applications: chatbots for ordering, customer service, and marketing. We have all experienced the pop-up chatbot on a screen as we look at a website for a particular product or have a question that needs to be answered. Some of the same issues concerning account privacy or credit/debit card security encountered with healthcare arise in ecommerce. Money is at stake.

We know that AI tools and chatbots can be hijacked by villains. Chapter 7 collects all of the nefarious uses of AI in one place: phishing, deep fakes, ransomware, plagiarism, privacy, and extremism. The author notes that training chatbots, both initially and in ongoing updates, consumes a significant amount of energy. The topic of the last chapter is artificial general intelligence (AGI), which we do not have yet but could be emergent. Will it affect human society negatively? If so, how can it be contained? Can we pull the plug on it if necessary?

This is a fairly short book that attempts to do many things, some of them well and some not so well. The book seems to target the general reader. For me, the most positive aspects of it are its history of chatbots, overview of linguistics, the general catalog of chatbots in use, the medical and ecommerce applications, the subversion of AI for mischief, and the reflection on AGI. The weaker parts are the technological discussions (including the missing figure illustrating a neural network with hidden layers). There is enough to tantalize readers, but not enough to avoid baffling them. The discussion of “hallucinations”--when AI generates nonsense at best or errors at worst--could be much stronger and include the issue of explanation. Chatbots seem to be black boxes generating garbage with little evidence of where things went wrong. When AI produces artificial stupidity, human smarts have to take over.

Reviewer:  Anthony J. Duben Review #: CR147762
Bookmark and Share
  Reviewer Selected
Featured Reviewer
 
 
General (I.2.0 )
 
Would you recommend this review?
yes
no
Other reviews under "General": Date
Artificial experts: social knowledge and intelligent machines
Collins H., MIT Press, Cambridge, MA, 1990. Type: Book (9780262031684)
Apr 1 1991
Catalogue of artificial intelligence techniques
Bundy A., Springer-Verlag New York, Inc., New York, NY, 1990. Type: Book (9780387529592)
Aug 1 1991
Knowledge and inference
Nagao M., Academic Press Prof., Inc., San Diego, CA, 1990. Type: Book (9780125136624)
Oct 1 1991
more...

E-Mail This Printer-Friendly
Send Your Comments
Contact Us
Reproduction in whole or in part without permission is prohibited.   Copyright 1999-2024 ThinkLoud®
Terms of Use
| Privacy Policy