Computing Reviews
Today's Issue Hot Topics Search Browse Recommended My Account Log In
Review Help
Search
Ethical machines : your concise guide to totally unbiased, transparent, and respectful AI
Blackman R., HARVARD BUSINESS REVIEW, Cambridge, MA, 2022. 224 pp. Type: Book (1647822815)
Date Reviewed: Dec 16 2022

It has been nearly a quarter-century since Ray Kurzweil considered the emergence of “spiritual machines” [1], and twice that time since Arthur Clarke introduced us to HAL’s murderous insanity [2]. The field of artificial intelligence (AI) endured a “winter” of pessimism, in the 1980s, about its utility and resulting decline in funding; yet today, thanks to numerous and spectacular successes (along with much Hollywood hype), it is now one of the primary areas of massively funded computer science research and extensive commercial application. We are no longer debating whether computers can think and reason like humans, but instead how they do so and if they can be taught to make ethical decisions.

Author Reid Blackman is a well-known author, researcher, chief ethics officer, and practitioner in technology ethics. His new book adds to the growing library of “ethical AI” commentary, speculation, praise for its successes, and warnings about its potential misuse. The book’s subtitle is quite aspirational: your concise guide to totally unbiased, transparent, and respectful AI. Adding to the difficulty of meeting that challenge is the problem of defining what we really mean in human terms concerning ethics, fairness, and trust.

Rather than focusing on prescriptive “AI for good,” Blackman concentrates on “AI for not bad.” That is, for his primary target readership of corporate leaders, how to stay out of AI trouble by understanding and minimizing its risks. He confidently asserts that implementing an AI ethics program is not as difficult as one would suppose, and, as he is also a professor of philosophy, takes both a theoretical as well as practical view of the subject.

The book focuses on three major AI risks--bias, explainability, and privacy--and organizes their discussion around the “content” of an AI ethics program and the “structure” for implementing it. The author ultimately recommends the development of an AI ethical risk program, starting with value statements based on avoidance of “ethical nightmares” and connected to organizational mission or purpose. Readers will need to consider whether his claim that “AI ethics isn’t so difficult after all” is applicable to their unique corporate characteristics. He concludes by noting that the book’s approach and recommendations are not specific to AI and are also useful for general ethics programs.

Humans have been struggling with how to constrain the power of their inventions, especially the “smart” ones. Asimov’s “Three Laws of Robotics” [3] was a hopeful start, and is still debated today. Whether we can tame AI to our admittedly imperfect human ethical standards remains to be seen. Blackman’s book provides some useful guidance.

More reviews about this item: Amazon

Reviewer:  Harry J. Foxwell Review #: CR147523 (2302-0022)
1) Kurzweil, R. Age of spiritual machines. Viking, New York, NY, 1999.
2) Clarke, A. C. 2001: a space odyssey. New American Library, New York, NY, 1968.
3) Asimov, I. I, robot. Gnome Press, New York, NY, 1950.
Bookmark and Share
  Editor Recommended
Featured Reviewer
 
 
Ethics (K.7.m ... )
 
Would you recommend this review?
yes
no
Other reviews under "Ethics": Date
Ethical issues in information systems: a book of readings
Dejoie R., Fowler G., Paradice D., South-Western Publishing Co., Cincinnati, OH, 1991. Type: Book (9780878355624)
Jun 1 1991
Computer ethics
Johnson D. (ed), Prentice-Hall, Inc., Upper Saddle River, NJ, 1985. Type: Book (9780131640054)
Jun 1 1991
Ethical issues in the use of computers
Johnson D. (ed), Snapper J., Wadsworth Publ. Co., Belmont, CA, 1985. Type: Book (9780534042578)
Jun 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