Ethics may save us from the possibly negative effects of artificial intelligence (AI). The author, Stahl, leads the SHERPA consortium, whose work is the basis of this book. Project Sherpa (https://www.project-sherpa.eu) stands for ”shaping the ethical dimensions of smart information systems.”
First, AI is surveyed from the perspective of how it is viewed by society rather than what AI is actually capable of doing. For example, the author does not believe that a general AI like HAL is possible; however, he does believe that the possibility of a HAL-like system influences society.
Ethics is presented from an ecosystem perspective for “human flourishing.” Four ethical principles of AI are developed: respect for human autonomy, prevention of harm, fairness, and explicability. This leads to seven key requirements: human agency and oversight; technical robustness and safety; privacy and data governance; transparency; diversity, non-discrimination and fairness; social and environmental wellbeing; and accountability.
SHERPA project findings include 39 ethical issues that were derived from interviews with experts. These were categorized into three areas: issues arising from machine learning; living in a digital world; and metaphysical issues. The three ways these issues can be addressed are discussed: policy level, organizational approaches, and guidance for individuals and groups.
Current efforts are identified and classified into categories in a very interesting chart. The Collingridge dilemma is discussed: it is easier to address issues early in the life cycle, but consequences are not clear; when consequences become clearer, changes are harder. Finally, the basics of AI ecosystems for human flourishing are discussed and recommendations are presented.
The book is very current and well written. Every chapter has extensive references.