Guardrails

Oxford philosopher Paula Boddington delivers a reference-grade volume that is timely, helpful, and clear.

This is the first book about codified ethics for AI. There’s a general belief that ethics are essential in guiding how such a powerful yet unknown technology will evolve. And while directed to AI, many of the principles in this book will suit other domains.

Boddington applies real-world practicality and how to use ethics as guardrails instead of tablets. She is a philosopher based in the Oxford computer science faculty. The tugs between philosophical and applied approaches strengthen every chapter.

Near the end she shows how her thinking works in a real situation. She analyses a code developed by the Future of Life Institute, an advocate for ‘beneficial AI.‘ Members developed a code to inform AI development. Boddington’s critical assessment shows how creating a code with a homogeneous group it not always optimal. The analysis presents a very useful case study of her approach.

The table of contents doubles as a road map to the moral issues we face with intelligent machines. This can be a go-to resource after first reading.

SEE ON AMAZON.COM
we do not receive a referral fee

Towards a Code of Ethics for Artificial Intelligence
Springer | 2017 | by Paula Boddington

Review by Andrew Cochran | November 2018