Our Final Invention

American author James Barrat weighs the possible benefits against the risks of artificial intelligence at or beyond human level, risks he says include the end of the human race. The book starts by explaining how a general AI could improve itself repeatedly until it became a superintelligence, then traces the history of the field, including I. J. Good's work, up to today's researchers. Barrat mixes explanations of AI concepts with profiles of technologists, industry watchdogs and AI systems, and with interviews with figures such as Eliezer Yudkowsky and Ray Kurzweil. His tone stays cautionary, stressing how hard it would be to control or predict a mind far smarter than any person. A later second edition adds a new preface.

Elon Musk said it is also worth reading on the subject of AI risk.