Superintelligence: Paths, Dangers, Strategies

These two fascinating books have much in common. Both, though written by philosophers, are aimed primarily at a non-philosophical audience; both provide state-of-the-art surveys of a wide range of academic and non-academic literatures; each deals with a potentially urgent threat to contemporary lifestyles and values; and each canvasses possible responses to that threat. Anyone interested in the future of humanity will learn much from these books. This review concentrates on what philosophers—and particularly moral philosophers—might learn.

Bostrom explores the possible future emergence of an artificial superintelligence: a machine (or other AI) with an ability to perform general cognitive tasks that ‘is smart in the sense that an average human being is smart compared to a beetle or a worm’ (p....

...