Risks of Artificial Intelligence

If the intelligence of artificial systems were to surpass that of humans, humanity would face significant risks. The time has come to consider these issues, and this consideration must include progress in artificial intelligence (AI) as much as insights from AI theory.Featuring contributions from leading experts and thinkers in artificial intelligence, Risks of Artificial Intelligence is the first volume of collected chapters dedicated to examining the risks of AI. The book evaluates predictions of the future of AI, proposes ways to ensure that AI systems will be beneficial to humans, and then critically evaluates such proposals.The book covers the latest research on the risks and future impacts of AI. It starts with an introduction to the problem of risk and the future of artificial intelligence, followed by a discussion (Armstrong/Sokala/OhEigeartaigh) on how predictions of its future have fared to date.Omohundro makes the point that even an innocuous artificial agent can easily turn into a serious threat for humans. T. Goertzel explains how to succeed in the design of artificial agents. But will these be a threat for humanity, or a useful tool? Ways to assure beneficial outcomes through 'machine ethics' and 'utility functions' are discussed by Brundage and Yampolskiy.B. Goertzel and Potapov/Rodionov propose 'learning' and 'empathy' as paths towards safer AI while Kornai explains how the impact of AI may be bounded. Sandberg explains the implications of human-like AI via the technique of brain emulation. Dewey discusses strategies to deal with the 'fast takeoff' of artificial intelligence and, finally, Bishop explains why there is no need to worry because computers will remain in a state of 'artificial stupidity'.Sharing insights from leading thinkers in artificial intelligence, this book provides you with an expert-level perspective of what is on the horizon for AI, whether it will be a threat for humanity, and how we might counteract this threat.