★★★★★ 5
Gripping and Informative, a Must-read
Format: Hardcover
As someone who struggles to finish books in their entirety, I found Our Final Invention by James Barrat highly readable, deeply informative, and utterly gripping. The book contains a powerful message: through competition, distrust, desire and curiosity, humans will inevitably create an artificial intelligence (AI) that rivals or surpasses our own. Thus, it is wise and necessary to invest now in mitigation efforts and potential safeguards -- increased research and advocacy for AI risk and, most importantly, producing friendly AI.
Barrat covers a lot of ground, but his main argument is summarized as follows: Currently, we humans regularly utilize narrow AI technology (technology capable of achieving specific, programmed goals through unassisted human computing -- Siri, Google search, IBM's Watson, etc). We are also experimenting with "black box" tools and techniques (programs where inputs and outputs are understood and measurable, but the processes in between aren't -- genetic algorithms/programming and software that writes better software) and artificial neural networking (ANN), as seen through efforts to reverse engineer the human brain. And, below the surface, there is an ongoing race between world powers (driven mainly by national security, defense, and international business interests) and guided by AI developers to develop and achieve artificial general intelligence (AGI) -- human-level artificial intelligence. The problem is that once AGI is achieved it will be very difficult to manage, and may very well result in the manifestation of artificial super intelligence (ASI) -- greater than human-level intelligence.
ASI could theoretically become thousands of times smarter than the smartest human being alive. It won't think like us, won't want to be ruled by us, and, most crucially, it won't want to be turned off. In fact, ASI would likely regard us as potential fuel for its quest to duplicate and improve itself exponentially in order to achieve its goals.
Throughout the book, Barrat refers to interesting psychological phenomena and concepts (such as the normalcy bias), while drawing on personal experiences, historic events, and interviews with computer programmers, inventors and philosophers, to tactfully illustrate how progress in AI development is dangerously rapid. Adequate checks and balances are not in place to deal with a non-ideal intelligence explosion or hard take-off (AGI quickly leading to ASI).
I highly recommend this book to anyone interested in learning about both human beings and the advancement of machines. I suspect that the prominence of AI, as a research field and topic for discussion, will only increase in time (it already has in recent years -- drones, smart technology, Wall Street high frequency trading (HFT), financial modeling), making Our Final Invention a valuable guide or stepping stone for anyone trying to understand our world and the path of the future.
WAS THIS REVIEW HELPFUL?YesReportShare
Reviewed in the United States on July 3, 2014
