I 
know what you’re thinking: another book about AI, Callan? Really?
Yes, really. I don’t know what compelled me – but perhaps that the author’s name was so similar to my own spurred me onward. Surviving AI is Calum Chase’s summary of the current state of affairs with AI risk management (specifically, of course, relative to artificial superintelligence). It’s a well balanced book in that it plainly states where there is agreement and disagreement by experts in the field of artificial intelligence. It’s concisely and clearly written (in contrast to the jargon-dense prose Nick Bostrom deploys in Superintelligence) and would serve adequately as an introduction to the topic without any histrionics or undue fretting. It’s a sober book. That said, I don’t think I gleaned too much new from reading it: this isn’t my first time reading a primer on AI. I’d send neophytes instead to James Barrat’s Our Final Invention as the best place to dip into the topic for the first time. They cover much of the same territory, but while Chase’s book is leaner due to a more spare writing style and less indulgence in personal anecdotes, it lacks some of the vitality and passion that made Barrat’s book so powerful. I guess the one thing I’d say I got from this book was Chase’s clear articulation of the tremendous uncertainty about almost every aspect of this subject. That uncertainty shouldn’t stop us: it should motivate us to learn more.
Both authors emphasize the same conclusion though: We really need to be talking about artificial superintelligence as a society, as a species, so we’re sure we’re doing all we can to be prepared for it when it arrives. And for us to talk about it, we all need to get up to speed on the topic. So: please pick one of these introductory books up and start reading!