A review by thebobsphere
Superintelligence: Paths, Dangers, Strategies by Nick Bostrom

4.0

 In cinema the depiction of Artificial Intelligence (A.I.) killing mankind is a common trope; There’s HAL from 2001: A Space Odyssey , the unforgettable Ultron from the Marvel Cinematic Universe even Pixar had a shot at it in WALL-E with AUTO. Clearly Hollywood loves it when a robot outwits it’s creators.

In a way Nick Bostrom also agrees that one day A.I. will take over the universe but not in the way we’ve seen in films.

Superintelligence is about how A.I. has become a part of society. In gaming, computers can outwit a human, in certain mechanical tasks A.I. can outdo a human as well. Bostrom does predict that there will be a time when robots and computers will dominate society completely. However this is not Bostrom’s worry.

Bostrom argues that with such rapid development with A.I. it is possible that it will wreak its revenge and enslave society. I know that this sounds like the premise of a science fiction film but Bostrom warns that cinema’s way of displaying this is nonsensical. He does state, though, that the possibility is there. He then gives a detailed plan on how by limiting our resources A.I. can dominate mankind.

The final part of the book deals with problem solving. How do we stop A.I. from becoming a menace? Is it possible? Does A.I. have any limitations? and can we live in peace with A.I.?

Superintelligence is fascinating. Yes, there are plenty of nightmare scenarios but the arguments that Bostrom presents are clear and fair so this is not a ‘doom and gloom’ book. As for someone who knows absolutely nothing about this topic, I learnt a lot about the development of A.I. and the current role it has in society. For the more hardcore technophiles, there’s a detailed notes section where Bostrom expands on the mathematical aspects his theories so there’s something for everyone. Superintelligence: Paths, Dangers, Strategies is a great primer for someone who is curious about A.I. but does not want to be bogged down with too much scientific jargon.