Human Compatible: Artificial Intelligence and the Problem of Control Audiobook (Free)

‘The most significant book on AI this year.’ –The Guardian

‘Mr. Russell’s fascinating book will go deep, while dazzling with dried out witticisms.’ –The Wall Street Journal

‘The most significant book I’ve read in quite some period’ (Daniel Kahneman); ‘A must-read’ (Utmost Tegmark); ‘The reserve we’ve all been waiting for’ (Sam Harris)

A respected artificial intelligence researcher lays out a new approach to AI that will enable us to coexist successfully with increasingly intelligent devices

In the favorite about Human Compatible: Artificial Intelligence as well as the Issue of Control imagination, superhuman artificial intelligence can be an approaching tidal wave that threatens not just jobs and human relationships, but civilization itself. Conflict between humans and machines sometimes appears as inevitable and its own outcome all too predictable.

Within this groundbreaking book, distinguished AI researcher Stuart Russell argues that this scenario could be avoided, but only when we rethink AI from the ground up. Russell begins by exploring the idea of intelligence in humans and in machines. He explains the near-term benefits we can expect, from smart personal assistants to vastly accelerated scientific study, and outlines the AI breakthroughs that still have to happen before we reach superhuman AI. He also spells out the methods humans already are finding to misuse AI, from lethal autonomous weaponry to viral sabotage.

If the predicted breakthroughs occur and superhuman AI emerges, we could have created entities far more powerful than ourselves. How do we ensure they never, ever, have power over us? Russell suggests that we can rebuild AI on a new foundation, according to which devices are made to be inherently uncertain about the individual preferences they must satisfy. Such devices would be humble, altruistic, and committed to pursue our goals, not theirs. This fresh foundation allows us to create machines that are provably deferential and provably beneficial.