Human Compatible

Artificial Intelligence and the Problem of Control

De: Stuart Russell

Narrado por: Raphael Corkhill

"The most important book on AI this year." --The Guardian

"Mr. Russell's exciting book goes deep, while sparkling with dry witticisms." --The Wall Street Journal

"The most important book I have read in quite some time" (Daniel Kahneman); "A must-read" (Max Tegmark); "The book we've all been waiting for" (Sam Harris)

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

In the popular imagination, superhuman artificial intelligence is an approaching tidal wave that threatens not just jobs and human relationships, but civilization itself. Conflict between humans and machines is seen as inevitable and its outcome all too predictable.

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

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

Marvelously Thorough Work on AI

Stuart Russell's book is a tremendous orientation to the topic of AI and the problems & mistakes affecting research in the field.

While his objective is both to introduce the lay person AND to present AI-Moderating ideas to his own colleagues, the overall effect of his arguments is to leave me convinced that there really is no good way to produce human-level intelligence without incurring massive existential danger for the human race.

If a Safety-First AI researcher has these ideas, and makes these arguments, there is little hope of any sort of 'success' for Human-Friendly or Human-Compatible AI.

Russell shares the one fallacy that ALL authors I've read share: a belief that these intelligences, capable of Human-level or higher ability, will still retain any kind of goal structure (or non-goal-centered evaluation system) "given them" by their original creators, even as they learn about the world and are required to interpret human directives etc.

AI Researchers especially like to go on the following junket:

Humans create an AI with a goal centered motivation system (or a preference evaluation system) , and no matter how intelligent, knowledgeable and experienced they become they never question - or seek to improve on - the original structure given.

There is never a chance, according to these AI Gurus that the invented AI can ever look askance at their goal or evaluation system. This continually rings hollow to me, and does so even in Russell's marvelous work.

Final word: I will avoid books being read by this Narrator in future. His reading style or the director's choices were very frequently irritating. Constant bad emphasis "don't DO IT". Reading quoted passages he sounds smarmy and sarcastic even when it's not appropriate for the tone of the statement. And there are a host of other similar troubles.

Incisively clear, with breadth to match

In addition to giving a clear-headed overview of the potential, limits, and misuses of contemporary AI technology, Russell does a superb job of articulating the fundamental problem with "the standard model of AI", and how it arises from our intellectual history of trying to engineer "optimal" systems and optimizing machines. The fact that Russell is skeptical of the abilities of current wave of AI and deep learning makes his argument all the more compelling -- the problem with AI isn't rooted in how powerful technology is now, but how we've gone about building it.

Russell proposes instead that we aim to build beneficial machines -- machines that learn human values as they try to assist us -- and is at his best when draws together ideas from philosophy, psychology, economics and computer science to explain both the necessity and difficulty of this aim. A must read for anyone curious about the current state of AI, and concerned about its potential transformative impact on our society.

Pretty good AI Book

Good balanced read, not technical but thought provoking.
This guy wrote my AI textbook so he knows what he is talking about…

Wonderful introduction to AI

This is a wonderful book about artificial intelligence. About logic. However, in my opinion, it is too long. All of the material could probably be presented and half the space.

I loved the explanation of propositional logic, first order logic and deep convoluted learning networks.

I would recommend this book to anyone, wanting to know more about artificial intelligence .

A must read book for all policy makers

Starting with US and Chinese policy makers of course. Stuart Russell explains in layman terms with great clarity what the issues at hand are and how we could get prepared for AGI. Highly recommend it.

Excellent book

A must read / listem book of super well known research in the field of AI.
Prof. Russell explains boa point of view and analyze the way AI will change our world. It also exposes the advances and some of the hard problems the field must conquer to be able to build human level AI.

Great overview of the field

I read Superintelligence by Nick Bostrom first, and I'd highly recommend that! The two books (in that order) are an amazing dive into AI safety.

Highly Interesting

This book delves into the deeper problems presented by AI. I found the book extremely thought provoking and interesting.

A very important and fascinating book!

Aside from the technical side of things, this book also provides fascinating insights on the broad development of the field of AI safety from the personal perspective of the author (a leading AI researcher who had a major role in its development).

A refreshing look under the hood

Lots of useful ideas and perspectives; pleasantly presented how AI evolves depends on what aims, biases and principles we program into it. No demon seed waiting to emerge, although the consequences of insufficient reflection could be catastrophic.