In the mass consciousness, superhuman artificial intelligence is a technological tsunami that threatens not only the economy and human relations, but also civilization itself. The conflict between people and machines is seen as inevitable, and its outcome is predetermined. Prominent AI researcher Stuart Russell argues that this scenario can be avoided. In his groundbreaking book, the author reveals how humans have already learned to use AI, ranging from lethal autonomous weapons to manipulating our preferences, and what else they can teach it. If this happens and superhuman AI emerges, we will be faced with an entity much more powerful than ourselves. How to ensure that a person does not become subordinate to a superintelligence? For this, Russell believes, artificial intelligence must be built on new principles. Machines should be modest and altruistic and solve our problems, not their own. The reader will learn about what these principles are and how to implement them from this book, which the most authoritative publications in the world called the main book about artificial intelligence. Everything that civilization can offer is a product of our intellect; gaining access to vastly superior intellectual capabilities would be the greatest event in history. The purpose of this book is to explain why this may be the last event of civilization and how we can prevent such an outcome. The introduction of the concept of utility - an invisible property - to explain human behavior through mathematical theory was stunning for its time. Moreover, unlike monetary amounts, the value of various bets and prizes in terms of utility is not available for direct observation. The first who will really benefit from the introduction of robots into the home will be the elderly and infirm, for whom a useful robot can provide a degree of independence that is unattainable by other means. Even if a robot performs a limited range of tasks and has only a rudimentary understanding of what is happening, it can be very useful. Obviously, the actions of loyal machines will have to be limited by rules and prohibitions, just as the actions of people are limited by laws and social norms. Some experts suggest absolute responsibility as a solution.