Scientists, representatives of information technology companies, and politicians are increasingly recognizing the opportunities and dangers of machine intelligence for the quality of people’s lives. They therefore repeatedly call for the use of technology for the benefit of people, but do not say what constitutes people’s well-being. Even happiness research provides hardly any guidance for a desirable use of the technology. Evolution controls us according to the principle of homeostasis through happiness and unhappiness. Humans strive for actions that lead to perceptions with positive feelings and avoid actions that cause suffering. Evolution, on the other hand, is aimed at progress, not at human happiness. Quality of life, i.e. happiness and unhappiness, is a relative and fleeting quantity: We measure ourselves against our demands, compare ourselves with our peers, and accommodate to both positive and negative situations. Hope seems more important than the actual achievement of goals. Since Plato, people have been striving not only for short-term satisfaction of their needs (hedonia), but also for lasting satisfaction with themselves and the environment (eudaimonia). As the rapidly growing datasets increasingly contain indicators for the well-being of people, data collections and pattern recognition make the quality of life more and more objectively measurable. Even if our knowledge about the use of machine intelligence for the benefit of mankind is still very modest, it would be grossly negligent, in view of the enormous changes, not to use the existing knowledge for the benefit of mankind.