Human morality is a set of cognitive devices designed to solve social problems. The original moral problem is the problem of cooperation, the “tragedy of the commons” — me vs. us. But modern moral problems are often different, involving what Harvard psychology professor Joshua Greene calls “the tragedy of commonsense morality,” or the problem of conflicting values and interests across social groups — us vs. them. Our moral intuitions handle the first kind of problem reasonably well, but often fail miserably with the second kind. The rise of artificial intelligence compounds and extends these modern moral problems, requiring us to formulate our values in more precise ways and adapt our moral thinking to unprecedented circumstances. Can self-driving cars be programmed to behave morally? Should autonomous weapons be banned? How can we organize a society in which machines do most of the work that humans do now? And should we be worried about creating machines that are smarter than us? Understanding the strengths and limitations of human morality can help us answer these questions.
- 2017 Festival