The AI Dilemma: Uniting Four Logics of Power

Automated systems: We can't live with them and we can't live without them. AI technology in the wrong hands is inarguably dangerous; but in the right hands it can be beneficial to all. This, say the authors, is the AI dilemma. In this excerpt from their book of the same title, they summarize four 'logics of power' that must be integrated in order for humanity to benefit from AI: engineering logic, social-justice logic, corporate logic and government logic. They show that AI systems consistently raise significant, complex problems that will not simply resolve themselves, and that having diversity of thought and perspective weighing in on these matters is incredibly important. If we want trustworthy AI systems, they conclude, we need to bring all four perspectives together, keep them in mind simultaneously, and make the effort to understand why others feel and think the way they do.