Culpability

Culpability explores the dilemma of autonomous AI systems operating in human society: Can AI ever be culpable for its decisions and the impact they have in the real world? Lorelei’s book, Silicon Souls, emphasizes the fact that AI are human-made machines; therefore, there is always an element of human responsibility for the machines’ behaviors: “Is a machine responsible for everything it does? Of course not. It is we who are responsible for the consequences of the very freedom we grant these objects of our creation” (246). Lorelei argues that AI must be programmed with morality in mind, as the way the algorithms calculate risk will have material consequences for human life. The excerpts from her book include the kinds of philosophical dilemmas that engineers like herself program into their algorithms. For example, the trolley problem “illustrates a common dilemma for ethical input in AI systems” (34), as it asks the AI to determine the path of least harm. AI systems must sometimes literally choose between saving one life over the other, so the programmer needs to train the machine on the most moral decisions to make. Lorelei feels responsible for the car crash and the Yemeni drone strike—and any future malfunctions with NaviTech vehicles—because the system uses her algorithm, making her morality the basis for its decision-making.