Frankenstein's Children: Artificial Intelligence and Human Value

Abstract

Artificial Intelligence gives rise to unique ethical issues as humanity approaches the final extrapolation of a technology which simulates human cognitive processes in a fundamentally nonhuman medium. Simulation AI, taken to its limit, will produce conscious machines. I argue that computer consciousness is not science fiction, but rather that conscious machines are conceptually and, ultimately, technically feasible. Because such a machine is conscious and created by humans, we stand in a unique ethical relation to it, complicated by the possibility that a machine consciousness does not share many of our human values and goals. On balance, our Promethean ambitions are likely both unethical and unwise. For these and other reasons, we should not seek to create fully sentient AI.

Keywords: Artificial Intelligence; AI; ethics; consciousness; value