[2E5-OS-4b] OS-4

[2E5-OS-4b-01] The Merits and Side Effects of Mechanising Morality

〇Minao Kukita¹ (1. Nagoya University)

Keywords:artificial moral agent, dual-process theory, moral philosophy

The creation of artificial agents capable of making moral judgments or taking moral actions has sparked a variety of debates. This presentation focuses particularly on the argument that ``artificial moral agents could potentially be more moral than humans because their judgments are not influenced by emotions.'' It is generally believed that human decision-making involves two processes: an intuitive and fast process, and a deliberative and slow process. The argument for the advantages of artificial moral agents mentioned above is based on the premise that moral judgments should primarily engage the deliberative process. This premise aligns with the dominant view in the tradition of Western moral philosophy, which emphasizes the importance of suppressing emotions and relying on reason in moral matters. However, attempting to replace intuitive processes with purely deliberative, mechanical processes in morally significant decisions may have unintended consequences. This presentation examines this issue by analyzing real-world cases where artificial intelligence is employed to make decisions with morally significant consequences.