Artificial Wisdom? A Potential Limit on AI in Law (and Elsewhere)
40 Pages Posted: 12 Mar 2019 Last revised: 22 Oct 2019
Date Written: March 1, 2019
Artificial intelligence (“AI”) may soon perform all tasks that belong to the factual realm more effectively than can human beings. AI may become superior—likely far superior—at describing, predicting, and persuading. Its competitive advantage in these pursuits may well extend to legal and judicial practice. Does that mean human participation in the law will be rendered obsolete? Not necessarily. This Essay suggests three propositions may hold true that would justify an ongoing—perhaps permanent—role for human beings: (1) that moral judgment is necessary for legal and judicial practice; (2) that the first person perspective (or subjectivity) is necessary for moral judgment; and (3) that AI is incapable of attaining the first person perspective. After briefly addressing the first two propositions, the Essay focuses on the third. It explores ways in which the best scientific accounts of various phenomena related to the first person perspective—consciousness, free will, and the unified self—seem incompatible with an internal experience of the first person perspective, particularly when it comes to decision-making. AI seems to be a creature of science, suggesting it too may be incompatible with the first person perspective. If so, while we must recognize the staggering potential of artificial intelligence, there is no similar prospect for artificial wisdom. The need for wisdom—understood here as involving moral judgment—preserves a role for human beings in legal decision-making.
Keywords: artificial intelligence, AI, jurisprudence, moral judgment, free will, consciousness, legal decision-making, wisdom, subjectivity, computers
Suggested Citation: Suggested Citation