In this paper, the author applies philosopher Thomas Nagel’s distinction between agent-relative reasons and agent-neutral reasons to the challenges associated with a U.S. artificial intelligence (AI) strategy. In Nagel’s formulation, a reason is agent-neutral if it can be stated in a general form without any essential reference to the person who has it (e.g., the badness of suffering gives everyone a reason to reduce suffering). A reason is agent-relative if its general form necessarily includes such a reference, such as when a duty arises from one’s particular role, relationship, or commitment.
Nagel’s account is well suited to the challenges involved in implementing a U.S. AI strategy: He argues that both kinds of reasons—those arising from an objective, impersonal standpoint and those arising from the standpoint of a particular agent—generate legitimate demands on action; he also argues that practical reasoning cannot eliminate the tension between the two reasons.
The author argues that national leaders face this exact tension in a strategist’s dilemma: Leaders bear agent-relative obligations to protect their own citizens, yet transformative AI generates agent-neutral obligations to prevent grave harms to anyone, regardless of nationality. Rather than treating these as a single balance to be struck, the author argues that catastrophic and irreversible risk should constrain what counts as a permissible strategy and that agent-relative trade-offs properly arise only within that constraint.
This publication is part of the RAND expert insights series. The expert insights series presents perspectives on timely policy issues.
This document and trademark(s) contained herein are protected by law. This representation of RAND intellectual property is provided for noncommercial use only. Unauthorized posting of this publication online is prohibited; linking directly to this product page is encouraged. Permission is required from RAND to reproduce, or reuse in another form, any of its research documents for commercial purposes. For information on reprint and reuse permissions, please visit www.rand.org/pubs/permissions.
RAND is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND’s publications do not necessarily reflect the opinions of its research clients and sponsors.
