While military planners have a robust history of using exercises, wargames, and simulations to identify capability gaps and inform doctrine, translating this data into actionable, long-term change remains a critical challenge for the nuclear enterprise. In the contemporary landscape, many of these simulations are either "white celled" to prevent escalation or terminated as soon as nuclear hostilities begin, limiting their ability to realistically modernize doctrine for a dynamic, multi-actor arena. To close the gap between lessons identified and lessons implemented across both 8AF and 20AF, what specific systemic bottlenecks, manpower constraints, or organizational culture barriers must be overcome? What models exist to enhance nuclear-focused exercises and simulations for tomorrow's challenges—drawing on historical lessons of interwar wargaming realism and assumptions—while ensuring that operational deterrent forces remain on-alert to face today's threats? Ultimately, how can the nuclear enterprise leverage automation or AI-enabled analysis to capture recurring observations, streamline the lessons-learned pipeline, and foster genuine organizational learning over legacy habits? References: Playing War by John Lillard; Rhetoric and Reality of Air Warfare by Tami Davis Biddle; Strategy in the Missile Age by Bernard Brodie
- Keith, Andrew J., "Alignment: National Security Objectives in Cold War Computer Simulations," SAASS thesis, 2025, 117 pgs.
- Keith addresses this by exploring the RAND Strategy Assessment System (RSAS), a highly sophisticated automated political-military wargame developed in the 1980s to analyze nuclear forces, strategic balance, and escalation control. RSAS utilized rule-based artificial intelligence—represented by automated decision-makers known as "Sams" (United States) and "Ivans" (Soviets)—to simulate global nuclear warfare, crisis decision-making, and deterrence. However, Keith highlights that while such models hold vast potential to enhance nuclear-focused wargames, their historical implementations faltered because the AI's complex behavioral rules became too opaque for government users and military staff to interpret or control, ultimately leading to their abandonment as practical decision-making tools.