Joint Force Quarterly
Abstract
Artificial intelligence technologies have reached a tipping point after decades of development. They are diffusing widely across defense and national security applications. Twenty-first century warfighters rely on analytic models in all systems, at all echelons, and in all domains. As more powerful models built on ever larger data sets become ubiquitous, militaries are in a new competition to deploy artificial intelligence. Operational art must embrace “analytic superiority.” This is the operational advantage from collecting and ingesting data, building robust models and computing infrastructure, deploying the models into operational systems, and denying adversaries' ability to do the same
This article explains how the warfighter can bolster combat power by incorporating the most advanced research on data and analytics into doctrine and practice. It provides an enterprise-level framework that identifies the many steps involved in moving from data to operational value. Adopting this framework can help overcome common reasons why most machine learning and artificial intelligence projects fail. The article emphasizes the importance of developing and implementing an analytic strategy aligned to mission priorities. It urges leaders to anticipate surprise for there is no one way to achieve analytic superiority.
Included in
Artificial Intelligence and Robotics Commons, Databases and Information Systems Commons, Defense and Security Studies Commons, Military History Commons, Peace and Conflict Studies Commons, Public Policy Commons, Theory and Algorithms Commons