Applications of Computational Command Leadership AI Models (ACCLAIM)

A military without good leadership is a mob. Identification, preparation, guidance and mentoring of potential leaders are critical functions across all services. Leadership training and support can be substantially improved by refinement of theories and models on what constitutes good leadership, and what enables individual to be effective leaders. DoD needs better models of how leaders learn, decide, and act—cognitive models that include social and emotional influences, and support computation for explanatory tracking, prediction, and/or prompting. Stottler Henke proposes to work with Roger Schank’s Engines for Education organization to develop computational theories and models of command leadership, built on artificial intelligence (AI) technology, shaped by cognitive science insights, and driven by the functional demands of selected applications. The resulting theory and system—Applications of Computational Command Leadership AI Models (ACCLAIM)—will capitalize on observations and data on how real leaders think, communicate, and act. During Phase I we will gather data and prior theory on command leadership; elaborate our initial theories; identify promising applications; use those applications to clarify theory requirements; develop preliminary ACCLAIM designs suiting those requirements; characterize metrics and methods for evaluating ACCLAIM; and develop a Phase II plan to follow through on theory, design, and evaluation.