|
| |
| Activity title | Artificial Intelligence in Military Training and Education | Activity Reference | HFM-MSG-ET-218 | Panel | HFM | Security Classification | PUBLIC RELEASE | Status | Active | Activity type | ET | Start date | 2024-04-17T00:00:00Z | End date | 2025-04-17T00:00:00Z | Keywords | Artificial Intelligence, Competency Development, HumanAgent Teaming, Large Language Models, Learning Engineering, Skill Acquisition, Team Training Intelligent Tutoring Systems, Training and Education | Background | Artificial Intelligence (AI) is maturing at a rapid pace. Developments in Machine Learning have been creating new possibilities for training data analysis and assessment, and more recent advancements in generative AI are forcing governments and industry sectors to react quickly. Of particular interest is examining how these technologies, tools, methods and techniques apply specifically to training and education requirements across the NATO alliance. Advancements in large language models and generative AI, digital twins, and multi-modal data analysis have the potential to significantly impact training practices across all domains of military operation. Combined with advances in cognitive science, AI enables sophisticated intelligent tutoring, generation of tailored training content, and improved fidelity in training support agents. These capabilities form critical pillars for adaptive and personalized training that can improve the ability of organizations to meet training requirements and develop technologies and strategies for using limited time and resources to meet the specific needs and learning preferences across individual soldiers and varying team structures. At the same time, advanced AI technologies can create significant challenges for training systems, including a higher requirement for technical competencies across the training enterprise, larger upfront development and infrastructure costs, a strategy for collection and management of data across large learning cohorts, and ethical implications with the misuse of generative AI by students, or with career-changing assessments performed by machines. The promise of advanced AI for training needs to be balanced against these challenges and complexities, which requires careful analysis of the technology landscape and how the NATO alliance can take advantage of the recent advancements in this exciting field of study.As respective organizations and laboratories across the NATO alliance execute research and development on the utility of AI to support training requirements, it is critical to leverage shared investments across our collective community to help accelerate and mature capability, while focusing on best practices and principles for applying it at scale to meet future training, education and human performance requirements. | Objectives | The primary objective of this Exploratory Team is to provide a focus and strategy for exploitation of AI tools and methods to support training and education requirements aligned with NATO-alliance readiness.This Exploratory Team aims to:a. Define and categorize a framework built on an expert informed collective view of what we want AI to do as an end-state in the training, education, learning and development space.b. Establish a taxonomy to define and coordinate capabilities and trends in AI aligned to the end-state framework, examining where different generations of AI align and whether we can leverage those successes to meet end-goals.a. Define limitations and gaps across the NATO alliance based on a review of existing systems and services.
c. Create a prioritization list across the recognized capabilities and trends, and down select and build justification for focused exploration through a research task group Technical Activity Proposal (TAP). | Topics | The following topics/tasks will be explored by this activity. These topics associate with common Intelligent Tutoring System (ITS) and Adaptive Instructional System (AIS) paradigms, which aim to leverage AI to personalize and optimize technology mediated learning opportunities.This is not a comprehensive list. The initial goal will be to define technologies, techniques, and trends across these application spaces. This list will be refined at the kick-off and across the execution of the ET. The resulting AI capabilities will be voted upon to create a prioritization list across the team, with the intention to identify specific AI topics and tasks for more formal investigation by a Research Task Group. These topics will be explored from cognitive science, human factors, and human system integration perspectives. This will involve aligning capabilities to use cases focused on training and education of individuals and teams, and supporting overall competency development across domains that require cognitive, psychomotor, and affective skill sets to be proficient in their roles and to maintain readiness. Topics for consideration include:a. Large Language Models[1] Content Creation[2] Conversational Agents[3] Informal Learning / Self-Developmentb. Human Representation in Digital Engineering[1] Video Capture and Avatar Rendering[2] Learning process modelling[3] Cognitive models for realistic behavior representation[4] Digital Engineering c. Multi-Modal Assessment / Ensemble Modeling[1] Assessment in live and virtual environments/adaptive training systems[2] Computer Vision, Natural Language[3] Support to assessment by trainers (e.g., in classroom)[4] On-the-job and informal assessment d. Learning Technology Standards and AI[1] TLA / XAPI[2] Policies and Procedures[3] Ethics around AI in Training and Educationi. Mis-use of generative AI models by students; implications for assessment
e. Acquisition Requirements[1] Data Strategy Policy[2] Capability Design Document ‘Requirements’f.Workforce Demand[1] Certification and Competency Standards[2] Implications for assessment and skill fade[3] Implications of on-the-job or informal assessment for career progression or personnel issues (e.g., warnings on pers file) | | Contact Panel Office |
| |
|
|
|
