[Editor’s Note: The Mad Scientist Laboratory has continuously tracked the rise of Artificial Intelligence (AI) since its inception in 2017. In our continuing mission to explore the evolving Operational Environment and identify the ramifications of emergent trends affecting the U.S. Army, we’ve explored how AI is an emerging game-changer, how it could affect battlefield operations, and even how our adversaries could use it to generate novel Biological/Chemical agents.
One of the most compelling near-term applications of AI is enhancing Soldiers’ intellectual performance. SGM Kyle J. Kramer made the case to “integrate AI into Soldier training, forging a force equipped to excel on the modern battlefield.” LtCol Joe Buffamante, USMC, explored how applying human-machine teaming can enhance Professional Military Education (PME). Dr. Billy Barry and LTC Blair Wilcox recently addressed how hybrid intelligence can amplify learning during Army wargames.
Today’s post adds to our body of knowledge on how AI can augment, enhance, and empower Soldier performance. Guest blogger Ben Van Roo explores how Generative AI (GenAI) and Large Language Models (LLMs) can amplify our Soldiers’ and Leaders’ capabilities, while “enabling TRADOC to execute its mission more effectively than ever before” — Read on!]
In an era where complexity defines the battlefield, TRADOC sits at the center of the U.S. Army’s effort to shape the fielded force. Tasked with responsibilities ranging from training Soldiers to crafting doctrine, TRADOC’s mission is both expansive and urgent. Operational environments (OEs) have never been more multifaceted, encompassing political, military, economic, and social domains, all while presenting commanders with incomplete or conflicting data. It’s a daunting challenge, but also one that aligns perfectly with the strengths of generative AI (GenAI) and large language models (LLMs).
Per NVIDIA, the AI industry’s leading chipmaker:
GenAI “enables users to quickly generate new content based on a variety of inputs. Inputs and outputs to these models can include text, images, sounds, animation, 3D models, or other types of data.”
LLMs “are deep learning algorithms that can recognize, summarize, translate, predict, and generate content using very large datasets.”
These technologies won’t replace human expertise; instead, they promise to augment it. The vision is clear: to help commanders, trainers, and planners cut through the noise, adapt to rapid change, and make better decisions, faster. Yet, as with any transformative technology, the promise of GenAI comes with significant hurdles, from technical implementation to ensuring trust and accuracy. The path forward will require not only innovation but also patience and a strategic approach to integration.
Operational environments are inherently dynamic and filled with shifting variables, from geopolitical upheaval to adversarial maneuvers and civilian considerations. Preparing for these complexities requires processing a mountain of information, from intelligence reports to historical precedents and real-time updates. Traditionally, synthesizing this data can take weeks—time we often don’t have. LLMs, with their unparalleled ability to process and organize information, offer a path forward. Imagine a commander needing a comprehensive overview of tensions in Eastern Europe. An LLM could synthesize disparate reports, extract key trends, and provide actionable 
insights in minutes. It’s not just about speed; it’s about clarity and focus. Soldiers can engage with these models conversationally, asking natural-language questions like, “What are the key political and military risks in this region?” or “Explain how this terrain impacts our logistics.” This transforms a sprawling, overwhelming dataset into a navigable, interactive resource.
This potential is transformative, but it isn’t without its challenges. The sheer volume and diversity of content for any operational environment pose significant obstacles, even for cutting-edge AI. Training these models to prioritize what matters without overwhelming users with irrelevant details requires ongoing refinement. It’s a process that demands meticulous effort, collaboration, and feedback from Soldiers and analysts alike.
TRADOC’s Decisive Action Training Environment (DATE) is one of its most valuable tools for simulating realistic adversaries and scenarios. Yet, in a world where adversaries continuously evolve, realism is a moving target. Static training manuals and scenarios often fall short of preparing Soldiers for dynamic, unpredictable threats. Here, GenAI shines. LLMs can automate updates to the DATE Force Structure, incorporating the latest intelligence into training scenarios almost instantly. They can simulate adaptive enemy behaviors, forcing trainees to think critically and respond dynamically. Imagine a training exercise where every iteration introduces new tactics, unpredictability, and depth—just like a real adversary would. This kind of evolution keeps training relevant and Soldiers ready.
Another critical challenge TRADOC faces is the overwhelming volume of knowledge it maintains. From field manuals to intelligence reports and doctrine, this information is invaluable but often inaccessible due to its complexity and scale. Soldiers describe the experience as trying to drink from a firehose. LLMs offer a practical solution, allowing Soldiers to query this knowledge base conversationally. A Soldier could ask, “What’s the effective range of the T-90 tank’s main gun?” and receive an instant, precise answer. This isn’t just about saving time; it’s about empowerment. Soldiers can learn, adapt, and prepare on the fly, even in high-pressure situations.
LLMs also hold immense potential in building better Orders of Battle (OOB). Traditionally, OOB creation requires a meticulous analysis of enemy capabilities, positions, and likely actions—an art and science that can consume 
valuable time. LLMs can rapidly organize and synthesize vast amounts of information, identifying patterns and constructing plausible OOBs in minutes. This efficiency frees planners to focus on strategic decision-making rather than being bogged down by data collection. However, it’s critical to recognize that LLMs are not perfect reasoning engines. They excel at organizing information and summarizing patterns, but lack the nuanced judgment and deep strategic insights that come from years of military experience. OOB creation, like most aspects of military planning, will continue to require a mix of tools, from simulations and optimization models to human expertise.
One of the less obvious but equally transformative roles for LLMs lies in content creation. TRADOC’s content needs are immense, spanning training manuals, PMESII-PT briefs, and operational plans. Creating these materials by hand is labor-intensive and slow. GenAI offers a way to dramatically accelerate 
this process. It can draft detailed training scenarios, generate comprehensive analyses, and even suggest predictive questions for planners to consider. For example, in preparing for operations in a fictional conflict zone, an LLM could produce adversary profiles, environmental challenges, and even strategic contingencies. This allows human experts to shift their focus to higher-order thinking and decision-making.
The vision for GenAI and LLMs in TRADOC isn’t just about efficiency—it’s about readiness. These tools offer a way to outpace adversaries, anticipate challenges, and ensure that our Soldiers are prepared for the complexity of future conflicts. But none of this will be easy. The volume of data, the 
intricacies of OEs, and the precision required for military planning make this a daunting task. Still, the stakes couldn’t be higher. In a world where adaptability and information advantage win wars, TRADOC’s embrace of GenAI could ensure that the Army remains not just a step ahead, but miles.
The question is no longer whether GenAI and LLMs can make a difference. The question is how quickly we can make them operational and ensure they fulfill their potential. These tools won’t replace the expertise of Soldiers and Leaders, but they will amplify it, enabling TRADOC to execute its mission more effectively than ever before. The future isn’t waiting, and neither should we.
If you enjoyed this post:
Check out TRADOC Pamphlet 525-92, The Operational Environment 2024-2034: Large-Scale Combat Operations
Explore the TRADOC G-2‘s Operational Environment Enterprise web page, brimming with information on the Operational Environment and how our adversaries fight, including:
Our China Landing Zone, full of information regarding our pacing challenge, including ATP 7-100.3, Chinese Tactics, BiteSize China weekly topics, People’s Liberation Army Ground Forces Quick Reference Guide, and our thirty-plus snapshots captured to date addressing what China is learning about the Operational Environment from Russia’s war against Ukraine (note that a DoD Common Access Card [CAC] is required to access this last link).
Our Russia Landing Zone, including the BiteSize Russia weekly topics. If you have a CAC, you’ll be especially interested in reviewing our weekly RUS-UKR Conflict Running Estimates and associated Narratives, capturing what we learned about the contemporary Russian way of war in Ukraine over the past two years and the ramifications for U.S. Army modernization across DOTMLPF-P.
Our Iran Landing Zone, including the latest Iran OE Watch articles, as well as the Iran Quick Reference Guide and the Iran Passive Defense Manual (both require a CAC to access).
Our Running Estimates SharePoint site (also requires a CAC to access), containing our monthly OE Running Estimates, associated Narratives, and the 2QFY24, 3QFY24, 4QFY24, and 1QFY25 OE Assessment TRADOC Intelligence Posts (TIPs).
Read following related Mad Scientist Laboratory AI content — spanning the gamut of potential applications:
Artificial Intelligence (AI) Trends
Takeaways Learned about the Future of the AI Battlefield and associated information paper
Artificial Intelligence: An Emerging Game-changer
Artificial Intelligence: Shaping the Future of Biological-Chemical Warfare, by Jared Kite
Training Transformed: AI and the Future Soldier, by proclaimed Mad Scientist SGM Kyle J. Kramer
The AI Study Buddy at the Army War College (Part 1) and associated podcast, with LtCol Joe Buffamante, USMC
Hybrid Intelligence: Sustaining Adversary Overmatch and associated podcast, with proclaimed Mad Scientist Dr. Billy Barry and LTC Blair Wilcox
Rise of Artificial Intelligence: Implications to the Fielded Force, by John W. Mabes III
Integrating Artificial Intelligence into Military Operations, by Dr. James Mancillas
“Own the Night” and the associated Modern War Institute podcast, with proclaimed Mad Scientist Bob Work
Bringing AI to the Joint Force and associated podcast, with Jacqueline Tame, Alka Patel, and Dr. Jane Pinelis
Thoughts on AI and Ethics… from the Chaplain Corps
The AI Study Buddy at the Army War College (Part 2) and associated podcast, with Dr. Billy Barry, USAWC
Gen Z is Likely to Build Trusting Relationships with AI, by COL Derek Baird
Hey, ChatGPT, Help Me Win this Contract! and associated podcast, with LTC Robert Solano
Chatty Cathy, Open the Pod Bay Doors: An Interview with ChatGPT and associated podcast
The Future of Learning: Personalized, Continuous, and Accelerated
The Guy Behind the Guy: AI as the Indispensable Marshal, by Brady Moore and Chris Sauceda
AI Enhancing EI in War, by MAJ Vincent Dueñas
The Human Targeting Solution: An AI Story, by CW3 Jesse R. Crifasi
An Appropriate Level of Trust…
About the Author: Ben Van Roo is the Co-Founder and CEO of Yurts, a generative AI company partnering with the US Department of Defense to advance mission-critical systems. He holds a PhD in Operations Research and has significant experience developing AI solutions for defense and national security applications.
Disclaimer: The views expressed in this blog post do not necessarily reflect those of the U.S. Department of Defense, Department of the Army, Army Futures Command (AFC), or Training and Doctrine Command (TRADOC).
