Mastering human-machine warfighting teams

An article by War on the Rocks explores the critical role of human-machine teaming in the future of warfare, focusing on the evolving integration of AI, autonomous systems, and human operators within military contexts. The author argues that successfully mastering human-machine teams will be essential for future military advantage, and proposes that this integration requires rethinking current approaches to both human training and AI development.

Military leaders, such as the Chief of Staff of the Air Force and the Commanding General of Army Futures Command, view mastering the cooperation between humans and increasingly capable AI as a critical advantage in future warfare. This vision goes beyond merely using machines to replace humans in high-risk scenarios. Instead, the goal is to leverage the complementary strengths of humans and machines for superior team performance. Human capabilities, such as judgment, context, and tacit knowledge, complement the machine’s strengths in data processing, speed, and repetitive tasks.

The opinion piece highlights several challenges in forming effective human-machine teams. Many efforts to train human personnel in human-machine teaming have focused on AI "explanations" — attempting to get AI systems to explain how they make decisions. However, the author critiques this approach, arguing that these explanations often do not enhance the trust or effectiveness of the team. Instead, they may inadvertently encourage human operators to defer to AI decision-making, even when the AI's reasoning is flawed.

One of the most significant arguments in the piece is that humans remain the most important part of any human-machine team, particularly in warfare. While AI can provide powerful assistance, it lacks the tacit knowledge and strategic reasoning that humans bring to the table. The author emphasizes that AI may be excellent at tasks requiring speed, data processing, and pattern recognition (e.g., locating enemy forces), but it cannot understand the context or purpose behind military objectives. Warfare is inherently a human activity, and AI, regardless of its capabilities, cannot replace the moral, ethical, and strategic reasoning that humans provide.

The author advocates for training mental models for human-machine teams. Just as humans develop a deep understanding of their teammates' strengths and weaknesses over time, repeated interactions with AI could help humans understand how best to collaborate with machines. This training, conducted through realistic military exercises, would help human operators develop the intuition needed to work effectively with AI systems in combat situations.

Another key proposal is that AI models should be designed to complement human abilities, rather than replicating or replacing them. For example, AI could be trained to detect improvised explosive devices (IEDs) in ways that humans might miss, such as using multiple sensor types. While AI may not always perform as well as human operators on simpler tasks, its greatest value lies in augmenting human capabilities for tasks that would be difficult or dangerous for humans to perform alone.

The piece also urges caution in the face of the AI hype. While AI’s progress in military applications is undeniable, the risks of overreliance on AI—particularly in military decision-making—could be dangerous. If the US military becomes too dependent on AI, it could lose sight of the human element that remains critical to warfare. The ethical and strategic implications of AI must always be guided by human oversight to ensure responsible use.

The article makes a compelling case that humans and machines should be complementary, but it also recognizes the limitations of current AI technology. The suggestion that AI models should not simply perform tasks that humans are good at but should instead focus on tasks where they can augment human capabilities seems like a reasonable and necessary strategy.

The critique of "explainable AI" and how it leads to human over-reliance on machines is a strong point. The piece argues that humans naturally trust other humans and that AI’s lack of emotional cues makes humans more likely to blindly accept AI decisions. This highlights a key ethical consideration: how to ensure humans maintain control over the decision-making process in critical combat scenarios.

The opinion piece emphasizes that war, at its core, is a human activity. This is an essential point, as it underscores the need for human judgment in military operations. Machines may be effective at tasks like logistics or intelligence gathering, but they cannot fully grasp the moral and strategic imperatives that underlie decisions made in war. Thus, the human-machine team must retain human oversight at all times.

The article points to the difficulty of integrating AI into military operations in a way that enhances, rather than hinders, performance. One significant challenge is human-machine trust, and the suggestion that humans need to develop stronger mental models for collaborating with AI is an important one. The article suggests that putting AI into military exercises as soon as possible, even with limitations, will provide valuable training for future integration.

The article’s skepticism of AI overhype is important and timely. While AI is clearly a powerful tool, it’s essential to recognize its current limitations and ensure that human judgment remains the primary decision-making force. The military’s commitment to human-machine collaboration should always prioritize human control, especially in situations that involve moral choices or the potential for catastrophic consequences.

The piece is a thoughtful examination of how the military should approach human-machine integration in warfare. It calls for a balanced and nuanced understanding of AI’s role in military operations, highlighting the complementary strengths of both human and machine intelligence. While AI can provide substantial advantages in areas like speed, data analysis, and precision, humans must remain the central decision-makers. To successfully integrate AI into military operations, the US military must focus on building trust, ensuring complementary skillsets, and maintaining human oversight at all times.

By Vafa Guliyeva