"It's a security risk not to have it. At this point, we have to have it."
With these urgent words, Air Force Secretary Frank Kendall described his experience after flying in an AI-controlled fighter jet. His feeling reflects a growing consensus among defense leaders that artificial intelligence is not merely a technological luxury but an operational necessity for America's military supremacy.
However, this enthusiasm for AI as America's next great military advantage may be overlooking the enduring realities of warfare.
The Pentagon's faith in AI stems from a long tradition of seeking technological "offsets" to counter adversaries' numerical advantages.
During the Cold War, when Soviet forces vastly outnumbered NATO in Central Europe, America developed nuclear weapons as its first "offset" -- a technological advantage that could deter or defeat a numerically superior opponent.
The second offset emerged in the 1970s and 1980s through precision-guided weapons and stealth technology, again providing America with a crucial military edge.
In 2016, the Department of Defense announced a third offset strategy that would focus on "combinations of technology, operational concepts and organizational constructs."
Today, as China fields the world's largest military with approximately 2.18 million active military, including the biggest navy and a rapidly modernizing air force, defense planners are hunting for a new offset. AI, with its revolutionary advances in computing power and machine learning, seems to offer that promise.
Modern AI systems, powered by sophisticated graphics processing units and deep neural networks are enabling computers and machines to do things traditionally done by humans. As a result, we are now seeing unprecedented capabilities in autonomous systems, cyber warfare, and military command and control.
Yet the dream of AI as a war-winning technology faces several sobering realities.
First, unlike nuclear weapons, AI is not fundamentally changing the nature of great power competition. While nuclear weapons made direct conflict between major powers potentially civilization-ending, AI's impressive capabilities -- though significant -- do not yet create the same paradigm-shifting deterrent effect.
For those banking on AI to be the key to unlocking this third offset, the ongoing war in Ukraine provides a stark reality check.
Both Russia and Ukraine are employing AI-enabled systems at unprecedented levels, from autonomous drones to sophisticated electronic warfare capabilities. According to recent Space Force assessments, the conflict has featured "more electronic warfare than we have ever seen before."
Yet despite these advanced technologies, neither side has achieved a decisive advantage. Instead, the war has devolved into a grinding battle of attrition, eerily reminiscent of conventional conflicts from the past century.
History offers numerous examples of promising technologies that failed to deliver offset-level advantages.
Nazi Germany's V-1 and V-2 missiles, despite representing revolutionary advances in propulsion technology, caused significantly fewer civilian casualties in Britain than conventional bombing raids. Some historians argue that Germany's massive investment in these advanced weapons was a strategic mistake, diverting crucial resources and industrial capacity away from more conventional military needs during a critical phase of the war.
Similarly, although the helicopter revolutionized battlefield medicine and dramatically reduced combat fatality rates in Vietnam by enabling rapid medical evacuation of approximately 900,000 troops, this tactical advantage did not alter the war's ultimate strategic outcome -- eventual withdrawal in the face of a determined adversary willing to absorb substantial losses.
None of this suggests that the U.S. should abandon AI development. In modern warfare, every technological edge matters. However, we must resist the temptation to view AI as a silver bullet that will deter or defeat numerically superior opponents.
The Ukraine conflict demonstrates that even with advanced AI systems, future wars between major powers will likely still be won through industrial capacity, logistical superiority, and the ability or willingness to sustain losses.
Instead of placing excessive faith in AI alone, America needs a comprehensive approach that leverages all of its national advantages. This means maintaining our robust industrial base, strengthening our network of global alliances and preserving our economic dynamism.
These fundamental strengths -- more than any single technology -- have historically proven to be the true foundation of American military power.
In preparing for potential future conflicts, particularly in regions like the Taiwan Strait where we might face numerically superior forces, we must remember that history favors nations equipped for sustained conflict.
The side with the most sophisticated AI will certainly have advantages, but victory will likely go to the competitor best prepared for a prolonged struggle -- the one with the industrial capacity to replace losses, the economic strength to sustain a war effort, and the national will to endure hardship.
As we invest in AI capabilities, we must also maintain a realistic perspective on its limitations. History showed that the true test of military power is not mere technological sophistication, but the ability to integrate technology with other elements of national power to achieve strategic objectives.
In this broader context, AI is an important tool, but not a guarantee of victory in future conflicts.