In the interwar period, some military officers did not believe the combustion-engine vehicle would replace the horse for combat operations. Cavalry officers from around the world, including the United States and United Kingdom, argued strongly against trading their lovely mounts for new tanks. Many insisted that heavy armored vehicles could never conquer the hilly terrains in which horses thrived.
In 1925, Field Marshal Sir Douglas Haig, the World War I commander in chief of the British Expeditionary Force on the Western Front, described the new airplanes and tanks as “only accessories” to warfare that would still be determined by “man and horse.” In 1938, Brigadier General Hamilton S. Hawkins, a prominent American crusader for horses in combat, lamented the “sheep-like rush to mechanization.”
These arguments ensured that the U.S. and British militaries kept mounted cavalry units until the outbreak of World War II. But when the horse came up against the tank, even the most strident defenses of a horse-bound army became an embarrassing footnote in military history.
The world’s biggest militaries, most notably that of the United States, are now at a similar turning point. The deployment of cheap unmanned vehicles on a mass scale, in the war between Russia and Ukraine and more recently in the U.S. war with Iran, has revealed the vulnerability of hugely expensive tanks, aircraft, and ships and the near impossibility of making major advances by infantry on the ground. Yet Western militaries have failed to reckon with these innovations in warfare, planning and spending as if nothing new is afoot.
In 2024, Brigadier General Douglas Wickert, the U.S. Air Force official in charge of AI piloting technology, insisted that robotized warfare was still “centuries” away. A few weeks ago, the Department of Defense requested more than $5 billion to cover just the 2027 fiscal-year costs of its next generation of manned combat aircraft, the Boeing F-47. At about $300 million a jet, the F-47 is the most expensive fighter in the world. Flight tests aren’t expected to start until after 2031.
The U.S. isn’t the only major military sinking enormous sums into legacy systems. The French and Germans are designing a new tank, which is not slated to be ready until 2040. The Chinese are similarly continuing their buildup of massively armed and staffed aircraft carriers, and will soon start work on the largest nuclear aircraft carrier in the world.
[Simon Shuster: Building tanks while the Ukrainians master drones]
Yet the lessons of the wars before us are that cheaper, nonhuman piloted systems are changing the balance of fighting. The future will be dominated by weapons of war that do not have a human onboard. These include cruise and ballistic missiles, a long-familiar category of weapons that can effectively attack targets anywhere in the air, on land, or in water. More and more, warfare will also include unmanned ground vehicles, which can fight, deliver goods, and evacuate the wounded; sea drones, which threaten ships and containers around the world; and unmanned aerial vehicles, or drones, which Ukraine, Russia, China, Israel, and the U.S. are making by the millions.
By still spending so much on conventional manned equipment, the United States and many other countries are diverting funds from more efficient systems. Placing a human in a tank or plane makes these vehicles larger, more easily trackable, and trickier to deploy than their unmanned counterparts. Adding a human to a cockpit requires many more millions of dollars in cost, including small items such as the explosive bolts needed to launch ejector seats and specialized helmets. The helmets worn by F-35 pilots, for instance, were so expensive to develop and build that they are estimated to have cost up to $700,000 each.
Like most bureaucracies, militaries can be resistant to change. Procurement officers would rather stick to familiar ideas and actions. Defense contractors would prefer to keep building what they know how to make, and lawmakers want to preserve jobs in their districts. Defense-industry leaders have moved quickly to dismiss the ways that drone technology is reshaping the battlefield in Ukraine.
Taking the human out of the cockpit does not take humans out of the fight; it just puts them in a safer place. The Ukrainians, for example, made strides in their war against Russia when they reduced the number of human soldiers on the front line and turned over much of the fighting to remotely controlled systems. By some accounts, drones are responsible for more than 90 percent of Russia’s losses.
Questions over whether autonomous vehicles and weapons are handy supplemental tools or disruptive innovations are in keeping with a long-running debate over what constitutes a revolution in military affairs. The notion that modern technology was radically changing the battlefield originated with Soviet military planners—who noticed in the 1970s that Western equipment incorporated more electronic advances than their own did—but gained particular force in 1991 after the lopsided results of the first Gulf War, in which the U.S. used precision-guided weaponry to swiftly defeat the large and supposedly well-armed Iraqi forces of Saddam Hussein.
A growing faith in the focused, targeted lethality of a digitized battlefield helped convince the United States that it could indefinitely rely on its technical lead to win future battles—despite the evidence otherwise in Iraq and Afghanistan. Throughout most of the post-9/11 wars, the U.S. military focused on counterinsurgency warfare—a lower-intensity form of combat than in the first Gulf War or the 2003 invasion of Iraq—and deployed drones of its own. But the Pentagon’s confidence in its priciest, most complex weapons systems went unabated.
Sadly, subsequent experience has done little to curb the U.S. military’s smug complacency. Even as less-expensive drones have been wreaking havoc for years in Ukraine and Russia, the U.S. has seemed to think that its military is far too strong and technologically advanced to have to learn any new rules of warfare from the Ukrainians. This blinkered thinking about the superiority of America’s military is also what left President Trump insufficiently prepared in his war against Iran.
[Simon Shuster and Nancy A. Youssef: The glaring oversight in the U.S. war plan]
The growing role of machines that are operated by a human at a distance—or programmed to operate autonomously—represents a fundamental “de-humanizing” of war. This has enormous implications for how wars start, end, and are waged. Instead of wasting time and resources debating whether future military systems should be manned or not, the U.S. military and others should be wrestling with the degree to which unmanned systems should remain under direct human control or be guided by artificial intelligence.
Autonomous AI-controlled weapons are developing apace as well. By crunching massive amounts of data in a short time, AI can help identify and strike enemy targets. The U.S. has claimed to have used such weapons in the bombing of Iran. The Ukrainians have announced a new air-defense system that can operate completely autonomously from launch, and can make its own “decisions” on how best to attack a Russian target.
Such innovations challenge conventional assumptions about air superiority. Cheap Iranian drones have been effectively targeting American bases in the Persian Gulf, regardless of the patrolling of fancy F-35 jets. Digital advances also raise pressing questions about the ethics of outsourcing life-and-death decisions to machines. These are the discussions we should be having. The nature of war has changed, and countries that fail to acknowledge it run the risk of sounding like nostalgic defenders of a mounted cavalry—and of being left behind.