The future of war is the future of society

The most prophetic post I’ve ever written wasn’t about economics — well, not directly at least. It was about military technology. I’m not much of an expert in that subject, but I managed to make some predictions that were unpopular at the time but which have been borne out spectacularly in the years since.

The original article, written for Quartz, was in 2013, but right now I can only findthis republished version from 2020. Here’s what I wrote:

The human race is on the brink of momentous and dire change. It is a change that potentially smashes our institutions and warps our society beyond recognition. It is also a change to which almost no one is paying attention. I’m talking about the coming obsolescence of the gun-wielding human infantryman as a weapon of war…

You may not even realize you have been, indeed, living in the Age of the Gun…But imagine yourself back in 1400. In that century…the battlefield was ruled not by the infantryman, but by the horse archer—a warrior-nobleman who had spent his whole life training in the ways of war. Imagine that guy’s surprise when he was shot off his horse by a poor no-count farmer armed with a long metal tube and just two weeks’ worth of training. Just a regular guy with a gun…

For centuries after that fateful day, gun-toting infantry ruled the battlefield…But sometime in the near future, theautonomous, weaponized dronemay replace the human infantryman as the dominant battlefield technology. And as always, that shift in military technology will cause huge social upheaval.

The advantage of people with guns is that they are cheap and easy to train…The hand-held firearm reached itsapotheosiswith the cheap, rugged, easy-to-use AK-47; with this ubiquitous weapon, guerrilla armies can still defy the mightiest nations on Earth…

But another turning point in the history of humankind may be on the horizon. Continuing progress in automation, especially continued cost drops, may mean that someday soon, autonomous drone militaries become cheaper than infantry at any scale.

Note that what we call drones right now are actually just remote-control weapons, operated by humans. But that may change…Sometime in the next couple of decades, drones will be given the tools to take on human opponents all by themselves…meanwhile, technological advances and cost drops in robotics continue apace. It is not hard to imagine swarms of agile, heavily armed quadrotor drones flushing human gunmen out of buildings and jungles[.]

That was 12 years ago. At the time, when I floated this idea on Twitter, people jumped to scoff at it. They told me that electronic warfare would be too powerful for drones to overcome, that drones wouldn’t have the firepower to dominate the battlefield and so on. They pointed out — quite correctly — that when it comes to military technology, I don’t have any expertise.

And yet, as of 2025, my prediction has been utterly vindicated. Reports from the battlefield in Ukraine tell of a battlefield so completely dominated by drone warfare that experts are forced to go out of their way to argue that traditional artillery still has a role. Here’sMichael Kofman(my favorite Ukraine expert):

Drones continue to be responsible for most daily casualties, with the front line defined by overlapping drone and artillery fire engagement zones 20-25km from the forward line of troops. This is commonly referred to as the ‘kill zone.’..Drone units work to suppress and displace the opponent’s drone crews further from the front…[L]onger range drones strike artillery, logistics, and enemy drone teams further in the rear…

On the Russian side Rubicon formations remains a leading problem for drone operators, not only the drone companies themselves, but because they train other Russian drone units to replicate their approaches focused on AFU logistics, drone crews, and intercepting ISR…One of the observed changes is the balance of casualties in the AFU has shifted from infantry to supporting roles, drone operators, logistics, etc. There is very little infantry forward, and in many AFU brigades infantry now bears less of the casualties…

Most units now have a UGV platoon, company, or battalion. These require greater skill and training to employ, but hold considerable promise, reducing casualties…The airspace has become even more contested for longer-range ISR, with both sides establishing dense tactical radar coverage to detect drones, and one way attack munitions…

Artillery remains important to suppressing enemy forces and shaping how they attack, especially in bad weather, which is more prevalent this time of year. Fog, wind, and rain significantly degrade drone operations[.]

Other reportsall tell the same story.

The drone is increasingly regarded as the infantryman’s basic weapon. The US Army isordering a million dronesto equip its soldiers (a war would require many, many times that). Drones are replacing artillery, now having the capability to take out infantry, tanks, artillery, and basically anything else at a fairly long range.

Strike drones are supplementing bombers and long-range missiles as a way of dealing damage behind the lines; Ukraine’s drone strikes aredegrading Russia’s oil industryfrom thousands of miles away.

And drone technology is still in its infancy. Currently, drones are still piloted by humans. This makes them subject to electronic warfare that jams the link between pilot and drone, forcing them to use spools of fiber-optic cable to maintain a secure connection.

And it means that drone operators have to stay somewhat near the front, exposing them to enemy strikes. Skilled human operators are a valuable resource that limits the amount of drones that can be used at once.

This is about to change. Advances in AI are going to enable drones to behave autonomously — the “killer robots” out of a science fiction novel. In fact, Ukraine has alreadyexperimented with autonomous drone swarms.

Drones are going to first supplement and then replaceboats,fighter jets,submarines, and every other manned weapon of war. Human infantry and human-crewed vehicles will become obsolete due to their sheer expense. Soldiers and big vehicles cost a lot; drones cost much less.

All of these predictions are fairly obvious and easy to make. AI is only getting better. And machines are generally cheaper than humans, who are only going to get more expensive over time.[1]Those two facts are all I really needed[2]in order to predict the rise of drones back in 2013, and neither has changed since then.

What happens next is harder to foresee. Obviously, everyone will look for ways to shoot down swarms of drones. At first, this will involve very fast guns, likeRheinmetall’s Skyranger. A drone is cheap, but a bullet is cheaper.

Even cheaper, eventually, will be a puff of light;laser weapons are being developedthat can shoot down drones cheaply, quickly, and very accurately. Eventually, we may see big battleships and tanks bristling with point defense lasers force their way through swarms of drones, while defenders try to take them out with big, fast missiles.

Maybe that will result in the return of WW2-style maneuver warfare. Or maybe missiles will cheaply take down any big vehicle, creating static battlefields more like World War 1 (or the current Ukraine war), where the only way to win is to have your economy produce more drones than the enemy. Recall that I’m not actually an expert in military technology, so I can’t say how this will shake out. I’m not sure if anyone knows yet.

But what I do think is very likely is that the organization of human societies will have to change. Take a look atthe long-term history of warfare. Our numbers are pretty patchy, but as far as we can tell, there have been three really big waves of warfare over the last millennium:

1. The Mongol conquests in the 1200s (and follow-up conquerors in the 1300s like Timur)
2. The Thirty Years’ War and the fall of the Ming Dynasty in China in the 1600s
3. The World Wars and communist revolutions of the 1900s

People argue a lot over why there were these three big outbreaks of war all over the world. Some blame climate change, while others blame patterns of trade, population growth, and so on. But I think one big plausible factor ismilitary technology.

Each of the three waves of war coincides with a dominant package of military technology. The Mongols ran circles around their opponents with stirrup-equipped horses, and outranged them with recurved bows. The wars of the 1600s represented the peak of gunpowder warfare, while the wars of the 20th century were the peak of industrial warfare — planes, tanks, metal ships, and so on.

Interestingly, none of those big wars happened right after the key technologies were introduced. There was always a substantial lag. Most of the bow and stirrup technologies that made the Mongols so fearsome wereinvented a millennium earlier by the Xiongnu(the predecessor of the Huns). Cannon and muskets were invented a century before the cataclysms of the 1600s.

The World Wars saw rapid innovation, but the machine gun, the howitzer, the ironclad battleship, and other key technologies were pioneered earlier. There were constant incremental improvements in all of these technologies, of course, but it’s unlikely that they reached some special threshold of lethality that caused wars to suddenly get much much bigger and deadlier.

Instead, what changed were thesocietiesthat made use of the weapons.

There were plenty ofsteppe empiresbefore the Mongols, but they usually weren’t able to overcome densely populated, settled civilizations like China. Only once Genghis Khan implemented reforms like meritocracy, writing, and so on were the steppe warlords able to break through and conquer the world.

Mongol tactics — especially the ability of mounted units to remain separate for days and then all converge on the same place at the same time — required sociocultural innovations to prevent defection/betrayal and ensure cooperation among highly mobile subcommanders.

Similarly, firearms and artillery saw wide adoption by the 1400s, but it wasn’t until the 1600s that gunpowder armies grew to truly massive size — hundreds of thousands of men instead of tens of thousands. That increase in size drove much of the higher death toll of the wars of the 1600s, since those giant armies had to live off the land, which caused famines and massacres (the battlefield death toll was higher too, obviously).

Those big armies required lots of money, which required lots offinancing— better tax collection and more bank loans.Paul Kennedy notesthat the countries that won wars in the gunpowder era tended to be the ones that were the best at tax collection.Charles Tilly arguesthat gunpowder-era wars made the modern state, because the regimes that survived were the ones that developed complex bureaucracies in order to collect more taxes to fund their wars.

The World Wars were also made possible by innovations in social organization — modern corporations, even larger bureaucracies, modern supply and logistics, and continuous research and development. Those social innovations emerged partly as a way to make countries more effective in war — if youdidn’tdevelop those things, you were liable to be conquered.

It’s likely that all three eras of warfare required innovations in persuasion, ideology and communication as well. Genghis Khan had to convince a plethora of fractious steppe warlords to all stick together and cooperate. The great powers of the 1600s used a combination of religion and local loyalties — often communicated using the printing press — to help motivate their large armies to fight.

In the 20th-century wars, ideologies like communism, fascism, and democracy were key, and radio was an important new tool. All of those ideological changes changed society as well.

It was these social changes that allowed wars to get so big and cataclysmic in each era. So it might seem like the changes I’m describing are a bad thing — and in fact, I do think it’s a distinct possibility that we’ll see a cataclysmic global drone war sometime in the future. But from any one country’s point of view, those social changes were absolutely necessary, since the only thing worse than winning a war is losing a war.

And historically, warfare acted as a vector for the beneficial spread of both physical and social technologies. Countries in the early modern period had to adopt modern fiscal systems in order to be able to resist conquerors who already had those systems in place. Industrial production spread not just because it made citizens wealthier, but because you needed it to make guns, ammunition, railroads and so on.

In other words, it’s very difficult for countries to resist the changes that new dominant packages of military technologies necessitate. In the 1700s and 1800s, you either learned to be a gunpowder empire or you got conquered and enslaved. In the 20th century, you had to become a modernized industrial state, or you got plowed under. If there is a single driving, irresistible force of history, I think it must be the innovations that we use to kill each other.

Right now, we’re in the middle of a revolution in military affairs that will be just as profound as the Industrial Revolution, the introduction of gunpowder, or the rise of steppe warfare. The two great inventions of our time — AI andthe electric tech stack— are rapidly eclipsing industrial warfare.

That will probably make war more capital-intensive, as human labor is increasingly removed from the equation (and because AI is very capital-intensive). It will almost certainly make it moreknowledge-intensive, as understanding of how to apply AI effectively becomes decisive.

These changes are going to force our society to change and adapt. The America that won World War 2 didn’t look like the America of Thomas Jefferson’s time. And if we want to remain powerful and secure, the America of the 21st century won’t look like Roosevelt’s America, either. It can’t.

Right now, China looks like it’s outpacing other societies in terms of adapting itself to the new requirements of war. Possibly alone among all countries, it has mastered every part of the drone supply chain.

China’s emphasis on manufacturing, while perhaps not economically efficient, has also probably prepared it better for prolonged capital-intensive war. Right now, they would be able to out-produce the rest of the world in terms of drones.

If we are to match them, we’re going to need to make better use of industrial policy to patch the holes in our supply chains, stop wrecking those supply chains with tariffs on our allies, and form closer partnerships with those allies so that we canachieve the scaleto match China.

China also seems to be adapting the new communications technologies of social media more effectively than many of its rivals — or at least, more effectively in terms of warfighting ability. We scoff at China’s massive system of internet thought control, butit’s possible that this is the only wayto keep a modern nation from fracturing and spiraling into chaos in the age of social media.

I hope there are ways to hold together a modern nation-state without resorting to centralized thought control. But we really need to startlooking forthose techniques, instead of just assuming that stability will somehow naturally reemerge from the chaos of X and TikTok.

Most of all, America — and Europe, and other developed countries — must prepare for dramatic, wrenching change. These days, we often find ourselves looking backward —resisting the adoptionof new technologies, blocking development and wallowing in nostalgia forthe glory daysof the 20th century.

But the kind of society we had then will not serve us as well in an age when flying robots rule the battlefield. Just as Thomas Jefferson’s America gave way to Franklin D. Roosevelt’s, we must find a way to preserve the core benefits and advantages of our liberal democracy while transitioning to the next thing.

Right now, I don’t think either Americans or Europeans are mentally prepared for that change. I hope it doesn’t take a catastrophic defeat in a major war in order to wake them up.

Notes

1 The cost of a human soldier goes up with both lifetime earnings and with the subjective value that society places on each human life. Both of those things go up as economies grow.

2 Technically, I also had to know that explosives carried by drones would win the battle against new kinds of armor. But that’s just obvious from basic physics. It takes a huge amount of energy.

Thisarticlewas first published on Noah Smith’s NoahpinionSubstack and is republished with kind permission. Become a Noahopinionsubscriberhere.