LOFTER-网易轻博

The New Revolution in Military Affairs

War’s Sci-Fi Future

By Christian Brose

Essay May/June 2019 Issue SecurityScience & Technology 

美国期刊《外交事务》

In 1898, aPolish banker and self-taught military expert named Jan Bloch published The Future of War,the culmination of his long obsession with the impact of modern technology onwarfare. Bloch foresaw with stunning prescience how smokeless gunpowder,improved rifles, and other emerging technologies would overturn contemporarythinking about the character and conduct of war. (Bloch also got one majorthing wrong: he thought the sheer carnage of modern combat would be so horrificthat war would “become impossible.”)

What Bloch anticipated has come to be known as a“revolution in military affairs”—the emergence of technologies so disruptivethat they overtake existing military concepts and capabilities and necessitatea rethinking of how, with what, and by whom war is waged. Such a revolution isunfolding today. Artificial intelligence, autonomous systems, ubiquitoussensors, advanced manufacturing, and quantum science will transform warfare asradically as the technologies that consumed Bloch. And yet the U.S.government’s thinking about how to employ these new technologies is not keepingpace with their development.

This is especially troubling because Washingtonhas been voicing the same need for change, and failing to deliver it, eversince officials at the U.S. Department of Defense first warned of a coming“military-technical revolution,” in 1992. That purported revolution had itsorigins in what Soviet military planners termed “the reconnaissance-strikecomplex” in the 1980s, and since then, it has been called “network-centricwarfare” during the 1990s, “transformation” by U.S. Secretary ofDefense Donald Rumsfeld in these pages in 2002, and “the third offset strategy” by DeputySecretary of Defense Robert Work in 2014. But the basic idea has remained thesame: emerging technologies will enable new battle networks of sensors andshooters to rapidly accelerate the process of detecting, targeting, andstriking threats, what the military calls the “kill chain.”

The idea of a future military revolution becamediscredited amid nearly two decades of war after 2001 and has been furtherdamaged by reductions in defense spending since 2011. But along the way, theUnited States has also squandered hundreds of billions of dollars trying tomodernize in the wrong ways. Instead of thinking systematically about buyingfaster, more effective kill chains that could be built now, Washington pouredmoney into newer versions of old military platforms and prayed fortechnological miracles to come (which often became acquisition debacles whenthose miracles did not materialize). The result is that U.S. battle networksare not nearly as fast or effective as they have appeared while the UnitedStates has been fighting lesser opponents for almost three decades.

Yet if ever there were a time to get seriousabout the coming revolution in military affairs, it is now. There is anemerging consensus that the United States’ top defense planning priority shouldbe contending with great powers with advanced militaries, primarily China, and that new technologies, onceintriguing but speculative, are now both real and essential to future militaryadvantage. Senior military leaders and defense experts are also starting toagree, albeit belatedly, that when it comes to these threats, the United Statesis falling dangerously behind.

This reality demands more than a revolution intechnology; it requires a revolution in thinking. And that thinking must focusmore on how the U.S. military fights than with what it fights. The problem isnot insufficient spending on defense; it is that the U.S. military is beingcountered by rivals with superior strategies. The United States, in otherwords, is playing a losing game. The question, accordingly, is not how newtechnologies can improve the U.S. military’s ability to do what it already doesbut how they can enable it to operate in new ways. If American defenseofficials do not answer that question, there will still be a revolution inmilitary affairs. But it will primarily benefit others.

It is still possible for the United States toadapt and succeed, but the scale of change required is enormous. Thetraditional model of U.S. military power is being disrupted, the way Blockbuster’sbusiness model was amid the rise of Amazon and Netflix. A military made up ofsmall numbers of large, expensive, heavily manned, and hard-to-replace systemswill not survive on future battlefields, where swarms of intelligent machineswill deliver violence at a greater volume and higher velocity than ever before.Success will require a different kind of military, one built around largenumbers of small, inexpensive, expendable, and highly autonomous systems. TheUnited States has the money, human capital, and technology to assemble thatkind of military. The question is whether it has the imagination and theresolve.

NEW TECHNOLOGIES, OLD PROBLEMS

Artificial intelligence and other emergingtechnologies will change the way war is fought, but they will not change its nature.Whether it involves longbows or source code, war will always be violent,politically motivated, and composed of the same three elemental functions thatnew recruits learn in basic training: move, shoot, and communicate.

Movement in warfare entails hiding and seeking(attackers try to evade detection; defenders try to detect them) andpenetrating and repelling (attackers try to enter opponents’ space; defenderstry to deny them access). But in a world that is becoming one giant sensor,hiding and penetrating—never easy in warfare—will be far more difficult, if notimpossible. The amount of data generated by networked devices, the so-called Internet of Things, is on pace to triplebetween 2016 and 2021. More significant, the proliferation of low-cost,commercial sensors that can detect more things more clearly over greaterdistances is already providing more real-time global surveillance than hasexisted at any time in history. This is especially true in space. In the past,the high costs of launching satellites required them to be large, expensive,and designed to orbit for decades. But as access to space gets cheaper,satellites are becoming more like mobile phones—mass-produced devices that areused for a few years and then replaced. Commercial space companies are alreadyfielding hundreds of small, cheap satellites. Soon, there will be thousands ofsuch satellites, providing an unblinking eye over the entire world. Stealthtechnology is living on borrowed time.

On top of all of that, quantumsensors—which use the bizarre properties of subatomic particles,such as their ability to be in two different places at once—will eventually beable detect disruptions in the environment, such as the displacement of airaround aircraft or water around submarines. Quantum sensors will likely be thefirst usable application of quantum science, and this technology is still manyyears off. But once quantum sensors are fielded, there will be nowhere to hide.

The future of movement will also be characterizedby a return of mass to the battlefield, after many decades in which the trendwas moving in the opposite direction—toward an emphasis on quality overquantity—as technology is enabling more systems to get in motion and stay inmotion in more places. Ubiquitous sensors will generate exponentially greaterquantities of data, which in turn will drive both the development and thedeployment of artificial intelligence. As machines become more autonomous,militaries will be able to field more of them in smaller sizes and at lowercosts. New developments in power generation and storage and in hypersonicpropulsion will allow these smaller systems to travel farther and faster thanever. Where once there was one destroyer, for example, the near future couldsee dozens of autonomous vessels that are similar to missile barges, ready tostrike as targets emerge.

Technology will also transform how those systemsremain in motion. Logistics—the ability to supply forces with food, fuel, andreplacements—has traditionally been the limiting factor in war. But autonomousmilitaries will need less fuel and no food. Advanced manufacturing methods,such as 3-D printing, will reduce the need for vast, risky, and expensivemilitary logistics networks by enabling the production of complicated goods atthe point of demand quickly, cheaply, and easily.

In an even more profound change, space will emerge as its own domain of maneuverwarfare. So far, the near impossibility of refueling spacecraft has largelylimited them to orbiting the earth. But as it becomes feasible to not justrefuel spacecraft midflight but also build and service satellites in space,process data in orbit, and capture resources and energy in space for use inspace (for example, by using vast solar arrays or mining asteroids), spaceoperations will become less dependent on earth. Spacecraft will be able tomaneuver and fight, and the first orbital weapons could enter the battlefield.The technology to do much of this exists already.

THE MILITARIES OF TOMORROW

Technology will also radically alter howmilitaries shoot, both literally and figuratively. Cyberattacks, communication jamming,electronic warfare, and other attacks on a system’s software will become asimportant as those that target a system’s hardware, if not more so. The rate offire, or how fast weapons can shoot, will accelerate rapidly thanks to newtechnologies such as lasers, high-powered microwaves, and other directed-energyweapons. But what will really increase the rate of fire are intelligent systemsthat will radically reduce the time between when targets can be identified andwhen they can be attacked. A harbinger of this much nastier future battlefieldhas played out in Ukraine since 2014, where Russia has shortened tomere minutes the time between when their spotter drones first detect Ukrainianforces and when their precision rocket artillery wipes those forces off themap.

The militaries that embrace andadapt to these technologies will dominate those that do not.

The militaries of the future will also be able toshoot farther than those of today. Eventually, hypersonic munitions (weaponsthat travel at more than five times the speed of sound) and space-based weaponswill be able to strike targets anywhere in the world nearly instantly.Militaries will be able to attack domains once assumed to be sanctuaries, suchas space and logistics networks. There will be no rear areas or safe havensanymore. Swarms of autonomous systems will not only be able to find targetseverywhere; they will also be able to shoot them accurately. The ability tohave both quantity and quality in military systems will have devastatingeffects, especially as technology makes lethal payloads smaller.

Finally, the way militaries communicate willchange drastically. Traditional communications networks—hub-and-spokestructures with vulnerable single points of failure—will not survive. Instead,technology will push vital communications functions to the edge of the network.Every autonomous system will be able to process and make sense of the informationit gathers on its own, without relying on a command hub. This will enable thecreation of radically distributed networks that are resilient andreconfigurable.

Technology is also inverting the current paradigmof command and control. Today, even a supposedly unmanned system requiresdozens of people to operate it remotely, maintain it, and process the data itcollects. But as systems become more autonomous, one person will be able tooperate larger numbers of them single-handedly. The opening ceremonies of the2018 Winter Olympics, in South Korea, offered a preview of this technology when1,218 autonomous drones equipped with lights collaborated to form intricate picturesin the night sky over Pyeongchang. Now imagine similar autonomous systems beingused, for example, to overwhelm an aircraft carrier and render it inoperable.

Further afield, other technologies will changemilitary communications. Information networks based on 5G technology will becapable of moving vastly larger amounts of data at significantly faster speeds.Similarly, the same quantum science that will improve military sensors willtransform communications and computing. Quantum computing—the ability to usethe abnormal properties of subatomic particles to exponentially increaseprocessing power—will make possible encryption methods that could beunbreakable, as well as give militaries the power to process volumes of dataand solve classes of problems that exceed the capacity of classical computers.More incredible still, so-called brain-computer interface technology is alreadyenabling human beings to control complicated systems, such as roboticprosthetics and even unmanned aircraft, with their neural signals. Put simply,it is becoming possible for a human operator to control multiple drones simplyby thinking of what they want those systems to do.

Put together, all these technologies willdisplace decades-old, even centuries-old, assumptions about how militariesoperate. The militaries that embrace and adapt to these technologies willdominate those that do not. In that regard, the U.S. military is in bigtrouble.

A LOSING GAME

Since the end of the Cold War, the United States’approach to projecting military force against regional powers has rested on aseries of assumptions about how conflicts will unfold. The U.S. militaryassumes that its forces will be able to move unimpeded into forward positionsand that it will be able to commence hostilities at a time of its choosing. Itassumes that its forces will operate in permissive environments—thatadversaries will be unable to contest its freedom of movement in any domain. Itassumes that any quantitative advantage that an adversary may possess will beovercome by its own superior ability to evade detection, penetrate enemydefenses, and strike targets. And it assumes that U.S. forces will suffer fewlosses in combat.

These assumptions have led to a force builtaround relatively small numbers of large, expensive, and hard-to-replacesystems that are optimized for moving undetected close to their targets,shooting a limited number of times but with extreme precision, andcommunicating with impunity. Think stealth aircraft flying right into downtownBelgrade or Baghdad. What’s more, systems such as these depend oncommunications, logistics, and satellite networks that are almostentirely defenseless, because they were designed under the premise that noadversary would ever be able to attack them.

This militaryenterprise and its underlying suppositions are being called into question. Forthe past two decades, while the United States has focused on fighting wars inthe Middle East, its competitors—especially China, but also Russia—have beendissecting its way of war and developing so-called anti-access/area-denial (orA2/AD) capabilities to detect U.S. systems in every domain and overwhelm themwith large salvos of precision fire. Put simply, U.S. rivals are fielding largequantities of multimillion-dollar weapons to destroy the United States’multibillion-dollar military systems.

China has alsobegun work on megaprojects designed to position it as the world leader inartificial intelligence and other advanced technologies. This undertaking isnot exclusively military in its focus, but every one of theseadvanced-technology megaprojects has military applications and benefits the People’sLiberation Army under the doctrine of “military-civil fusion.” Whereas the U.S.military still largely treats its data like engine exhaust—a uselessbyproduct—China is moving with authoritarian zeal to stockpile its data likeoil, so that it can power the autonomous and intelligent military systems itsees as critical to dominance in future warfare.

The United States’position, already dire, is rapidly deteriorating. As a 2017 report from theRAND Corporation concluded, “U.S. forces could, under plausible assumptions,lose the next war they are called upon to fight.” That same year, GeneralJoseph Dunford, chairman of the Joint Chiefs of Staff, sounded the alarm instark terms: “In just a few years, if we do not change the trajectory, we willlose our qualitative and quantitative competitive advantage.”

The greatestdanger for the United States is the erosion of conventional deterrence. Ifleaders in Beijing or Moscowthink that they might win a war against theUnited States, they will run greater risks and press their advantage. They willtake actions that steadily undermine the United States’ commitments to itsallies by casting doubt on whether Washington would really send its military todefend the Baltics, the Philippines, Taiwan, or even Japan or South Korea. Theywill try to get their way through any means necessary, from coercive diplomacyand economic extortion to meddling in the domestic affairs of other countries.And they will steadily harden their spheres of influence, turning them intoareas ever more hospitable to authoritarian ideology, surveillance states, andcrony capitalism. In other words, they will try, as the military strategistSun-tzu recommended, to “win without fighting.”

THE FUTURE IS HERE

The United Statesis still betting that by incrementally upgrading its traditional militarysystems, it can remain dominant for decades to come. This approach might buytime, but it will not allow the U.S. military to regain superiority over itsrivals. Doubling down on the status quo is exactly what Washington’scompetitors want it to do: if the U.S. government spends more money in the sameways and on the same things, it will simply build more targets for itscompetitors while bankrupting itself.

Washington must banish the idea that thegoal of military modernization is simply to replace the military platforms ithas relied on for decades. 

 

It’s time to thinkdifferently, and U.S. defense planners should start by adopting more realisticassumptions. They should assume that U.S. forces will fight in highly contestedenvironments against technologically advanced opponents, that they will beunlikely to avoid detection in any domain, and that they will lose largenumbers of military systems in combat. Washington must also banish the ideathat the goal of military modernization is simply to replace the militaryplatforms it has relied on for decades, such as fighter jets and aircraftcarriers, with better versions of the same things. It must focus instead on howto buy systems that can be combined into networks or kill chains to achieveparticular military outcomes, such as air superiority or control of the seas.Finally, the old belief that software merely supports hardware must beinverted: future militaries will be distinguished by the quality of theirsoftware, especially their artificial intelligence.

What would amilitary built on those assumptions look like? First, it would have largequantities of smaller systems: swarms of intelligent machines that distributesensing, movement, shooting, and communications away from vulnerable singlepoints of failure and out to the edges of vast, dispersed networks. Such anapproach would impose costs on competitors, as they would no longer be able toconcentrate on a few big targets and would instead need to target many thingsover larger spaces.

Second, thosesystems would be cheap and expendable, which would make it easier to endurelarge-scale losses in combat. If it takes the United States’ competitors moretime and money to destroy U.S. systems than it does for the United States toreplace those systems, the United States will win over time.

Finally, thesesystems would be unmanned and autonomous to the extent that is ethicallyacceptable. Keeping humans alive, safe, and comfortable inside machines isexpensive—and no one wants to pay the ultimate price of lost human life.Autonomous systems are cheaper to field and cheaper to lose. They can also freehumans from doing work that machines can do better, such as processing rawsensor data or allocating tasks among military systems. Liberating people fromsuch work will prove crucial for managing the volume and velocity of the modernbattlefield, but also for enabling people to focus more energy on making moraldecisions about the intended outcomes of warfare. In this way, greater autonomycan not only enhance military effectiveness; it can also allow more humans topay more attention to the ethics of war than ever before.

Building this kindof military is not only desirable; it is becoming technologically feasible. TheU.S. military already has a number of programs in development aimed at justsuch a future force, from low-cost autonomous aircraft to unmanned underwatervehicles that could compose an artificially intelligent network of systems thatis more resilient and capable than traditional military programs. For now, noneof these systems is as capable as legacy programs such as the F-35 Joint StrikeFighter or the Virginia-class submarine, but they also carry a small fractionof the costs. The goal should be not to buy more individual platforms but to buyfaster kill chains. The money currently invested in one legacy system could buydozens of autonomous systems that add up to a superior capability.

The purpose ofthis kind of military—one that relies heavily on swarms of thousands of small,low-cost, autonomous systems that can dominate all domains—would not be toprovoke war. It would be to deter it, by demonstrating that the United Statescan destroy any force its competitors put onto the battlefield in any domain,replenish its combat losses faster and cheaper than they can, and sustain afight until it wins by attrition. The purpose of preparing for war will remainto never have to fight one.

A FAILURE OF IMAGINATION

Militarymodernization of this kind will not happen all at once. Autonomous systems mayrely on legacy systems, including aircraft carriers, for some time to come. Buteven this will require significant changes to how traditional systems areconfigured and operated. Some leaders in Congress and the executive branch wantto embrace these changes, which is encouraging. But if this transitionfails—and the odds of that are unsettlingly high—it will likely fail forreasons other than the ethical opposition that is the focus of current debates,which seeks to “ban killer robots” or ensure that commercial technologycompanies do nothing to benefit the U.S. military.

There are seriousethical concerns. The military use of advanced technologies such as artificialintelligence requires sober debate, but that debate should not be reduced to abinary decision between human and machine control. If framed clearly, many ofthe technological and moral questions facing policymakers can be answeredwithin the confines of existing law and practice. For example, the legalconcept of “areas of active hostilities,” in which the threshold for usingviolence is reduced in limited geographic areas, could provide useful answersto the moral dilemmas posed by lethal autonomous weapons.

The real challengefacing policymakers is how to imbue intelligent machines with human intent, andthat is not a new problem. And although this new technology will present ethicaldilemmas, it will also help resolve them.Autonomous systems will enable humans to spend less time on menial problems andmore time on moral ones. Intelligent machines will likely become more capableof differentiating between, say, tanks and other vehicles, than a scared19-year-old is. Americans will naturally be apprehensive about trustingmachines to perform what have traditionally been human tasks. But the greaterdanger right now is that Americans will move too slowly and not be trustingenough, especially as China and Russia are proceeding with fewer ethicalconcerns than the United States. Unless Washington is willing to unilaterallycede that advantage to its rivals, it cannot allow itself to become paralyzedby the wrong questions.

If the UnitedStates fails to take advantage of the new revolution in military affairs, itwill be less for ethical reasons and more as a result of the risk-averse,status quo mentality that pervades its domestic institutions. Former U.S.Secretary of Defense Robert Gates explained why in his memoir, Duty:

The military departments develop theirbudgets on a five-year basis, and most procurement programs take many years—ifnot decades—from decision to delivery. As a result, budgets and programs arelocked in for years at a time, and all of the bureaucratic wiles of eachmilitary department are dedicated to keeping those programs intact and funded.They are joined in those efforts by the companies that build the equipment, theWashington lobbyists that those companies hire, and the members of Congress inwhose states or districts those factories are located. Any threats to thoselong-term programs are not welcome.

 

This is whatSenator John McCain, a Republican from Arizona, once called “themilitary-industrial-congressional complex,” and its entire livelihood dependson developing, producing, acquiring, operating, and maintaining traditionaldefense systems in traditional ways.

Some in thiscomplex may seem welcoming to advanced technologies now because they stilldon’t view them as threats. For a transitional period, advanced technologieswill indeed support, rather than replace, traditional systems. But as thebackers of traditional systems come to see intelligent machines as substitutesfor those systems, they will resist change. Bureaucrats who derive power fromtheir mastery of the current system are loath to alter it. Military pilots andship drivers are no more eager to lose their jobs to intelligent machines thanfactory workers are. Defense companies that make billions selling traditionalsystems are as welcoming of disruptions to their business model as the taxi cabindustry has been of Uber and Lyft. And as all this resistance inevitablytranslates into disgruntled constituents, members of Congress will haveenormous incentives to stymie change.

Overcoming theseobstacles will require leadership at the highest levels of government to setclear priorities, drive change in resistant institutions, remake theirincentive structures, and recast their cultures. That may be too much toexpect, especially amid Washington’s current political turmoil. There are manycapable, well-intentioned leaders in the Pentagon, Congress, and the privatesector who know that the U.S. defense program needs to change. But too often,the leaders who understand the problem the best lack the power to address it atthe scale required, while those with the most power either don’t understand theproblem or don’t know what to do about it.

This points to abroader problem: a fundamental lack of imagination. U.S. leaders simply do notbelieve that the United States could be displaced as the world’s preeminentmilitary power, not in the distant future but very soon. They do not have thevision or the sense of urgency needed to alter the status quo. If that attitudeprevails, change could come not from a concerted plan but as a result of acatastrophic failure, such as an American defeat in a major war. By then,however, it will probably be too late to alter course. The revolution inmilitary affairs will have been not a trend that the United States used todeter war and buttress peace but a cause of the United States’ destruction.