network cyber
Army Offensive Cyber Brigade soldiers during an exercise at Fort Lewis, Washington in 2015. (U.S. Army photo)
WASHINGTON — In the words of one U.S. Army general, Ukraine has become a “command graveyard.” In the age of drones, large command centers have become major targets, and Russia in particular has been losing generals one after another to precision attacks.
Here, a war between the United States, which has a large joint air operations center (CAOC), and China, whose system destruction warfare doctrine sees these high-tech sites as pressure points for attack with long-range missiles and cyberattacks, could be argued. Let’s imagine.
“You can’t put 100 staff members, non-commissioned officers, etc. in a tent and hope they don’t get killed,” said Alexander Kott, an AI expert who has worked for DARPA and the Army. “To this day, our command post looks like a lot of vehicles, a lot of antennas, a lot of tents. We know how long this will last on the Russian-Ukrainian front, right? A few minutes. will survive.”
“The Western system today is very hub-and-spoke structured,” agrees Sidharth Kaushal, a research fellow at the Royal United Services Institute in London. “Processing happens at a central node, for example a CAOC, and these systems are inherently vulnerable. And adversaries understand this. [hence] It’s a system-destroying war. ”
However, Kaushal told reporters at a press conference in January that AI has the potential to strengthen the chain of command. “This means moving to the edge more and more significantly.”
Centralized management and decentralization to the “tactical edge” have different advantages and disadvantages, require different forms of AI to be employed, and above all, different organizational cultures. These cultural factors may give America a lasting advantage in an era when new technology, especially software, spreads rapidly around the world.
But that is no reason for America to be complacent. There are other models for modernization than the U.S., the Pentagon has its own problems with top-down bureaucracy, and a close reading of Chinese military literature suggests that at least some officers are seeking more flexibility. Experts warn that it is clear that
“The Nazis were quite authoritarian, but that didn’t prevent them from being innovative when necessary,” said Dean Chen, senior adviser for China at the U.S. Institute of Peace. “They invented rockets, assault rifles, and blitzkrieg. So I think the idea that ‘authoritarians can’t innovate’ is very ahistorical.”
As far as U.S. innovation goes, Cott, who retired last year as chief scientist at the Army Research Institute, said, “Unfortunately, I’m not very impressed with what we’re doing.”
“It seems like we’re just building increasingly expensive and complex piped communication systems,” he says.
Related: Pacific Fleet Chief Paparo says the big lesson China learned from Ukraine: Win quickly.
Meanwhile, as in the former Soviet Union and, more specifically, in interwar Germany, there is a surprisingly lively debate about military theory in authoritarian China.
“If you look at their writing on so-called ‘informatization,’ you see a very strong emphasis on pushing authority down to the lowest level,” Kevin Polpeter, senior research fellow at CNA, told Breaking Defense. Told. “Whether that actually happens is debatable.”
“At least some of the PLA’s writing envisions what AI and emerging technologies more broadly can do to inform decision-making for frontline commanders,” said Randland researcher Nathan. Beecham Mustafaga agrees. “The question is whether these ideas actually have any weight. [with] Chinese military leadership. It’s very difficult to assess that from the outside…I think it’s too early to judge. ”
Combined Air Operations Center (CAOC) at Al Udeid Air Base, Qatar. (U.S. Air Force photo by Staff Sergeant Alexander W. Riedel)
center and edge
Centralized control and empowerment at the edge is not an either/or, but a sliding scale, with forces constantly moving up and down as technology and strategy evolve.
“In reality, it’s always a mix,” Kot says. Meanwhile, he told Breaking Defense: [for instance]Two soldiers with drones can do a lot of damage…so that would be an argument for bringing decision-making closer to the edge. ”
At the same time, wars are becoming increasingly complex and widespread, Kot continued. Weapons such as long-range missiles and cyberattacks can be focused on a single target from thousands of miles away, if someone can coordinate them. “It has to be done at a higher level,” he said.
Given the power of AI, networks, and surveillance technology to bring vast amounts of data to a single headquarters, “there’s a perfectly consistent argument.” [for] He is a fully empowered theater commander,” Kaushal agreed. “If all goes well, they can be very deadly.”
Therefore, one option is to use AI to further centralize command and control. The idea is to build a super CAOC with supercomputers that allows top commanders and their staffs to algorithmically absorb vast amounts of data from across the war zone and view it with God’s eyes. . While these palatial panopticons are appealing to bureaucratic organizations, they also tap into powerful trends in how technology itself has evolved.
For example, cloud computing, which the Department of Defense has been trying to implement, boils down to a hub-and-spoke structure. Increasingly, a small number of large server farms store data for many users and also run software. We have made power-hungry generative AI, such as ChatGPT and DALL-E, that require vast amounts of computing power to run, available to millions of people on their laptops and smartphones. This is a hub-and-spoke model. A central warehouse full of servers can run AI that is too powerful for a single user’s machine, make the most efficient use of scarce technical talent, and gain other economies of scale.
Therefore, in peacetime, centralization makes a lot of sense. It also works well in war, allowing it to coordinate attacks from far-flung units that would otherwise be difficult to act in coordination. But as soon as the enemy is able to fight back, centralization means he puts a lot of delicate eggs in one big basket.
“The challenge facing these structures is that they are inherently vulnerable to ideological adversaries,” Kaushal told reporters. “Given the fundamentals of war, going back to Clausewitz, that it is characterized by fog, friction and a certain degree of confusion, it seems to me that there is a strong possibility that the initiative needs to be pushed to a lower level. ”
Even if the command center itself is a difficult target, buried underground, surrounded by missile defenses, or simply out of range deep inside mainland China, targets that are not, i.e. aircraft, warships, ground Must communicate with troops (both robots and ground forces). Unmanned. To get anything done, central command requires hundreds or thousands of connections with distributed subordinates. This is to feed data and send orders to central command. If the link is sabotaged, hacked, or physically blown up, the commander’s divine perspective on the battlefield becomes deaf and blind.
That is the essence of the PLA’s “system destruction warfare” doctrine, Chen said. There is a possibility of jamming or cyber attacks. …It’s also psychological. How much can you trust his communications if they change every third message?”
AI can help here too. “How do you communicate when the battlefield is full of jammers?” Kot asked. “Squeezing all the vitamins out of the little products that are available is another role for AI.”
Autonomous algorithms already manage mobile phone networks, optimizing frequencies and switching calls from one cell phone tower to another without the user knowing. Military AI can automatically detect when a relay has been sabotaged, hacked, or vandalized and work around it. This is an enhanced version of the so-called “self-healing” network. At a higher level of intelligence, Kott suggested that AI could also limit usage based on pre-set priorities. For example, denying the transmission of bandwidth-consuming surveillance video to free up transmission capacity for telegram updates of target coordinates.
However, AI is not only useful in central hubs, it is also becoming increasingly available and powerful at the edge. This dovetails nicely with Starlink putting thousands of communications satellites into low-Earth orbit, allowing front-line users to communicate without relying on a small number of high-end, high-orbit birds. For example, advanced analytics company Royce GEO is using Starlink in field tests to download satellite images to portable computers running a slimmed-down version of the company’s object recognition AI.
“Starlink was game-changing,” said Dave Sterling, a former NGA analyst and founder of Royce GEO. “Before, you had to pack a 160-pound antenna and it took three men to carry it. … Everything has changed in the last few years. one of them will be in charge of it. [with] Rugged battery pack to run machine learning [algorithms] We’re leveraging computer vision from the comfort of a pickup truck in some of the harshest environments around the world. ”
People’s Liberation Army Honor Guard soldiers stand at attention in front of a poster of President Xi Jinping near Tiananmen Square in Beijing during the 2020 coronavirus outbreak. (Photo by Kevin Frayer/Getty Images)
new techniques, new tactics
Throughout history, from the Bronze Age to the Blitzkrieg, victorious commanders have used new technologies not just to fight better the old ways, but to enable entirely new kinds of tactics. For example, both China and the United States are considering AI-based decision-making support that can generate “courses of action.” Essentially, it automates the arduous staff work of tracking supplies, scheduling convoys, coordinating strikes, drafting orders, and determining countless other details. Enter military planning.
“If you look at PLA writing on AI, a lot of it focuses on thinking about AI as like a chief of staff. If the AI is in a command post somewhere, you can give it a list of parameters. You can…and it will give the commander a choice, but ultimately it’s the commander who makes those decisions.”
This suggests the potential for forward command posts to use AI to plan complex operations without the need for large personnel, and to attack large-scale operations without becoming large, slow-moving targets. Kaushal said this suggests that the command could gain coordination capabilities.
But will military commanders, and more importantly, their political leaders, allow junior officers to exercise this kind of AI-driven initiative? In Washington, let alone Beijing? It’s not easy to sell that.
“What people often overlook or don’t realize is that the People’s Liberation Army is the party’s army. It’s not the country’s army, it’s the armed wing of the Chinese Communist Party,” Polpeter said. “If anything…the role of political commissars has been strengthened under the Xi Jinping administration.”
While the XI purges senior generals, the PLA’s own public self-criticism is causing intense self-doubt. “Especially when you look at their own writing about soldiers and officers,” Kaushal said.[there’s] What they call the “Five Incompetencies” are five areas in which they still feel there are significant deficiencies. One of these is the ability of junior officers to perform their duties under conditions of uncertainty, fog, and friction. ”
A lack of trust in subordinates suggests that China’s top leaders are reluctant to delegate, a weakness that Western powers can exploit.
