Artificial Intelligence and Big Data, two buzzwords that are colloquially interchangeable but subtly nuanced, are not silver bullets poised to handily solve all of the US military’s problems.
Unpopular opinion: the US military and the defense industrial complex are currently giving up one heck of a run to the Chinese Communist Party. Note how I didn’t say that we’re losing… yet.
This century saw a solid first quarter, with US domination that witnessed the rise and fall of a competing Soviet Union and the establishment of global American hegemony both militarily and economically. We have enjoyed decades of unipolar dominance. When you’re on top, it typically feels like you could never lose.
However, the rapidly shifting political landscape and return to great power competition have America reeling. The Chinese Communist Party and its People’s Liberation Army are mounting a comeback.
While we were worried about cracking terrorist networks of bearded men with AK-47s in caves, the Chinese were speeding towards advanced technologies, hypersonic weapons, and the very defenses required to put up a front against the predictable American military machine.
The Chinese have seized this strategic opportunity. While the West was distracted, Beijing sunk billions of dollars into anti-access and area denial capabilities (A2/AD), a defensive posture aimed at the American way of fighting. They have also amassed massive amounts of data required to weaponize and harness the benefits of “Big Data.”
Chinese tech companies and government-sponsored research initiatives have built massive data sets while we were preoccupied with Iraq and Afghanistan. These are precisely the requisite data sets necessary to train Machine Learning algorithms and AI neural networks.
While we were building social networks by word of mouth of terrorist cells, the Chinese were collecting intelligence and building advanced systems for moving data with little regard for civil liberties, privacy, or data protection. Not that I advocate for it, but it is amazing what you can do when you ignore ethics or societal norms.
In defense tech news, all I read about is AI solving the joint, all-domain command and control problem, or Big Data providing a potential solution for some multi-domain capability gap. Perhaps we just desire an easy, one size fits all solution in the form of a Big Data Band-Aid?
Indeed, it seems like our greatest adversary and the second greatest existential threat to the American way of life after nuclear war has already found the elixir of life in Big Data, so why can’t we?
Artificial Intelligence’s Combat Capabilities
For starters, artificial intelligence is not the Terminator. It is not a killing machine that is easily weaponized, deployed, and employed to combat adversary capabilities. Even the most cutting edge artificial intelligence tools today are narrow in scope and limited in application.
While that will change eventually, algorithms are currently fantastic for vehicle routing, search engine optimization, facial recognition, asking Siri to set a timer, and other modern technical conveniences that we all carry around in our pockets. These are simple applications of AI. These are not weaponized, military applications that result in warheads on foreheads or power projection.
AI is great at parsing through billions of bits quickly and making sense of it all; creating information from data is its strong suit. This is not complicated. At their core, these algorithms rely on data configuration and formatting to sort and shape vast matrices full of different variables, perform some sort of reduction or matrix operation, and compare this reduction to a set of user-programmed decision criteria.
There is a difference between artificial intelligence and decision making. AI facilitates expedited data to decision throughput, but it does not make its own decisions in a vacuum.
Next, AI is slightly more complicated versions of the matrix math you were probably introduced to in algebra. This advanced linear algebra is advanced applied statistics. By itself, it does not result in a major weapons system delivering effects against an adversary position. Just like space and cyber effects at your favorite large force exercise, you cannot simply sprinkle some “Big Data” on top and bring added military capability to bear to win the 21st-century fight.
On their own, AI and Big Data do not result in increased competition by the US military. They don’t produce a capability to which the Chinese Communist Party has no solution. While they can expedite paths through a particular kill-web to deliver effects, they aren’t a standalone military capability.
Another reason why AI and Big Data won’t solve the A2/AD problem is because of the laws of physics. The US Indo-Pacific Command Area of Responsibility poses a geography problem for the US military. It requires ships and airplanes to travel farther to even get to the fight. Missiles can only go so far and fast, and AI does not provide a solution that creates a silver bullet hypersonic solution.
A2/AD is also a logistics nightmare. Posturing the supplies and equipment at disparate operating locations anywhere from the Philippines to Guam or to Alaska to support even a limited regional conflict is a hard nut to crack, and AI does not by itself solve the agile, global logistics problem.
I might sound exceptionally contrarian in my simplification of AI and Big Data. In truth, I’m a huge proponent of defense applications for AI and Big Data. Our military’s future hinges on it.
For the Department of Defense (DoD) to harness AI and to weaponize Big Data, the US military machine and industrial base need to integrate artificial intelligence into military systems.
The current generation of developmental systems needs to bake in advanced algorithms to take the human brain as a data filter out of the loop while introducing fusion, machine/deep learning, and the power of computation to military applications.
The old way of filtering data and enabling the military operator’s tactical decision making is irrelevant today. If the DoD can’t shift, adapt, and embrace this change, they’re doomed to fight the last war for the rest of this century.
Challenges in Weaponizing Artificial Intelligence Capabilities
The DoD, like many contemporary large organizations, will face many hurdles in weaponizing artificial intelligence capabilities.
One of the main challenges in this transition is simple integration. That’s something the DoD already isn’t good at. To abuse an overused example, the F-22 and F-35, arguably the world’s most advanced tactical fighters, cannot communicate via their tactical data links. While they were both developed by Lockheed Martin, their data links use different standards for their waveforms and are not interoperable. To oversimplify two prodigiously complex weapon systems, the F-22 is using AM radio and the F-35 uses FM.
This is partially the government’s fault but also the fault of the big defense contractors. Back to my data link example: in the 21st century, these capabilities are software-driven. However, major defense contractors are hardware companies.
During the early years of the American century, they mobilized and bent metal to create some of the last generation’s most capable machines. That said, they have a comparative advantage only in producing hardware, not in the software required to fight in the 21st century.
For the DoD to be successful in harnessing AI for the next conflict, it needs to foster relationships with organizations that operate in the tech space with a mastery of software development featuring AI applications.
The key is integrating AI and Big Data capabilities into military applications of all kinds, across the full spectrum of military operations. Global logistics, command and control, persistent ISR, and advanced weapons are untapped applications for AI that have not yet been touched by the tech space.
The traditional bloated defense contractor is not resourced for this, nor do they have the right skill sets. Only seasoned developers outside the typical defense industrial base have the know-how to actually succeed with this integration.
AI alone won’t compete with Chinese military capabilities. Applying the tenets of big data and weaponizing it to field advanced and lethal military capabilities is the future of power competition.
The Chinese are catching up and may one day challenge American global military dominance, but applied AI capabilities and advanced data science just might be the key to preserving American hegemony and protecting American interests domestically and abroad.
Alex Hillman is an analyst and engineer in the defense space. A US Air Force Academy graduate, he holds master’s degrees in operations research, systems engineering, and flight test engineering, and has previously served in various technical and leadership roles for the USAF. Alex is a graduate of the United States Air Force Test Pilot School and a former US Department of State Critical Language Scholar for Russian.
Disclaimer: The views and opinions expressed here are those of the author and do not necessarily reflect the editorial position of The Defense Post.
The Defense Post aims to publish a wide range of high-quality opinion and analysis from a diverse array of people – do you want to send us yours? Click here to submit an op-ed.