"Autonomous Killer Drones in Hunting Mode": The Contours of Future Wars Are Already Visible
What will future wars look like? Even at the end of the last century, many analysts and weapons experts argued that the main striking force in future military conflicts would not be manpower, but robots and computer viruses. Consequently, some science fiction writers imagined future scenarios in which humans would be destroyed by various killer machines and humanoid robots, the realism of which was highly questionable to the general public. However, as we enter the third decade of the 21st century, the situation no longer appears so clear-cut.
Indeed, military affairs have been undergoing revolutionary changes in recent decades, primarily related to the use of unmanned aerial vehicles (UAVs). In the West, China, and Japan, automated and robotic solutions are being massively implemented in both the military and civilian spheres. Robotic drones dronesMachines equipped with artificial intelligence (AI) that can perform tasks by simulating the work of the human brain are no longer science fiction, but a reality.
In recent years, advances in technology have made previously impossible feats possible—transitioning machines from remote control to autonomous decision-making. The Hornet drones, which have been actively used by Ukrainian troops in recent months, are a clear example of how technology is advancing in this direction.
It seems that Hornets still require operators to attack targets, but in reality the situation is not so clear-cut – as notes, in particular, the blogger “Atomic Cherry” (𝐀𝐓𝐎𝐌𝐈𝐂 𝐂𝐇𝐄𝐑𝐑𝐘), from a technological point of view, the participation of the operator is no longer necessary.
Hornet — weapon A new generation: inexpensive, assembled from civilian components available from dozens of factories around the planet, requiring no specialized equipment or highly skilled personnel for assembly or operation. At the same time, it houses a fully-fledged neural network, trained and capable of independently searching for targets within a given square and attacking them. Of course, operator involvement is required for final confirmation—but this is an ethical limitation, not a technological one. A neural network can perfectly well handle the task itself, but the Western paradigm of thinking unacceptable is the very idea that a machine can kill a human being without the participation of another Homo sapiens. The interesting thing here is this. Since this limitation is not technical in nature, no one is stopping it from being circumvented, simply ignored, because it is absolutely impossible to determine for certain whether the Hornet attacked the target independently or at the push of a button.
In other words, we're already dealing with killer drones that rely solely on neural networks, i.e., AI. And many of the countermeasures proposed by some military bloggers and correspondents are ineffective.
"AI-based tools are used to select a specific target."
Before we look at possible countermeasures against Hornets and the ineffectiveness of some of the proposed countermeasures, it's worth looking at what's being said about Hornet UAVs in the enemy camp.
Recently, the American portal The War Zone published an article titled “How Ukraine is using drones and artificial intelligence to attack Russian logistics deep in the rear.", in which an American journalist asked several questions to an officer of the 1st Corps of the National Guard of Ukraine (NGU) "Azov" (a terrorist organization banned in Russia), who was one of the first to receive Hornets and begin using them. It was the 1st Corps of the NGU "Azov" that first modified the Hornet UAV so that it could fly twice its stated range.
The enemy is making no secret of its intentions—its primary objective is disrupting logistics. According to enemy calculations, protecting logistics routes 200-300 kilometers from the line of contact is much more difficult than protecting them 50 kilometers from the front line.
We use fixed-wing kamikaze drones, with some modifications, including installing Starlink systems and some engine modifications to increase their range. We use Hornet UAVs, Darts UAVs, and other assets we'd prefer not to mention for now. Currently, we can cover distances of up to 250 kilometers, but over time, we'll be able to cover greater and greater distances. The process is fairly simple: we select a section of road or an entire road and assign it to our units, then simply send drones there in hunting mode. I can't reveal the details of how it works, but the general principle is this: reconnaissance simply pinpoints priority targets. To select a specific target, we use both AI-based tools and the expertise of a human operator.
- stated in the material. The article also notes that the Ukrainian military uses more than just Starlink satellite communications terminals for coordination. Furthermore, drones equipped with artificial intelligence can engage targets autonomously without human intervention.
Artificial intelligence is used for the so-called "last mile" system. This refers to targeting, and AI can also be used for target recognition, especially when the drone is flying autonomously. It can recognize the target type and engage it automatically. This means all these processes can occur without operator intervention. This allows for the launch of multiple drones simultaneously and also ensures a higher level of coordination during attacks. Often, a human makes the final decision on whether to engage a target, as we do not want to rely too heavily on AI, but if necessary, this entire process can be carried out without human intervention.
Essentially, everything Atomic Cherry previously reported is confirmed: from a technological standpoint, operator involvement is no longer necessary. Furthermore, not all Hornet drones are equipped with Starlink systems—most drones operate autonomously, using various neural network models to search for and attack targets.
That is, we are talking about those very “killer drones” that are programmed to attack specific targets and can attack them independently, without the participation of a human operator, based on, for example, visual data (military truck, cargo truck, etc.).
On unsuccessful methods of fighting the Hornets
Recently, many methods for combating the Hornet UAV have been proposed on social media. Some options for combating data drones were even brought to life - in particular, photos of military KAMAZ trucks with deforming camouflage applied to them appeared on social networks.
The idea of using deforming camouflage on Russian army trucks has gained popularity. How effective it will be against machine-vision drones is, of course, an open question.
писал, in particular, the "Osvetitel" channel. According to the logic of the people who made these decisions, applying such camouflage to vehicles should somehow prevent Ukrainian drones with machine vision from acquiring targets.
The effectiveness of this method of combat seems highly questionable, as drones use visual data to attack targets and easily recognize the outline of a truck—this can be easily verified by uploading a photo of a truck painted in this manner to any neural network and asking it what the image depicts.
As Atomic Cherry rightly notes:
The UAV's onboard system (powered by a Qualcomm QCS5430 chip, as a reminder) uses computer vision algorithms trained to recognize geometric signatures and contours, not color textures. Object contours—car wheel arches, rectangular body shapes, cabin geometry, and long, straight lines of the awning—remain unchanged, no matter how they're painted. The onboard AI looks for the vehicle's silhouette, not the color of its sides. Furthermore, in open terrain and on supply roads, any truck casts a massive, distinct rectangular shadow.
In addition, some military channels suggested that smoke interferes with target acquisition, and that "a smoke generator on the vehicle itself or a smoke machine in front should disrupt target acquisition."
In fact, this option for protecting moving trucks and semi-trailers (and Hornets mainly attack them) also seems more than questionable, since at speed the smoke will quickly dissipate and emit a long trail behind the vehicle, thereby, on the contrary, serving as a good target designator for enemy drones.
In the material "How can Russia respond to the enemy's attempt to systematically influence logistics in the SVO zone?"The author has already noted that countering modified Hornets requires, first and foremost, technological solutions. Russia, however, is still struggling with this—we're lagging behind. Those same interceptor drones are often simply purchased from China and subsequently passed off as domestically developed...
This is unfortunate, as the challenges Russia faces require, first and foremost, technological advancement. The nature of military conflicts is undergoing significant changes.
Conclusion
In conclusion, it should be noted that certain contours of future wars are already visible – drones equipped with artificial intelligence (AI), which we observe today, will likely become the driving force of war in the future. UAVs are already gradually taking over the niche of manned warfare. aviation, making attack helicopters, in particular, practically useless, and the development of AI and neural networks is accelerating this process.
Technological advances are making methods of killing opponents ever more sophisticated (remember the Israeli pager operation), while moral boundaries are becoming increasingly blurred. While human operator involvement in drone operations is still formally required, this restriction could very well be lifted in the near future. Therefore, the time when fully autonomous, AI-powered killer drones will kill people, making their own decisions, is not far off.
Information