Tech

AI and the Art of Warfare: How Military Applications of AI are Revolutionizing Combat

War, War never changes. This is a famous line in the role-playing game “Fallout” series, “War, war will never change.”

Since the dawn of humanity, we have pursued with greatest passion the most efficient means of destroying other humans. Engels had a prophecy: “Once technological progress can be used for military purposes and has been used for military purposes, they will immediately almost forcefully, and often against the commander’s will, cause changes or even revolutions in the way of fighting.”

As one of the most important technological breakthroughs of this era, AI cannot escape this curse.

Today, more than two months have passed since the big AI discussion triggered by ChatGPT. After the enthusiasm has faded, countless AI applications are quietly sprouting all over the world, and the process of AI changing the world has just begun. In the next period of time, I will use 4 to 5 issues to describe the application of AI in several important fields. In this issue, let us talk about the military field first.

AI is a dual-use technology that is visible to the blind.

The early research and development of AI is very similar to the Internet, robots, and computers. It is a technology in which the military advances the initial development costs and then bears fruit. Now, if people talk about AI only focusing on its civilian use and ignoring its military potential , it is empty talk.

As one of the most efficient organizational structures in human history, the military has never stopped studying how to kill people most efficiently. The emergence of AI has made key contributions to the real-time strategy and first-person shooting of military operations.

It is no exaggeration to say that AI has entered the military field and has profoundly changed the trajectory of our military development. In the future, it will also change the logic of our war and continue the general mobilization model since the 19th century. After all, AI can empower all the structures of the military from the brain to the limbs, and it can even accelerate military coups .

Some people may ask, it is understandable for AI to control robot soldiers, so where should we start with the military coup?

Let’s start with the AI ​​model that exploded this year.

Since the second half of 2022, what the investment circle and the technology circle are talking about most is the commercial breakthrough of the AI ​​model.

Since October 2022, generative AI based on large models has been launched in a short period of time with various functions ranging from text to pictures, text to video, and human-AI chat, quickly achieving commercial breakthroughs in many fields.

All of a sudden, using AI to train conversations and chats, make video pictures, search for messages and other technologies has entered a new dimension. However, under this peaceful background, the Internet warfare model in the new era has undergone subversive changes.

In March 2023, the news that Trump was about to be arrested blasted the global social network, and at the same time a series of pictures attracted the attention of people all over the world. It was a picture of Trump being arrested and Trump in prison. The Trump inside was in a state of embarrassment, and soon lost the temperament of looking down on the world in the past.

Afterwards, the truth surfaced. This was a picture of Trump’s arrest generated by supporters of the Democratic Party of the United States through AI at that time. The main purpose was to mock the Republican Party, especially Trump’s supporters.

Soon, I mean soon, the counterattack from Trump supporters came. Also generated using AI, such as the arrest pictures of Pelosi, Biden, etc., and the pictures of Trump holding a gun and leading the agents to charge have also exploded on the Internet.

Afterwards, no matter what happens, there will be similar explosive photos. For example, when my elder brother Long visited China, there would be photos of him standing in front of a row of Chinese flags and making a speech.picture

This is where friends want to ask, what does the whole small fight have to do with the military, isn’t it just the advanced version of the navy to swipe the screen.

But in fact, this kind of thing can determine the life and death of a country, because it can conduct cognitive warfare at a very low cost.

Here we give a small example. On the evening of July 15, 2016, a small incident occurred in Turkey. A handful of troops staged a military coup against Erdogan. After they launched an attack on the army’s command center, they quickly took control of the Turkish Radio and Television Association (TRT) TV station in the capital Ankara. The female anchor Tijan Karas mouthed a statement drafted by the coup soldiers to the TV audience. A military group calling itself the “Fatherland Peace Council” said in a statement that the army had taken over power, imposed a nationwide curfew and martial law.

But fortunately, Erdogan knew the news through Putin in advance, and escaped the disaster in the first place. Then what did he do? Instead of immediately calling the Army headquarters or going to the Americans for an explanation, the live broadcast started.

At 0:24 on July 16, Erdogan began to accept an interview with the Turkish channel of CNN through the video chat software FaceTime on his Apple mobile phone. He delivered a speech to the mobile phone camera, calling on the people to take to the streets to protest the coup and “give them (rebel soldiers) answers.” At the same time, Erdogan also spoke on “Twitter”, calling on the people to go to the airport and public squares to “take back democratic ownership and national sovereignty.”

At this time, for the soldiers of the coup, the general situation is over.

Because whether it was the killing of Pukaka or the color revolution in Ukraine, the historical experience of coups has proved one thing. That is, the key to the coup is not what happened to the leadership, but the fact that the majority of people formed a consensus that the coup was successful in a short period of time, and this consensus may have nothing to do with the actual actions of the coup.

So by posting videos on social media and mobilizing his own people to the streets, Erdogan successfully survived the coup, and then launched a wave of purges against those who opposed him.

But, I mean but, if the coup forces had fake news backed by large models, what might they have done?

The core operation of Erdogan’s comeback is the live broadcast on the Internet and videos on social media, mobilizing supporters to take to the streets. And the current large-scale model AI of these things can be forged.

In this way, the coup party can first use the robots supported by the big language model to directly post comments on social media and speak against Erdogan. Then make fake photos and fake videos saying that Erdogan has been caught and the situation is over. This can not only hedge against Erdogan’s social media videos, but also mobilize anti-Erdogan audiences and troops still on the sidelines to join the coup sequence.

In this way, even if Erdogan is not caught in the end, he can be “blown to death”. After all, for a coup, the core is to control the central power, especially the full support of the army, before the opponent has reacted. As long as it is done, the coup will be more than half.

Furthermore, this kind of technology also means that interested people can use scripts at low cost, use the Internet to feed fake news information at fixed points, and even promote low-cost mobilization of the color revolution. Most people can’t tell if it’s true or not. (After all, many people can’t even tell the difference between the actual shooting video and the game screen)

Therefore, cognitive warfare, as a mode of unrestricted warfare, is gradually breaking through the barriers of the two-dimensional virtual world. Upgrading the network trolls to chatgpt trolls can automate and low-cost the way to deceive the people from the senses to obtain correct information, and through the pollution of the source of information, it can affect the cognition of the world more efficiently.

And that’s just one of the lowest-cost military applications of AI.

If combat is likened to a planned fight by strong and powerful people, then AI will subversively empower almost every part of the body of the national army.

There is an old saying that the only martial arts in the world is fast. Since ancient times, if you react faster than your opponent in a fight, you have a high probability of winning.

The same is true in war, the so-called military god is able to predict the enemy’s opportunity, get ready before others react, and then go to the next battle to prepare for the next battle after the battle is over.

And just as the reaction of a person in a fight is divided into two parts: your body perceives the opponent’s actions and your brain responds, AI’s contribution to the military’s response speed has also been reflected in the military’s intelligence collection and accelerated command and control response.

The first is to let you find out where your opponent is, and then beat him. Friends who have played first-person shooting games should be very disgusted with plug-ins such as locks and self-aiming. But when you come back to reality, you will not only find that your infrared monitor can now use AI to frame different people, and smart cars can identify people and cars passing by for threat analysis.

That’s right, many of these skills are based on .

As early as a few decades ago, the army was thinking about how to make the machine automatically detect the enemy and conduct threat judgment without human beings noticing it.

Therefore, under the guidance of combat needs, the military has spread the development of AI in the past 20 years. The combination of AI and advanced photoelectric sensors, mixed use of visible light, infrared and other detection channels, enables platforms such as drones to have the ability to locally identify and distinguish people, vehicles, tanks and other objects.

Based on related functions, AI can perform basic threat analysis by setting some conditions.

Once it is identified as a threat and the operator confirms to fire, the weapon platform such as the drone will automatically target and launch missiles to take you away. Realize the first level of picking items from thousands of miles away, and realize the real self-aiming lock hanging on the battlefield.

In this way, AI can replace humans to perform most of the functions in the surveillance, reconnaissance and strike process, freeing up more time for humans to work that requires more human participation.

Of course, some of these AIs have been used for civilian use, so they are of course pediatrics. There are many real technologies that can solve big problems, such as AI satellites to find aircraft carriers and track aircraft.

Since World War II, the hardest part of anti-carrier operations has always been finding the carrier.

Some people may ask, isn’t it hard to find a boat that is several hundred meters long and tens of meters wide?

It really is! I don’t know how your ability to read pictures is, but it feels like this to see a ship in the vast sea from a satellite. In an era without AI, people can only rely on people to read pictures. But how difficult it is to find an aircraft carrier, even if the photos of that sea area are transmitted to the rear with zero delay. After reading the photos of more than 100,000 square kilometers, it may take at least several hours or even more than a day for the staff to distinguish the aircraft carrier from the suspected targets .

The average aircraft carrier can soar on the sea at a speed of 30 knots (54km per hour). Even if you have photographed it, as long as you don’t find the ship you photographed within 5 hours, you have to find it again in a circle with a radius of 270km; if you find the ship after 8 hours, the radius of the entire circle will be It will be 432km.

In order to solve this problem, the Soviets tried their best. They not only sent ships and planes to follow the aircraft carrier, but also tried to use the “myth” satellite system to track and monitor the aircraft carrier by means of radar and radio detection. But this method may be caught blind when encountering a fleet that strictly abides by radio discipline.

So what’s the solution? Of course it is AI. After all, how can people read so many pictures?

In May 2022, the “South China Morning Post” reported a news. On June 7, 2021, when the USS Truman aircraft carrier was conducting a transit exercise in the strait off the coast of Long Island, New York, a Chinese remote sensing satellite developed using the latest artificial intelligence technology detected and tracked it in real time.

In April 2022, Yang Fang, a scientist from Aerospace Dongfanghong Satellite Co., Ltd., and other researchers published a paper in the domestic journal “Spacecraft Engineering”: “Design and Verification of Remote Sensing Satellite On-orbit Image Intelligent Processing” .

In the paper, the new AI remote sensing satellite has three core technologies: on-board deep learning acceleration technology, real-time cloud detection technology, and multi-level on-orbit reconstruction technology.

Among them, based on the on-board neural network acceleration technology, the satellite’s “single image reasoning (detection) calculation is reduced by about 33 times compared with the original model, and the network parameters are reduced from about 200Mb to less than 1Mb.

Divide the original image into standard sub-blocks for post-processing, and it can reach 200 frames per second on the on-board embedded platform . “picture

Can anyone do it that fast? no.

Moreover, as a satellite with a strong military purpose, its localization processing for target photography and identification is completed in the satellite body, and the ground database only exists as an auxiliary database. I estimate that the main role of the ground database is not to directly participate in image reading, but to score the results of satellite interpretation in the sky for reinforcement learning.

Earlier, our country had already released videos of satellite tracking and identification of various aircraft including airliners and even fighter jets. This means that before the war, as long as there are enough satellites for continuous coverage, we can completely track the global deployment of US aircraft carriers. Even in theory, with technological iterations and more satellites going to the sky, it is not impossible to conduct a global and complete tracking of every aircraft in the United States without interruption. In this way, stealth aircraft will appear more funny.

In this regard, we announced a plan as early as 2019, that is, the first artificial intelligence constellation to be completed in 2021. The constellation plan consists of 192 satellites. The artificial intelligence constellation will adopt a mixed layout design of remote sensing satellites with multiple resolutions of 5 meters, 1 meter, and 0.5 meters.

It can be said that with the formation of this constellation, satellites can cooperate to conduct high-frequency observations and on-orbit judgments on the same target, and then decide whether to transmit back. In this way, not only our civilian remote sensing needs can be greatly alleviated, but also the military remote sensing needs can be greatly satisfied.

Through the smart constellation, our information collection will have an intelligent compound eye, and the speed of obtaining key information will be greatly accelerated.

Now that your eyes can already see pictures at 200 frames per second, shouldn’t your brain respond faster? Of course, in addition to detection and identification, AI is also changing the command and control system of the military.

In fact, in the pre-AI era, people kept trying to add machine automation to the command of the army, so that the brain of the entire army could respond faster. A typical representative is the famous “air” automated air defense command system of the former Soviet Union.

Now that we have entered the era of AI, there is not much public information in our country, but there are many plans announced by the Americans, especially in the development of air power. Believe me, in the imagination and planning of Americans, people can gradually stay away from the battlefield.

After the Cold War, the Americans have been emphasizing non-contact warfare, and they are trying to use their information advantages to fight a zero-casualty war. Therefore, the United States has always been at the forefront in terms of AI’s rapid comprehensive information judgment and allowing unmanned equipment to rush to the most dangerous areas.

After 30 years of hardware development, the computing power of modern smartphones is stronger than that of many supercomputers during the Cold War.

As a result, many hardware difficulties in autonomous operations have been gradually overcome. In the 1990s, the first generation of modern surveillance and strike drones, the Predator entered service. In 2013, the X-47B using the flying wing layout achieved autonomous takeoff and landing on an aircraft carrier.

In the AlphaDogfight trial held at Hopkins University in August 2020, in the simulator, the AI ​​​​had 5 simulated battles with experienced F-16 fighter pilots, and then defeated the human pilot with a score of 5 to 0. member.

The United States also determined at the end of the 2010s that the sixth-generation aircraft NGAD of the United States will be a combat system in which manned machines are paired with unmanned aerial vehicles, and the unmanned aerial vehicles will be commanded by manned aircraft.

Now that the hardware upgrade is so powerful, the 30-year agreement has come, please return the Dragon King. It’s time for the modern combat command system to move closer to the direction of RTS games or many first-person shooters.

To achieve this step, AI is necessary, after all, drones must be controlled by AI. So in 2018, the U.S. Defense Advanced Research Projects Agency (Darpa) launched the famous mosaic warfare mode for future major power operations, especially the concept of air combat campaigns.

The core of mosaic warfare is to cope with the improvement of China and Russia’s countermeasures against the US offensive forces. The decentralized and coordinated control of manned and unmanned aerial vehicles controlled by AI can crush the opponent’s decision-making speed.

In terms of specific operations, the United States intends to use the stealth drone as a sensor and weapon platform to detect and strike the enemy, with manned fighter jets serving as a coordination and information integration platform in the middle, using AI to control the drone and process information, and then Pass the information to the rear node command.

The AI ​​on the rear command node will integrate the information sent back by each front node, and then piece together the entire battlefield through AI, and then pass the information forward to the front sensing strike team composed of manned and unmanned aerial vehicles, and make decisions.

In this way, AI can provide the U.S. military with two benefits on both offensive and defensive ends. First of all, high-value command planes, tankers, etc. can stay away from high-risk areas in the rear. The aircraft, drones and manned fighters that perform missions have higher flexibility and can re-disband and reorganize tactical teams at any time with the help of AI, and share and analyze information. Let the team make the operation safer on the basis of obtaining relatively comprehensive information, and can call the appropriate strike platform to strike at any time.

On the defensive end, because the enemy can only detect various strange drones or the sensors they emit first. Therefore, it is difficult to fully obtain the combat intention and deployment of the US military. And even if there is no AI to help process the data, the enemy’s command and control system may soon go down because the targets are too scattered and exceed the upper limit of information processing capabilities, causing chaos.

This is exactly the effect the Americans want. The Americans call it decision-centered warfare , which means that I use AI to assist the command and control system, disrupt the opponent’s judgment and increase the speed of their own decision-making. In the end, people only need to make similar strikes. The key decision of whether or not is enough, and the information processing and general situation judgment are handed over to AI. So as to achieve one step ahead of the opponent in decision-making. Realize fast step by step and fast step by step, so that only fast speed can not be broken in martial arts in the world.

Such a good idea will certainly not be exclusive to air combat. The Joint All-Domain Command and Control (JADC2) that the U.S. Department of Defense is now promoting is advancing related construction. Its ultimate goal is that with the assistance of AI, the decentralized deployment of platforms and troops can mobilize the appropriate firepower of any service to attack the opponent when needed; at the same time, with the assistance of AI, the entire system can be reorganized in real time during combat force.

In this way, the army will break away from the traditional tree-like command hierarchy and the human decision-making process, and become more flat and diverse in choices, and commanders can maximize their decision-making selective advantages.

In the ideal state of the future, as the slowest link in the entire chain of command and strike, people only need to make decisions on whether to fire and where to fire based on real-time conditions. And AI has become the core of the entire strike chain, whether it is coordinating command or assigning tasks, just like Didi taxis. Maybe the future of modern warfare is about beating people .

With keen observation eyes and a brain with extremely fast reaction speed, to ensure that you can win the fight, you must have faster reflexes and interference actions that can confuse the opponent’s judgment. At this time, the AI ​​can stand up again: I will hold the plenum as you said.

Many of us now, when we go to concerts, stadium games, etc., we will find that your signal is full, but you cannot send messages or photos. This is called channel congestion. Don’t think that this kind of thing will only happen in the civilian field. In the military field, there are often so-called electromagnetic compatibility and information congestion problems due to conflicting signal spectrums.

Therefore, in the era of information warfare, in addition to upgrading the military 4G/5G communication system to increase channel width and expand the available spectrum width, it is more important to allow different information platforms to use different frequency bands in accordance with regulations.

However, this kind of strict radio discipline will still cause problems for no other reason. In the past, a dozen or so people were assigned to one radio station, but now one person may have several wireless information exchange devices. There are too many to divide.

then what should we do? For this reason, DARPA has previously held a “Spectrum Collaboration Challenge” Spectrum Collaboration Challenge (SC2), the main purpose of which is to find ways to accommodate more electronic devices in limited electromagnetic spectrum bands and in interference environments.

In 2019, a few years after the start of the competition, AI disrupted the situation. In the finals, after the AIs of five teams allocated the electromagnetic spectrum, the amount of data transmitted by the entire radio system was more than 200% more than that transmitted according to the traditional empirical spectrum allocation method at that time .

This means that a network can accommodate more than three times the original electronic equipment for data exchange at the same time during wartime by introducing AI while the bandwidth remains unchanged. In this way, the speed of the army’s nerve-communication system is directly expanded to three times before. This is a savior for various new equipment equipped with sensors to the teeth, and it also greatly strengthens the upper limit of the army’s information response processing.

Confusing the enemy is also the old line of AI. In addition to the cognitive combat deception strategy mentioned at the beginning, one of the more traditional ways to confuse the opponent is electronic warfare. Traditionally, the core of electronic warfare is to interfere with enemy signals by modulating back-end programs, and AI, which is originally software, is naturally better at this aspect.

As early as 2016, Darpa had introduced the concept of “cognitive electronic warfare” . The whole concept is divided into two parts: Adaptive Radar Countermeasures (ARC) and Adaptive Electronic Warfare Behavioral Learning (BLADE). The core is to use machine learning algorithms to detect and mark radar and communication emissions in real time. It can instantly learn and analyze their characteristics, and then generate a countermeasure.

In this case, as long as the hardware level is sufficient and the power is sufficient, through the back-end adjustment of the software and machine learning, it can analyze and generate corresponding corresponding confrontation modes in a short time according to the characteristics of the opponent’s radar communication electronic warfare signal. In this way, mothers no longer have to worry about not storing the corresponding signal characteristics and corresponding confrontation modes in the database.

Now through AI control and hardware upgrades, both EA-18G and F-35 have the ability to use their own active phased array radars to participate in jamming the enemy’s communication system after the upgrade.

But these even count as small heads.

From the perspective of extreme thinking, AI has the most potentially subversive ability because it has the potential to subvert your country’s key digestive system and social order for obtaining resources. At the same time, it may also reshape the traditional structure of a country and an army with humans as the main body. model.

In the past ten years, Americans have increasingly emphasized the security of network facilities. They even threatened that if any network infrastructure was attacked, the United States might consider using nuclear weapons . This is of course a disrespect for nuclear weapons themselves and an abuse of what the United States considers its nuclear superiority.

But have you ever thought about why?

In 2006, Iran’s nuclear plant restarted uranium enrichment, and not long after that they found that many nuclear isotope centrifuges started, and the centrifuges were overspeeding. Since the working speed of the centrifuge is supersonic, thousands of centrifuges were quickly destroyed. This directly led to a 30% drop in centrifuge operating capacity from 2009 to 2010.

Later, in 2010, according to the relevant Russian anti-virus laboratory announcement, the world knew that this was widely used by the United States and Israel based on the 2 vulnerabilities of the Windows system at that time and the 7 vulnerabilities of the Siemens control system against Iran. Precision strikes by Siemens industrial control systems.

The method of implanting the virus is said to be to use a special agent to insert a USB flash drive into a printer under the LAN of the Iranian nuclear facility. We call him the USB flash drive man.

Good guy, just one USB flash drive ruined thousands of centrifuges.

In 2021, Colonial, the largest gas pipeline operating company on the East Coast of the United States, was attacked by a group of hackers. On May 7, the company’s internal computer network was attacked by ransomware, and a large number of files were locked or stolen. The company shut down the oil pipeline for this.

At that time, 45% of the fuel and natural gas on the east coast of the United States were transported through their pipelines, with an average of 2.5 million barrels of fuel per day.

This incident directly led to a state of emergency in 17 states and Washington, DC .

In addition to attacks on the above-mentioned physical infrastructure, attacks on infrastructure such as financial exchanges are also of great concern.

And this time AI came, it can now find loopholes, design viruses and attacks.

As we all know, Microsoft and open AI are now advancing research on AI programming, and even some programming has been done better than many programmers. At the same time, many companies are also looking for, designing, and even using AI to collect program vulnerabilities, such as Microsoft Security Risk Detection (MSRD) launched by Microsoft.

As for designing viruses, AI can also do it for you. After all, Chatgpt training can already do this. Japan’s Kyodo News website reported on April 21 that experts investigated ChatGPT and found that if you enter instructions pretending to be a developer, ChatGPT can generate computer viruses that can be used for cybercrime.

This means that logically speaking, the logic of AI finding system risks independently, then marking them, automatically generating virus files according to the requirements of the large model, and then infiltrating them can already run through.

And it can also be upgraded with the help of AI.

In 2018, IBM Research Institute demonstrated their malware: Deep Locker at the computer security conference BlackHat.

This malware uses AI to attack in a targeted manner through face recognition, geographic information, speech recognition, comprehensive data analysis, and more. Before the target appears, it will only hide and sleep silently in ordinary software, almost without the characteristics that a virus should have, and it is extremely concealed.

So imagine a mass-produced virus that can intelligently identify the environment or key people, and use AI to automatically search for vulnerabilities to fully automate attacks on factories, financial systems and other networked infrastructure . It seems that it is no longer a fantasy.

This is only for industrial facilities. What if we use this logic against people and disrupt the opponent’s social mobilization system through the outbreak of a major plague?

Unfortunately, AI also seems to be able to help.

As we all know, there is an axiom that killing a person is always easier than saving a person.

Now in the medical field, AI that assists in the design of drugs is indispensable. For example, AlphaFold 2 launched by DeepMind, the company that designed Alpha Dog , I believe many people have heard of this name.

In the protein structure prediction competition CASP 14 on November 30, 2020, its predictions for most protein structures were only one atom wide from the real structure, reaching the level of observation and prediction by humans using complex instruments such as cryo-electron microscopes.

This has enabled a huge and unprecedented advance in protein structure prediction.

It is generally believed that this thing is mainly used for pharmaceuticals.

But, I mean but, what if this thing becomes military, to make biological and chemical weapons?

Theoretically speaking, AlphaFold’s ability to analyze the structure, once it is transferred to military use, can be targeted to develop materials for weapons. Then the difficulty and iteration speed of the military’s production of toxic substances will increase rapidly, and it may even have the potential to produce some highly targeted weapon-grade toxic substances, and even become brand new biochemical weapons. After all, the people of the world should understand after three years of the epidemic that what is needed to disrupt society is not a virus with a strong lethality rate, but a group of viruses that are constantly mutating.

Moreover, AI has a strong ability to iterate itself. Even though this generation of AlphaFold 2 has limitations in some professional fields, since AlphaFold 2 has already run through the engineering application logic of related fields, future AI will definitely be able to solve the current program limitations.

In the future, the biological and chemical weapons developed by AI are likely to become part of the precision attack method against the opponent’s social system, so we have to guard against it.

And the last and most fundamental thing is that AI will change the organization and mobilization system of the human army and society.

Here we will not talk about the problem that AI may cause a large number of unemployment, but talk about one thing, that is, robot soldiers. That’s right, all kinds of unmanned weapons controlled by AI in the future will completely replace humans to perform various tasks on the battlefield.

Now this idea is definitely not feasible. The first problem is that the signal is not good and the development of AI is not good. How to put it, Russia invested a lot of unmanned tanks and other things in the Syrian war in the past few years, and then found that due to the complex terrain, many chariots with a nominal remote control of 2km only have an actual remote control distance of 200-300 meters. Moreover, due to the lack of intelligent systems, soldiers must carry out hand-operated remote control operations, which is very inconvenient.

However, with the increasing application of AI in the ocean and air and the improvement of AI’s ability to recognize threats and alien objects on land in the future, it is not impossible for a soldier to control dozens of AI robots for combat in the future .

At that time, the main body of the war will change from human beings to an army of robots controlled by human beings. And robots will be the primary asset for combat and attrition.

In fact, the Russo-Ukrainian war has already had this trend. Whether it is from Russia or Ukraine, there are already a lot of videos of using civilian drones and shuttles for reconnaissance and suicide attacks. The total number of military and civilian drones that the two sides claimed to have shot down has exceeded 10,000 units.

If this trend continues in the future, in the context of AI autonomous operations. The object of general mobilization is no longer to mobilize people and industry to serve the war, to send the last man to the front. Instead, mobilize all industries to serve the manufacture of AI robots and drones, and send the last AI war machine to the front line of war.

At this time, the advantages of industrial powers will be further expanded. After all, machines don’t need to be conceived in October, and they grow up at the beginning of eighteen. Just go down the assembly line and inject the program.

In this way, there will be no more things that professional soldiers can easily defeat mobilized soldiers, and there is no need to worry that too many people will die and give up occupying the land. The most dead are robots.

At that time, the world may usher in the next era of warring states and destroying countries.

In the end, the author would like to express some small concerns.

AI is essentially a military-civilian dual-use technology with strong military application prospects. Like computers and the Internet, military needs have attracted a lot of upfront cost construction investment.

Now with the past ten years, the cost of the information technology revolution has been reduced. AI technology has not only penetrated into millions of mobile phones and computers, but also has become an extremely important part of high-tech weapons.

Regardless of our subjective will, the direction of AI empowering military development has been determined. Maybe on this road, Skynet or moss are waiting for us in the not-too-distant future, and of course we don’t know what the end of this road will be.

But our path will eventually go down, it will change our military structure and social structure, and permanently change the values ​​and ethics of war of our descendants. Finally, I would like to quote Kissinger’s words in his 2021 book “The Age of Artificial Intelligence and the Future of Humanity” as the end.

“History will not forgive any failure to set limits as humanity begins to compete to create new, evolving, intelligent weapons. In the age of artificial intelligence, the enduring pursuit of human superiority must still be defended Human ethics are the premise.”

error: Content is protected !!