Perspective on the “Gateway to Victory” of Cognitive Operations

  Cognitive warfare, after firepower, mobility, and information power, is another top of the battlefield to win. After thousands of years of hidden confrontation with the human brain, it has shown a new look on the modern battlefield.
The basis of cognitive success: Computing power is everywhere

  ”If the husband has not fought and the temple counts as the winner, it must be counted as more.” Although the “temple calculation” is not so precise, it is also an indispensable link in knowing the outcome of a war. If it is said that ancient wars were won by “temple computing”, then modern wars are won by the ubiquity of computing power.
  Computing power is ubiquitous, which is an upgrade of the cloud brain computing model. The typical cognition process in the current combat is that people are on the circuit, in the circuit or outside the circuit, and the data information is gathered in the “cloud brain”. The “cloud brain” uses powerful computing power to extract cognition, and then Return to the relevant combat platform. This “cloud-brain sending computing” model faces two major problems: one is the delay in computing power transmission, and the communication network often faces the dilemma of not being able to deliver computing power to the end in time. Although 5G provides a new solution, the base station The limited coverage is not conducive to its military application; the second is the cognitive delay of the cloud brain, and the probability of modern warfare is extremely high, followed by an instantaneous burst of computing power demand. When the calculation amount exceeds the maximum load of the cloud brain, the cognitive results will be timely. Presentation has become a luxury. A common problem faced by many Internet companies today is how high their “cloud center” computing power should be to cope with less than 1% of the computing peak time. In recent years, militaries around the world have been seeking effective solutions for the supply of computing power. Cloud-network integration, one-hop into the cloud, and cloud follow-up, etc., have quickly become new ways of computing power delivery on the battlefield, and are also upgrading every combat unit. into an organism capable of autonomous cognition.
  Computing power is ubiquitous, which is the impetus for the rise of the edge of combat.During the long evolution of combat methods, because the combat edge is difficult to control, the cognition on the edge side is often ignored or ignored, and the edge unit is therefore reduced to the executor of cognition. However, the experience of letting people who can hear the artillery fire to command the battle has repeatedly warned us that marginal cognition is often the accurate interpretation of the battlefield situation, which is also the key to the efficient operation of the hierarchical organization of the military. In recent years, distributed combat forces, decentralized mesh architecture, and highly intelligent autonomous platforms have increased the demand for edge participation in combat collaboration unprecedentedly. The edge must complete autonomous cognition on the battlefield of high-intensity confrontation in order to truly remove “Iron Foolish” hat. Scenes where fifth-generation fighter jets like the F-22 act as combat clouds can be seen everywhere. Portable tactical cloud equipment has become a companion of combatants. New types of optical fibers that can be sewn into military uniforms and can sense surrounding environmental information and perform simple calculations and analyzes are being developed. In research and development, some military experts predict that 90% of computing tasks on the future battlefield will be completed on the edge side. Computing power is ubiquitous with the rise of the edge side. Edge computing is becoming an important mode of computing power supply on the battlefield.

Edge computing refers to an open platform that integrates network, computing, storage, and application core capabilities on the side close to the source of objects or data, and provides the nearest end services

  Computing power is ubiquitous, which is the necessity of data-driven cognition. The data on the battlefield is growing explosively, and it is obviously impossible to deal with such a level of computing tasks only by the human brain. In fact, most of the huge amounts of data generated instantly on the modern battlefield are not for the combatants, but for thousands of embedded chips, smart terminals and portable computers. Only machines can deal with the data on the modern battlefield. Massive data, and only when a “data consensus” is reached between machines, that is, machine cognition, can the data be truly transformed into information that combatants can understand, so that it is possible to extract cognition from the information; otherwise, even Data that machines cannot read, or data that cannot be transferred between machines, will soon be annihilated due to the loss of the carrier on the battlefield. The pressure of processing massive amounts of data on the battlefield and the fact that machines can only recognize data force computing power to be embedded in weapon platforms, and are distributed in every corner of the battlefield along with combat terminals. The U.S. Department of Defense’s investment in microelectronics in fiscal year 2022 will reach 2.3 billion U.S. dollars. This is the U.S. military’s continuous investment in the cloud computing strategy, and it also shows that the U.S. military regards ubiquitous computing power as the focus of combat power building.
The starting point of cognitive victory: the emergence of swarm intelligence

  Cognition is formed in the brain. The brain of modern warfare is composed of human brain, computer brain and cloud brain, each of which has its own advantages and disadvantages. Only by close cooperation, communication and mutual recognition, and co-excitation of stress can the cognitive emergence effect be produced.
  The emergence of group intelligence is the basic way of cognition in modern warfare. The human brain is obviously unable to cope with so much data in modern warfare. High-level commanders can only read about 1,000 words per minute, and the size of a thousand-word electronic document is only a few hundred KB, but commanders are needed every minute on the battlefield. The data traffic displayed is more than hundreds or thousands of MB. The advantage of the computer brain is that it can downlink data that humans cannot understand, but it is too rigid, once the algorithm is implanted, it is difficult to change, and self-learning is just an algorithm optimization driven by similar data. When you have seen the situation, it seems powerless. The cloud brain stores a large amount of data, models and algorithms that cannot be memorized by the human brain, implanted by the computer brain, or stored by the network. Its goal is to be omnipotent, all-encompassing, and omnipotent, but it cannot be exhausted. All combat scenarios that didn’t happen. Although the human brain has limited computing power, its logical reasoning ability is super strong. Although the computer brain is a little rigid, it will not make mistakes when the data is given correctly. Although the cloud brain is far from the frontier of combat, it is all-encompassing. As long as the “human-machine-cloud” cooperates closely, no matter what What kind of scene appears on the battlefield can be made in a short period of time, and under the condition of incomplete information, a combat plan that may not be optimal, but can be regarded as an impromptu decision.

Microelectronics technology is a discipline that acts on microminiature integrated circuit systems on semiconductors

  The emergence of swarm intelligence is a subversion of the single-cloud cognitive architecture. At one time, whether it was the joint information environment of the US military or the unified information space of the Russian military, they all attempted to build a cognitive cloud environment with a single technical architecture. However, just as ships cannot be required to change their speed from “knots” to “Mach”, this single cloud cognitive architecture solves the problem of “unification”, but what it brings is that many weapons and equipment are excluded from the combat system. They even shackled their own side in the “information cocoon room” they constructed, and lost their accurate understanding of the entire battlefield. In December 2021, the U.S. Department of Defense overturned the original $10 billion cloud project contract with the new project “Joint Combatant Cloud Capabilities”. Products from different service providers are incorporated into the U.S. military’s combat system. To promote the emergence of swarm intelligence, it is necessary to change the “single cloud” into a “polymorphic cloud” cognitive architecture, integrating the data of various combat platforms such as ships, aircraft, bombs, and guns, and even non-military information equipment; otherwise, future battlefield confrontation A large number of originally very important data will be missed in the process, and the insufficiency of the data will bring the direct consequence of inaccurate cognition, which will be catastrophic for cognitive victory.

  The emergence of swarm intelligence is the result of mutual trust between human-machine and cloud cognition. Combatants’ recognition of autonomous systems, humanoid robots, and artificial intelligence equipment is continuously increasing, and people have become accustomed to giving instructions to machines to give them a planned marching route, or a weapon’s combat technical performance index parameters; Get used to sending a request to the cloud center for a scenario solution, or remote support for an algorithm model and software. All of this stems from the advancement of technology, as well as the trust in machines established by combatants after many personal experiences. The purpose of the U.S. Air Force’s Alpha air combat test is to allow human pilots to gradually feel the decisions made by the “loyal wingman” in the experience of confronting AI fighters, and then communicate with it cognitively on an equal footing. With the upgrading of technology, the phenomenon that combatants on the battlefield unidirectionally input instructions to machines or cloud brains instead of treating them as cognitive subjects will change rapidly or even reverse, and the “human-machine-cloud” The issue of cognitive mutual trust will no longer be a fetter, and the emergence of swarm intelligence will no longer be limited to human-to-human, machine-to-machine, and cloud-to-cloud, but will span the multi-dimensional world of human society, the machine world, and the information space. domain collision.

In December 2022, the U.S. Department of Defense decided to award contracts for joint operational cloud capabilities to Amazon, Google, Microsoft, and Oracle
Achievement of Cognitive Victory: Human-Machine Collaborative Game

  Although data, algorithms, and models are popular in modern warfare, the traditional combat philosophy of surprise and surprise cannot be discarded. The game of commanders on the traditional battlefield has become a game of human-machine cooperation on the modern battlefield.
  The game of human-computer cooperation is the advancement of cognitive barriers. In essence, the OODA cycle is a process of transforming cognition into action. Observation, judgment, and decision-making are all behaviors driven by cognition. The cognitive game on the traditional battlefield can only be known whether it is correct after fighting on the battlefield with real knives and real guns, which also makes every link in the operation of the OODA ring very important. Since the 1990s, previous wars have repeatedly promoted the rapid shortening of the OODA cycle, from dozens of days, to days, to hours, and now to the minute level, while the new intelligent weapon platform requires completion in seconds, otherwise it will The “window” for launching an attack will be missed, which essentially puts forward new requirements for the cognitive game on the battlefield, that is, the gate must be moved forward. On the modern battlefield, due to the ubiquitous computing power and the strong support of the cloud brain, commanders, combatants, autonomous platforms, and smart terminals can complete scenario assumptions, plan generation, and plan deduction in a very short period of time. Etc. For various situations, trends and possibilities in combat, speculate and predict various combat results. This adds a new prediction link to the OODA loop, making the cognitive game move forward from the real battlefield to before the battle. .
  Human-computer cooperation game is a penetration of complexity.War complexity is a new interdisciplinary subject that has grown very fast in recent years. Nowadays, when various digital models symbolize entities such as artillery, tanks, fighter planes, and ships in the combat system, coupled with data, connections, and empowerment, which are inherently virtual factors, people find that the battlefield The level of complexity has increased by several orders of magnitude compared to the past. The increase in the complexity of the combat system has actually given the combatants the opportunity to seize the opportunity in the game process. Whoever can penetrate the complexity of the combat system and quickly and accurately examine and recognize the situation and direction of the battlefield will be able to be ahead of the opponent. Find the key to victory on the battlefield in one step. To penetrate the complexity, to disassemble the combat system, the most effective way to disassemble is to use the means of connection when constructing, that is, first decompose the tasks, functions, and units in the combat system, and then connect them with the network , and use the flow direction, flow rate and flow rate of information to achieve the classification of combat units, the unification of group cognition and the association of combat operations. Such a combat system has evolved into an organism of “two-way mutual control” between the digital virtual space and the physical world entity, thus providing a new “entry” for penetrating the complexity of the combat system.

Computing power and cloud brain together accelerate the cycle of OODA, and make the cognitive game mentioned before the war

  The human-machine cooperative game is a transformation of the trial and error method. Scientific experiment is falsification, and combat game is trial and error, and the cognition after trial and error is the best. The traditional combat game is mainly completed in the commander’s brain, and trial and error is also flashed in the human brain. Combat games in modern warfare add machine calculations to the commander’s calculations, giving the commander greater opportunities for trial and error. Although this trial and error is virtual or simulated, it is also enough to make the cognitive on the battlefield The confrontation has been raised to a new level. Due to the abundance of battlefield data, in order to seize the opportunity during the trial-and-error game, whether it is using the “commander prediction-machine deduction” method or the “machine prediction-commander judgment” method, artificial intelligence technology must be relied on. auxiliary. Otherwise, the complexity of the combat system will make the cognitive decision-making under any rules fall into the dilemma of being unable to give results. Since 2019, the U.S. military has invested a lot of money in an attempt to build a cross-domain “kill net” similar to the “supplier-software agent-user” architecture, and use artificial intelligence technology to assist in the construction of an instant kill chain. Although this attempt has encountered many problems such as resource conflicts, response delays, and capability mismatches in practice, it is still a new way for machine intelligence to cooperate with human commanders to achieve game victory.