With the development of unmanned aerial vehicle technology and artificial intelligence technology, the form of air combat is also moving towards a deeper level of informatization and intelligence. Understanding the new connotation, characteristics, and trends of human-machine unmanned aerial vehicle collaborative combat is of great significance for mastering the mechanism of air combat victory and maintaining national airspace security. The evolution of aerial combat from "knight duel" to "system confrontation" over the past century is a history of constantly expanding platform capabilities and combat radius, while command and control are becoming increasingly centralized and system dependence is deepening. The aerial combat in World War I was a visual combat and close combat between pilots using a single platform with machine guns and cannons as the main body, within a range of several hundred meters. The victory or defeat depended on individual skills and courage. During World War II, although radar began to be used, the vast majority of aerial combat was still "knight duels" within line of sight. The jet age and the emergence of missiles have made beyond visual range air combat possible. But discovering long-range targets and guiding attacks far exceeds the capability of a single fighter jet. So, the early warning aircraft took to the historical stage. It lifts the radar into the sky, becoming the "eyes" and "brain" of the entire air formation, realizing the perception of the vast air situation and the command and guidance of multiple batches of fighter jets. In the "5-7 Air Battle" between India and Pakistan, the J-10CE aircraft achieved beyond visual range air strikes from a distance of 160 kilometers and shot down advanced fighter jets such as the "Rafale". At this point, the key to victory has shifted from the quality of individual pilots to the collaborative ability of the system under the command of early warning aircraft. The relationship between early warning aircraft and fighter jets formed the classic paradigm of "manned manned" collaboration in the early days, laying the foundation for systematic combat. In recent years, with the emergence of stealth fighter jets, traditional radar networks have been penetrated, and air combat has entered the "stealth era". The survival of early warning aircraft is facing challenges, and distributed and networked operations have become inevitable. Data link technology integrates all nodes such as fighter jets, early warning aircraft, ground air defense systems, etc., forming a "network centric warfare" capability. Each platform is both a receiver and a contributor of information. However, even in the network, reconnaissance, suppression, and consumption of high-risk areas at the forefront still require manned fighter jets to complete, with high risks and costs. In November 2025, a manned fighter jet of India's "Radiance" crashed at the Dubai Airshow, exposing issues such as the pilot's physiological limits and high-intensity deployment pressure, which have always been a bottleneck restricting the sustained generation of combat effectiveness. The development of air combat history clearly shows that air combat is bidding farewell to "individual heroism" and always pursuing two major goals: one is to extend the arm of perception and strike, and the other is to protect key high-value air decision-making nodes. In the traditional mode, there is a contradiction between the two. The maturity of drones has propelled air combat towards a new era of human-machine collaboration, becoming the driving force behind the shift from platform centric warfare to system centric warfare. Under the "complementary advantages" and "multiple modes" of collaborative combat, the concept of human-machine collaborative combat is constantly maturing, key technologies are gradually breaking through, and some prototypes have been verified. New combat capabilities are being transformed into real combat capabilities. Collaborative combat between manned and unmanned aerial vehicles (UAVs) refers to the implementation of coordinated direct combat actions through the use of UAV and UAV formations, information sharing, situational integration, task planning, collaborative control, human-computer interaction, and resource coordination, supported by a system, in order to jointly achieve combat objectives. The core concept is to integrate unmanned aerial vehicles (as command, decision-making, and high-end mission platforms) with various types, functions, and purposes of unmanned aerial vehicles (as cutting-edge nodes for perception, strike, decoy, electronic warfare, etc.) through a networked information system to form a distributed air combat system. According to relevant experimental analysis, the coordinated operation of manned and unmanned aerial vehicles can shorten reaction time by 50%, improve survivability by 25%, and increase mission success rate by 35%. Its winning logic is to "obtain information advantages through distributed nodes and transfer operational risks to low-cost unmanned platforms". At the 2024 Zhuhai Air Show, the J-20S dual seat aircraft model made its first public appearance, pioneering the world's dual seat fifth generation aircraft. The dual seat design upgraded the fighter from an excellent "combat platform" to a powerful "air combat command center". The front seat pilot focused on flight safety and tactical maneuvering, while the rear seat mission system officer was responsible for commanding and controlling the drone formation, significantly improving the efficiency of executing electronic warfare, joint strikes and other tasks in complex electromagnetic environments. This "flight brain+tactical brain" model will strongly support the reconstruction of the human-machine collaborative air combat mode. On September 3, 2025, at the military parade commemorating the 80th anniversary of the victory of the Chinese People's War of Resistance Against Japanese Aggression and the World Anti Fascist War, a new type of unmanned aerial vehicle, unmanned wingman, unmanned air superiority combat aircraft, and other unmanned aerial combat units were formed. It is a new type of unmanned aerial equipment independently developed by China, with land-based and shipborne combat capabilities. It mainly performs diversified combat tasks such as battlefield reconnaissance, surveillance, and strike, and can launch covert attacks, wide area coverage, and autonomous coordination. It will continuously innovate new styles of human-machine coordinated combat. Of course, the coordinated operation of UAV and UAV is not a single mode, but a dynamic combination of multiple modes according to mission requirements. One is the master-slave collaborative style ("loyal wingman"): this is currently the most mature model. Controlled by a manned fighter jet, 2 to 4 unmanned aerial vehicles with high subsonic speeds and carrying multiple mission payloads are used for combat. Drones are the 'brain', directly commanding multiple highly autonomous drones, responsible for advanced decision-making and final attack authorization; Drones are the 'arms' that undertake high-risk missions. The second is the cluster collaboration style ("swarm tactic"): a low-cost, miniaturized intelligent swarm is formed by releasing and guiding dozens or even hundreds of drones or large platforms. The bee colony achieves distributed perception, saturation attacks, or collaborative electronic suppression through autonomous collaborative algorithms. Using large early warning aircraft, bombers, and even transport planes as command hubs, the swarm can autonomously plan attack paths and task assignments by issuing macro commands such as "suppressing a certain area's air defense system". The third is the heterogeneous collaborative style ("functional puzzle"): combining drones with different functions with drones to form a "team", based on dynamic division of labor, equally assuming various combat responsibilities, and cooperating with each other to complete air combat tasks. Looking ahead to the future, with further breakthroughs in artificial intelligence, decentralized networks, and cross domain fusion technologies, the coordinated operation of manned and unmanned aerial vehicles will surpass the simple master-slave "long aircraft wingman" relationship and evolve into a highly intelligent and autonomous collaborative air combat "ecosystem". The key technologies are developing towards the direction of "more intelligent and autonomous". Through machine learning, AI can accumulate far more "aerial combat experience" than human pilots in simulated environments, become a new quality "ace pilot", and provide microsecond level tactical advice in real confrontations, even authorized to autonomously make offensive and defensive decisions under specific rules. At the same time, human-machine mixing will enter the physiological level, and brain computer interface technology will support pilots to transmit intentions to drone swarms more quickly through "thoughts", or directly receive "sensory information" from drone sensors, achieving true human-machine integration and thought control. This will greatly shorten the OODA (Observation Judgment Decision Action) cycle, achieving the state of "mind following thoughts, sword at will". The scale of collaboration is exploding towards the level of a "manned unmanned swarm". In the future, the main body of collaboration with drones will develop towards an "adaptive swarm", where thousands of drones will act like organic life forms, able to adapt to the battlefield environment. Even if some individuals are lost, the entire swarm can still reorganize and continue to perform tasks. Drone pilots become 'bee shepherds'. For example, there are accusations from behind fighter jets, loyal wingmen in the center, and unmanned swarms attacking and suppressing. The early warning aircraft simultaneously commands high-altitude long endurance unmanned aerial vehicles for persistent monitoring, medium electronic warfare unmanned aerial vehicles for network electronic countermeasures, and small unmanned aerial vehicle fleets for saturation attacks, forming a layered and three-dimensional combat system. The winning logic is leaping towards the dimensions of "decision-making speed and system intelligence". The future mechanism of human-machine collaboration for victory will leap to the level of "emergence of advantages through system intelligence, achieving a complete collapse of the opponent's decision-making cycle". A large number of unmanned intelligent platforms with diverse functions adaptively combine and dynamically allocate tasks on the battlefield, generating "emergent intelligence" that no single platform or traditional formation can match. It cannot defeat its key nodes through "decapitation", drag down through attrition warfare, or reach the decision-making speed of "intelligent life forms". The coordinated operation of manned and unmanned aerial vehicles is an inevitable stage in achieving intelligent air combat, which is reshaping the face of air combat with unprecedented strength, and is a deep integration of human intelligence and machine intelligence. In the future, when fighter jets are no longer lone "swordsmen", but accompanied by a group of "death squads", "thousand mile eyes", and "wind following ears" played by drones; When early warning aircraft no longer only focus on the trajectory of enemy fighter jets, but instead dispatch and manage numerous aerial robots, it marks a fundamental change in the form of aerial combat. From pursuing platform excellence to pursuing system intelligence; From relying on human courage and reaction, to integrating human intelligence with machine precision and endurance, the synergy between manned and unmanned aerial vehicles is the new driving force for this evolutionary development path. Like an experienced music conductor, it leads an orchestra with tacit division of labor and intelligent autonomy, playing a grand and precise symphony in the sky together. This transformation is not only a leap in military technology, but also a profound shift in the philosophy of war from "destroying platforms" to "paralyzing systems", and from "pursuing firepower density" to "pursuing decision-making speed". Whoever can take the lead in building and managing this intelligent, distributed, and resilient aerial combat ecosystem will be able to compose the magnificent symphony of human-machine collaboration earlier and better, and who will hold the key to unlocking the door to future aerial combat victories. (New Society)