ChatGPT’s Enlightenment to Future Battlefield Intelligence Perception and Decision-making

Author: Liu Xinyu

The combination of artificial intelligence technology and unmanned aerial vehicle will greatly change the existing theory and mode of military operations. The power of artificial intelligence plays a role at all levels of the national defense industry, and the rapid development of artificial intelligence will define the next generation of war. Recently, ChatGPT, a generative artificial intelligence technology developed by OpenAI, has been paid attention to and used all over the world. ChatGPT is a large-scale generative pre-training language model based on transformer. It can train in existing data sets and generate texts similar to human language. This unique ability makes it an ideal tool for military applications. The underlying natural language model and technology that provides power for ChatGPT may completely change the artificial intelligence on the battlefield and have a great impact on the situation awareness and independent decision-making methods in future wars.

Intelligent situational awareness in the whole information domain

In the future military field, there will be a battlefield environment with complex information, high confrontation and changeable tasks. The highly uncertain combat environment puts forward extremely high requirements for the independent perception and cognitive ability of combat equipment. Military equipment needs to have the ability of automatic target detection and identification and multi-sensor data fusion. It can detect and fuse enemy target information and its own support information through autonomous and receiving information acquisition methods, and perceive the battlefield situation and extract important information for subsequent decision-making on the basis of obtaining full information domain.

Based on the demand for intelligent situational awareness, the generative artificial intelligence technology ChatGPT can be integrated into military vehicles, aircraft and other combat systems. With the application of artificial intelligence language robots trained by a large number of models, the required real-time information can be effectively coordinated in a multi-domain environment, and the input data from various sensors can be analyzed to generate a complete, comprehensive and real-time updated operational environment map.

Intelligent technology can play a key role in future military operations. Generative artificial intelligence technology can rely on its strong creativity, understanding and response speed to obtain cross-domain intelligence and battlefield situation data, improve the ability of insight into intelligence, form a highly simulated situation scene, realize dialogue between people and battlefield environment, provide real-time information and situation prediction, enhance the real-time and decision-making of battlefield situation perception, and better support real-time decision-making in military operations.

Real-time and efficient intelligent independent decision-making ability

At present, the autonomous decision-making ability of aircraft has been initially characterized by intelligence and independence. The US Air Force uses artificial intelligence "decision-making assistant tools" in distributed common ground system (DCGS) to help sort out and integrate a large amount of data. This artificial intelligence system connects most airborne intelligence and monitoring and reconnaissance platforms of the US Air Force, and integrates artificial intelligence technology into training to expand into other fields.

If the generative artificial intelligence technology ChatGPT is introduced into the decision-making method of aircraft, it can provide real-time information about enemy positions, movements and capabilities, as well as the advantages and disadvantages of friendly forces in tactical situations based on the prior information database and real-time signal, data and image databases, and at the same time based on the real-time interaction between aircraft and environment, so as to analyze, reason and make decisions, and realize rapid response to battlefield decision-making.

The generative artificial intelligence robot can generate multiple sets of operational plans in a short time when the aircraft is faced with complex and uncertain operational conditions, and preview the battlefield process and results for each set of plans, so as to generate real-time optimal decisions in the face of complex requirements in terms of information acquisition, reaction time, calculation speed, tactical evolution and comprehensive evaluation, and support decision makers with diversified decision plans and deduction results.

Real-time and efficient intelligent autonomous decision-making method can cover the complex situation in real combat environment, play a similar role to human brain in high uncertainty environment, dynamically adjust attack and protection strategies according to real-time situation awareness and operational effectiveness evaluation, and realize efficient confrontation through closed loop of process.

Enlightenment and prospect of intelligent military

Generative artificial intelligence technology breaks the logic of time series calculation, and makes artificial intelligence in multiple sub-fields begin to merge technically. As a new technology in the field of artificial intelligence, deploying ChatGPT in military operations may enhance cross-domain combat capability and realize situational awareness and real-time independent decision-making in all information domains. In the future battlefield, if we can deploy the top-level layout and bottom-level algorithm of generative artificial intelligence with the guidance of military operations and equipment development, with its ability of understanding, responding and interacting with people, we can greatly improve the cognitive and decision-making methods in the battlefield, promote the technical upgrading in key areas, and realize the optimization iteration of combat capability.

From the point of view of data and subsequent development, generative artificial intelligence technology is a more advanced neural network deep learning algorithm, which has high requirements for training data, depends on the authenticity of training data and is easily disturbed by external information. Because of the long training time and billions of parameters, automatic machine learning is needed for multiple lines to generate better calculation results. Therefore, when transforming the technological achievements of generative artificial intelligence, it is necessary to take into account the parameterization requirements of scientific research technology development and the automation requirements of equipment application development, so as to achieve the balance between scientific research liquidity and industrial productization.

Generative artificial intelligence has brought a new paradigm to military applications and set a new route for the next generation of military operations. It is the main problem to apply generative artificial intelligence technology to the military field to actively explore the representation form of situation awareness and decision-making tasks for different military problems and consider how to use effective information for large-scale pre-training. The combination of artificial intelligence technology and military operations and technological innovation will reserve new ways for situation awareness and independent decision-making in future operations, and realize intelligent support for the development of new military equipment.