Scope
To understand the above-mentioned challenges to the military leadership in their operational environment.
Evolution & Applications of AI
AI has been known to the scientific community for decades. Its origins go back to 1956 when the word AI was first coined at the Dartmouth conference. Early AI efforts were rudimentary, using rules and symbols to represent knowledge and problem-solving. Its evolution through the decades is owed to the development of large-scale data sets, faster processing speeds, especially the GPUs (Graphic Processing Units), the introduction of neural networks and machine learning tools, and the proliferation of cloud computing and internet connectivity. All these and more have contributed to the present state of AI, where it has invaded households through Alexa and Siri. AI applications are now routine in medical diagnosis and care. Education and agriculture, too, have been overwhelmed by AI-based tools. Chat GPT has taken the business world by storm. Popular driverless cars and aircraft have stimulated people's imagination, and they have started believing they will live in the Star Wars era soon.
AI applications in the Military are already far-reaching. As mentioned earlier, the use of AI in the two recent conflicts has highlighted its utility in warfare. In the Israel-Hamas conflict, the IDF (Israel Defence Forces) relied heavily on the use of AI to determine the tunnels. However, the IDF's most significant success in AI came before their ground operations began. An AI process aided the month-long targeting of Hamas targets in the Gaza Strip. The IDF has a data target bank branch in its organisation. The task of this organisation is to store inputs on targets in the Gaza Strip received through various agencies: human and electronic intelligence (int) sources, satellite feeds, electronic surveillance and most importantly, visitors from Gaza. This bank also stores the data of previously hit targets and their status. In the days leading to the ground assault, this unit was the IDF's most secretive and active organisation. Based on live inputs coming from surveillance drones and human-int, the embedded AI systems could prepare a target list for the IAF (Israeli Air Force), which was commensurate with the war aims specified for the IDF. The AI systems' efficiency helped the IDF achieve ten times better results than in the previous conflicts. Later, The same branch could map out the fleeing targets or targets of importance for the ground forces, like the live locations of Hamas Commanders. They could also trace the ground columns to reach the objectives through routes of minimum interference.
Another dimension in which AI has widespread applications in current and future warfare is the field of drones and unmanned vehicles; in the Ukraine-Russia conflict, Ukraine has used thousands of drones against the Russian army. Many of these drones hit their targets. Significant damages to Russian logistic dumps and artillery were inflicted through the use of AI-enabled drones. Imagine a future scenario where a battlefield is saturated with AI-enabled small (low-cost) explosive-laden drones, far more than the enemy can take down. These drones, which are networked, will be capable of seeking enemy targets. Each will act as a command centre. In a superfast mode, they seek and destroy targets. Other functional drones take over tasks undone by damaged or destroyed drones. Such an attack by relatively cheap drones would wreak havoc on armoured vehicles, tanks, artillery batteries, communication centres and command centres on a battlefield. This scenario did not play out in the Ukraine war, but it could be the case in the future. In the Azerbaijan- Armenia war a few years back, Azerbaijan forces used Turkish-made suicide drones to somewhat similar effects. They could individually seek and destroy targets. The next level is what the potential holds when these drones are AI-enabled.
The two most significant changes AI would bring to the battlefield as part of the Revolution In Military Affairs (RMA) are the decision-making process (DMP) and the aerial vehicles. The decisions on a battlefield are made by commanders in their respective command centres (CCs). These CCs are connected with the seeker links and have dedicated shooter resources. Many other CCs are part of a tactical battlefield, which are networked. The DMP will be in conjunction with the networked systems. In future, they will be AI-enabled.
A sensor seeks an enemy target in the air, ground, or sea, analyses the target's nature, size and shape and allocates appropriate weapon systems for its destruction. Execute the mission and collect feedback in the post-destruction phase for further continuance or signalling the completion of the mission. However, during this process, the control remains with the commander, a human in the decision-making chain. If this chain is rid of human intervention, it will become a pure AI kill chain.
Ammunition Loading and delivery to the weapon is now within the AI-based systems' gamut. Minimum human assistance is required to monitor expenditures, raise demand, allocate ammunition from dumps based on optimum delivery requirements, transport, and even help in unloading. Such large-scale logistic operations in wartime could be executed through AI processes. In the Air Force and Navy, their decision support systems will also be slaved to the respective weapons platforms, whether an aircraft or a ship.
The future with AI will lead to a stage that is both an opportunity and a threat for military planners. Opportunity is easy to explain. However, what is the threat? The threat can be defined as the period of uncertainty in the commander's mind about the change the AI systems would bring into operations and their efficacy in providing optimum solutions to fight the enemy. The burden of transition to the AI-dominated environment would fall on the leadership, whether in the civilian world or the Military.
No comments:
Post a Comment