New Manhattan Project: What Will Be the Hiroshima and Nagasaki for AI Field Tests

The Pentagon expects that advanced American technology will give the U.S. military a decisive advantage.

Author: Oleg Farichev

The U.S. Armed Forces have begun developing a military planning system using artificial intelligence (AI). At its core is the "Thunderforge" project, which essentially applies AI to processes related to data analysis and operational planning.

It is believed that with this project, U.S. military personnel will be able to process large amounts of information more quickly, which is particularly important in modern conflicts where decision-making speed plays a key role.

The project can also simulate military scenarios, effectively replacing traditional headquarters. The AI will calculate action plans and propose optimal solutions, after which commanders can make decisions based on actual circumstances.

"'Thunderforge' marks a decisive shift towards AI- and data-based combat operations, ensuring that the U.S. Armed Forces can anticipate threats and respond quickly and accurately." The Pentagon stated this after signing a contract with Scale, a company renowned for its AI research, to implement the "Thunderforge" project.

Microsoft will participate in the implementation of the project. Preliminary results are expected by the end of this year. The U.S. Armed Forces' Indo-Pacific Command and European Command will be the first to introduce this automated system.

The "Thunderforge" project is part of a larger Pentagon initiative aimed at ensuring U.S. dominance in all possible military confrontation domains, including land, air, sea, space, and cyberspace.

2018 can be considered the beginning of actively applying AI technology to achieve these goals, when the United States adopted the new National Security Strategy. This strategy defined AI as "technology ensuring that the U.S. can fight and win future wars."

To further develop this strategy, the Pentagon adopted the "AI Strategy" document and established the Joint Artificial Intelligence Center of the Department of Defense. In turn, the U.S. government established a special AI Committee and a research center for advanced experience in this field.

Since 2018, the Pentagon has implemented over 600 projects to varying degrees involving AI technology. Today, experts estimate that spending on the largest AI-related government projects in the U.S. rivals the expenditures of the Manhattan Project in the 1940s for nuclear weapons production. It is expected that investment in applying AI technology to military space systems will increase more than fourfold from 2019 to 2025, reaching $2.2 billion.

In 2023, the Pentagon adopted an updated strategy to accelerate the use of AI systems in military operations. This strategy requires all branches of the military to apply AI technology wherever it can bring the greatest military value.

Kathleen Hicks, then Deputy Secretary of Defense, highlighted the "Replicator" project she said would accelerate the transition to "small, smart, inexpensive, and numerous platforms." It is expected that thousands of different autonomous systems will be created within two years.

To introduce a multi-level "military information ecosystem," since 2021, the Pentagon has been working to improve the operational command of the U.S. Armed Forces. Specifically, by 2030, all autonomous and semi-autonomous military robotic equipment using AI technology will be integrated into a unified network.

This year, some measures will be taken to enhance the efficiency of existing military automated command systems. A series of innovative methods will be adopted at the tactical, operational, and strategic levels to process large amounts of data. This will enable even the smallest changes in enemy actions to be detected and a unified combat situation map to be formed from scattered pieces of information.

The development of natural language processing neural network technology based on large language models is seen as a breakthrough direction. In 2023, the Pentagon established a dedicated Lima task force to study potential applications of generative AI (i.e., neural networks that create new data based on trained information).

In 2023, there were reports that the U.S. military successfully tested large language models.

These systems have already been tested in the Ukrainian armed forces. This includes the situational awareness system Palantir developed by U.S. Palantir Technologies.

Another example of AI application in the military domain is the attempt to create a system to enhance the U.S.'s defensive capabilities against potential nuclear missile attacks. It is expected that by analyzing large amounts of data, including satellite images used to determine the location and status of mobile intercontinental ballistic missile launchers, the efficiency of identifying reconnaissance features for preparing a nuclear missile attack will be improved.

People believe that timely exposure of the enemy's intentions will allow the U.S. government to take certain diplomatic steps to resolve the situation.

Moreover, if the AI concludes that a nuclear attack is unavoidable, the U.S. government will make a political decision to take military measures, including preemptive strikes.

To address military issues, the Pentagon is committed to getting AI technology into practical battlefields as soon as possible.

General James Ryne, Commander of the U.S. Army Futures Command, recently stated that lessons from the Ukraine incident prompted them to do so, clearly indicating that the nature of war is changing.

Drones allow people to scout enemy positions without personally venturing out and strike discovered targets from the air without manned aircraft. Now, the army is looking for ground equipment to alleviate the burden during combat.

Currently, the Army Command is formulating a multi-stage plan with the goal of introducing new integrated human-machine formations (HMIF) by 2027.

"The platoon of human-machine formations will include ground and aerial platforms, payloads and auxiliary equipment, as well as command vehicles from which soldiers will operate." Major Jonathan Judy, deputy commander of an experimental human-machine formation, stated.

The introduction of robotic systems into the military will mean that machines, not soldiers, will be the first to engage with the enemy.

New technologies are also being introduced into the U.S. Air Force at an equally rapid pace. In 2024, it was reported that the U.S. Air Force successfully tested AI for controlling fighter jets for the first time.

An F-16 fighter jet autonomously conducted a simulated dogfight. This event took place at Edwards Air Force Base in California, where test pilots trained the neural networks using specialized simulators. There was a pilot in the cockpit of the F-16 fighter jet, but he was merely monitoring the robot-controlled process.

The success of the test gave rise to the idea of equipping up to 1000 F-16 fighters with AI technology. Americans plan to deploy the first such aircraft to operational units by 2028.

Original source: https://www.toutiao.com/article/749785887497858878030316073/

Disclaimer: The article represents the author's personal views. You can express your attitude by clicking the "Like/Dislike" buttons below.