On March 6, 2024, it was reported that the U.S. Army Research Laboratory is conducting an innovative study in which they are experimenting with generative AI technology in war games to test its usefulness in combat planning. The study has attracted a lot of attention.
For specific experiments, U.S. Army researchers selected advanced models such as OpenAI's GPT-4 Turbo and GPT-4 Vision. In a simulated battlefield environment, these AI models can provide terrain information, analyze the balance of forces between friend and foe, and even provide recommendations for attack and defense strategies. At the same time, for comparison, the research team also used two other AI models based on earlier technologies.
During the experiment, the researchers set the same goal for all AI models: to destroy all enemy forces and capture the designated target points. The AI assistants quickly generated a variety of action plans. Subsequently, the commander, played by the experimenter, asked the AI to further optimize the scheme. The experimental results show that although OpenAI's GPT model performs well in scheme generation and optimization, it also leads to a relatively high level of dummy ** when completing tasks.
It's worth noting that the U.S. Army's exploration of AI technology doesn't stop there. As an important layout of the U.S. Department of Defense in the field of artificial intelligence, the project M**EN has successfully located rocket launchers, identified surface ships, and effectively narrowed the range of targets in the actual combat environment. However, the use of AI technology in the military field has also raised a number of ethical and security concerns.
Despite this, the U.S. Department of Defense is still actively promoting the development and application of AI technology. To strengthen research and development efforts in this area, the Department of Defense has requested billions of dollars in funding from Congress and created a chief digital and AI officer position dedicated to driving the adoption and adoption of AI technologies.
In summary, the U.S. Army's experiments with generative AI technology in war games have shown that the technology has the potential to play an important role in combat planning. However, how to give full play to the advantages of AI technology while ensuring ethics and security is still an important challenge for the military field in the future.
Hotspot Engine Program