Chinese air combat AI can explain its actions
Artificial intelligence is becoming an integral part of modern warfare. For example, Israel’s combat AI, Fire Factory, plans airstrikes to facilitate its military operations.
However, AI is becoming more sophisticated that it not only receives orders but learns from them.
READ: Tech firms unite against deceptive AI in elections
Article continues after this advertisementSouth China Morning Post reported that China’s latest air combat AI can explain the decisions. As a result, it improves as it enters more conflicts.
How does Chinese air combat AI work?
China creating AI air combat system that can explain decisions it makeshttps://t.co/s8EEoFwi9n
— Interesting Engineering (@IntEngineering) May 12, 2024
Interesting Engineering explained how the Chinese aerial AI functions. It says smart air combat systems are becoming more common, but they learn through deep reinforcement learning.
AI company v7labs explains the method involves promoting good behaviors and penalizing bad ones. However, most AI systems work like a “black box,” meaning we cannot fully explain why they choose specific decisions.
Article continues after this advertisementConsequently, Zhang Dong, an associate professor in the School of Aeronautics at Northwestern Polytechnical University in Xian, China, developed an innovative system.
First, the combat AI builds models to explain strategies and recognize intentions. Then, it calculates how important decisions are and the likelihood of intentions, making the decision-making process easier to understand.
Lastly, it adjusts the deep reinforcement learning algorithm based on these interpretations. As a result, an experienced pilot can pinpoint the combat AI’s errors so that it avoids similar mistakes.
How does US combat AI work?
Another Inquirer Tech article reported on another aerial combat AI in the United States. Last month, it successfully completed its first dogfight against a human pilot.
The AI piloted the X-62A VISTA, which stands for “Variable Stability In-Flight Simulator.” They uploaded dogfight scenarios into the X-62A and attempted them against a human-operated F-16 aircraft.
The USAF describes the AI-vs-human encounter as follows:
“Initial flight safety was built up first using defensive maneuvers, before switching to offensive high-aspect nose-to-nose engagements where the dogfighting aircraft got as close as 2,000 feet at 1,200 miles per hour.”
Overall, it seems artificial intelligence will transform how countries defend themselves in the future. That is why it is more important than ever to learn about the latest AI trends.