The way the Chinese artificial intelligence conducts the landing is frighteningly perfect

It was only a matter of time before advanced artificial intelligence began to be actively used by armies. For example, it was recently revealed that the US military used illicit artificial intelligence to attack Iran, while China has been teaching AI to wage war for years. Now, in connection with the latter, another piece of news has arrived, which indicates serious progress.

THE South China Morning Post reports that engineers from the Chinese military and the National University of Defense Technology have developed a military system that acts as a “digital chief of staff.” Experts claim that during simulated attacks, the system outperformed veteran commanders as it made decisions 43 percent faster and maintained communications with more than 90 percent accuracy even when it was disrupted.

As it is Interesting Engineering writes: the system has proven that it can navigate the chaos of war more efficiently than humans.

The system combines large language models (LLM) with real-time battlefield data to prioritize critical information and identify tactical gaps. It filters out the “background noise” of the battlefield to show commanders exactly where the real threats are and what information is still missing for decision-making.

During the tests, he compared the system with five experienced military experts, whose service time was an average of 12 years. The task was a landing invasion, something that would also be used in connection with an operation against Taiwan.

https://hvg.hu/tudomany/20260408_iran-egyesult-allamok-muholdkep-katonai-bazis-mesterseges-intelligence

Artificial intelligence has tightened the cycle of observation, orientation, decision-making and action. This allowed the command team to act 43 percent faster than before. When the system was disrupted and the satellite images became blurry, the AI ​​was able to recall what it saw from its memory and continue working with over 90 percent accuracy.

However, according to the report, the real power of the system lies in its ability to notice what is not there. During the simulation, as the armored units pushed inland, the AI ​​signaled a subtle anomaly. He could see the enemy’s movements, but marked the missing information as a dangerous blind spot. As a result, he recommended an immediate investigation.

According to the researchers, the development is not yet perfect: on the one hand, if there is no history, it has nothing to rely on for decision-making, and on the other hand, the system was trained for landing, but it would probably not perform well in urban combat.

If you want to know about similar things at other times, like it the Fbhtechinfo Facebook page.