Russian media attention: China has learned to distinguish between real and fake nuclear warheads!
On May 31, multiple media outlets such as the TASS reported.
According to the South China Morning Post, Chinese scientists have invented an artificial intelligence system that can distinguish between real and fake nuclear warheads.
This is the world's first solution based on artificial intelligence for arms control verification.
Russian media pointed out: Scientists used the Monte Carlo method for modeling, creating "millions of virtual nuclear components," some containing weapons-grade uranium, while others were covered with lead or low-enriched uranium.
"A multi-layer deep learning network was trained on a neutron flux model, achieving extremely high accuracy in distinguishing real warheads."
Russian media explained that in order to prevent artificial intelligence from directly accessing top-secret information about nuclear weapon designs, Chinese scientists built a polyethylene wall with 400 holes between the verification system and the actual warhead, encrypting neutron signals, masking the geometry of the warhead, while ignoring radiation characteristics.
Therefore, this system can verify the chain reaction capability of the core of a nuclear weapon without revealing design details.
Russian media cited Chinese media reports stating that the artificial intelligence system knows nothing about the design of the warhead but can still determine its authenticity through partially hidden radiation signals.
Scientists at the China Institute of Atomic Energy stated that in order to ensure transparency in verification, the artificial intelligence system can be jointly coded, trained, and verified by both the verifying and verified parties.
Original article: https://www.toutiao.com/article/1833591763365131/
Disclaimer: This article only represents the author's personal views.