shuí谁lái来kòng zhì控制rén gōng zhì néng人工智能？
Artificial intelligence is moving from laboratories into society and the military sphere.
Unlike nuclear weapons, which require large factories and special materials, artificial intelligence mainly depends on computers, data, and software.
Once a model is made public, it can be copied quickly and can also run on computers without an internet connection. Therefore, it is difficult to truly control it simply by shutting down websites or issuing bans.
This kind of technology can help people, but it may also be used for cyberattacks, creating false information, or controlling weapons.
In war, artificial intelligence can quickly analyze situations. But if people are given too little time to make judgments, a machine's recommendations may become actual orders.
Therefore, important military decisions must be made by people. Artificial intelligence should not automatically use lethal weapons, nor should it control nuclear weapons.
Countries also need to work together to inspect the safety of military systems.
The deeper problem lies not only in technology, but also in human fear, distrust, and competition for power.
Machines have no hatred, but they may rapidly carry out human hostility.
To use artificial intelligence safely, humanity must not only make rules, but also learn to cooperate, reduce violence, and take responsibility for its own choices.