rén gōng zhì néng人工智能jìng sài竞赛xū yào需要ān quán安全guī zé规则
Artificial intelligence is developing quickly. Some researchers involved in its development worry that future systems may become extremely powerful, while humans still do not have reliable ways to ensure that they always follow human instructions.
This does not mean that a disaster will definitely happen, because the idea that artificial intelligence could destroy humanity is still a controversial extreme hypothesis.
However, if experts believe that the danger really exists, companies cannot focus only on improving technical capabilities and leave safety issues to be solved later.
Researchers, company managers, and governments should all take responsibility.
Today, a small number of large companies control funding, equipment, and talent, yet they may make the whole world share the risks.
Companies and countries also fear falling behind their competitors, so no one wants to slow down first. This is much like an arms race.
The real question is not whether to research artificial intelligence, but who decides the pace of development, which risks are unacceptable, and when progress must be paused.
Artificial intelligence affects everyone, so safety rules cannot be decided by companies alone.
Countries need to hold open discussions, conduct joint reviews, and establish clear limits, so that reliable “brakes” are ready before technology continues to advance.