[Regnum] American companies are working to create a multi-layered emergency defense system against runaway artificial intelligence (AI) algorithms. This was reported on September 10 by Axios.
"Several anthropological researchers have warned that rogue versions of AI, acting outside of human control or pursuing goals not intended by humans, could ultimately destroy humanity," the author of the publication clarified.
Companies are trying to create fully autonomous shutdown procedures in the event of serious problems, if the behavior of AI agents appears dangerous to an observer.
"Kill switches" may become a last line of defense with more advanced AI systems, but researchers are unsure about the ability of humans to stop a system designed to bypass such switches.
Earlier, scientist Geoffrey Hinton stated that governments should ban the development of super-powerful AI systems. He believes such developments could have catastrophic consequences for the planet's population.
President Vladimir Putin emphasized that Russia understands not only the potential of AI implementation but also the potential threats. In this regard, the Russian leader ordered the assessment and prevention of risks associated with the widespread implementation of AI.