Anthropic is proposing that top AI companies coordinate a global way to pause the development of advanced AI system as a countermeasure if the AI system become too dangerous.
Nothing that AI models are douling their task capabilities every four months, raising the risk that humans could lose control over self-building AI systems.
Researchs at the Univer of Toronto say AI tools could be used to create a new kind of AI worm that adapts its hacking strategy as it spreads, suggesting that such an attack could be initiated through anything connected to the Internet.
ย
๋ค์๊ฒ์