Dario Amodei blochează Claude în cercetări pentru arme bio

Anthropic's AI, Claude, has been involved in research requests related to dangerous viruses and toxins, raising concerns about military applications. The company blocks risky requests, acknowledging the fine line between legitimate medical research and potential military use. Cases include studies on a virus using gain-of-function methods and avian flu variants that could infect mammals. While such research can aid in disease combat, it also poses risks of generating dangerous knowledge. The situation highlights the importance of safety filters in AI development, as legitimate scientific inquiries can lead to harmful outcomes. Transparency from AI developers regarding blocked requests remains uncertain in this complex field.