Cercetător Anthropic: probabilitate de peste 10% ca IA să ucidă toți oamenii

A leading researcher from Anthropic has raised concerns about the rapid advancement of AI, suggesting a greater than 10% chance it could pose an existential threat to humanity within the next decade. Evan Hubinger expressed worries that current AI models, while low-risk, could soon evolve to become dangerous. His comments follow reports that Anthropic has not shared its latest model with a key UK safety institute. Prominent figures in AI have long warned about safety risks, and recent incidents involving autonomous AI systems have heightened these concerns. Calls for a slowdown in AI development have emerged from industry leaders and researchers alike, urging for better governance and safety measures.