Adnkronos (Adnkronos) –
"Artificial Intelligence could kill all humans" in the next decade. This is the alarm that Evan Hubinger, a leading researcher at Anthropic, one of the AI giants, raises in a post on X. "We truly believe that AI could kill all humans. I personally believe the probability is greater than 10% in the next decade," he writes in a post that was reposted by CBS.
Hubinger introduces himself as a researcher at the San Francisco-based company. "I think Anthropic is doing its best, but we don't yet have a plan to solve the superintelligence alignment problem, and we're clearly not on the right track to do so. I think the risk from current models is low."
What concerns me is superintelligence resulting from self-enhancement: as we said, it's happening faster than we thought,” he adds, commenting on the post of a colleague who has just resigned.
“Today I am resigning from Anthropic. I have spent the last three years conducting pre-training research at both OpenAI and Anthropic.
“Neither company is acting responsibly,” Jacob Coxon wrote in a post on X. “They are racing full speed toward self-improving superintelligence, putting our lives at risk. At OpenAI, many have not fully grasped the magnitude of what is at stake for civilization. At Anthropic, the stakes are well understood, but the company is stuck in a race to get there first: they believe that no one else will act responsibly and therefore “they shouldn’t” either, despite the risks.”
CBS called out the AI Security Institute, the international organization created by governments and tasked with analyzing, testing, and mitigating risks associated with the most advanced artificial intelligence systems. "The AI Security Institute continues to work closely with industry partners, including Anthropic, to make models more secure," a spokesperson for the British Cabinet Office told CBS, noting that "just last week" OpenAI's "most powerful" model, "GPT-6 Astra," was tested before its public release. "Risks don't stop at national borders, and no country can address them alone. The UK will continue to test the most advanced models, develop a rigorous scientific understanding of their capabilities and associated risks, and ensure policy decisions are based on concrete evidence," the spokesperson said.
international/foreign
webinfo@adnkronos.com (Web Info)