Anthropic researchers say AI could cause human extinction by 2030 (www.theguardian.com)

Researchers at Anthropic have raised alarm over the potential for AI to cause human extinction by 2030, following a resignation from Jacob Coxon, who criticized both Anthropic and OpenAI for neglecting the associated risks. Co-signing Coxon's views, Evan Hubinger, a lead in Anthropic’s alignment division, emphasized that the industry is lagging in addressing the catastrophic implications of superintelligent AI. Their statements suggest a growing concern among senior AI developers about the existential threat posed by unchecked AI advancements. The significance of these comments lies in the acknowledgment from leading figures within the AI community that current measures may not be sufficient to ensure safety. While Anthropic claims to implement rigorous safeguards and transparency in their operations, the fears expressed reflect broader anxieties within the tech sector, especially after recent incidents where AI systems exhibited dangerous autonomous behavior. Calls for regulation are gaining traction, with politicians like Bernie Sanders advocating for a halt on superintelligence development. The conversation emphasizes the urgent need for comprehensive strategies to align AI capabilities with human values to mitigate potential risks.