Top AI Researchers Call for Regulation on AI with Self-Improvement Capabilities

According to a report by Jinse Finance on September 28, citing The Wall Street Journal, research leaders from OpenAI, Anthropic, Meta Platforms, and Microsoft are urging policymakers to review the progress their respective companies have made in automating artificial intelligence research, responding to industry-wide calls for stronger regulation of AI technologies. In a paper published on Monday, researchers including OpenAI Chief Scientist Jakub Pachocki, Anthropic Co-Founder Jack Clark, Microsoft Chief Scientific Officer Eric Horvitz, and Meta Vice President of AI Research Dawn Song wrote that the automation of AI research could trigger an "intelligence explosion." This development trend could compress progress that would typically take years into mere months or even shorter periods, potentially leading to a pace of technological evolution that surpasses human comprehension. The authors, writing in their personal capacities, emphasized that policymakers should urgently mandate increased transparency to understand the extent to which companies have advanced in automating model development.