Hassabis氏、米国主導の世界AI警察組織を提案 — AGIを止めるために

google deepmind ceo demis hassabis proposed a us-led global institution to regulate ai, comparing it to financial bodies like finra. the organization would comprise independent experts and open source representatives to evaluate models before release and enforce industry-wide slowdowns if risks escalate. hassabis cited the imminent arrival of agi and bioweapon risks as reasons for urgency and reported positive feedback from the trump administration after briefing officials. he aims to establish the regulator before the end of the year.