Eric Schmidt Warns Against US Push for Superintelligent AI

Former Google CEO Eric Schmidt, along with Scale AI CEO Alexandr Wang and Center for AI Safety Director Dan Hendrycks, has strongly opposed the U.S. government's proposed 'Manhattan Project' for developing Artificial General Intelligence (AGI). They warn that this aggressive pursuit could trigger international tensions, particularly with China, potentially leading to cyberattacks and destabilizing global relations. Their policy paper, 'Superintelligence Strategy,' argues for a cautious approach, suggesting that a race for superintelligent AI could replicate the fears of nuclear escalation seen during the Cold War. They introduce the concept of 'Mutual Assured AI Malfunction,' which emphasizes the risks of preemptive actions against rival nations' AI developments. This stance comes as U.S. officials advocate for a significant investment in AI similar to the atomic bomb project, raising concerns about the implications of such a strategy. The authors call for a shift in focus to deterrents rather than competition in superintelligence development.