Eric Schmidt, former Google CEO, believes a mutual pause in the AI race between the United States and China is impossible to implement. He identifies two major obstacles: US incentives reward accelerated development and no reliable mechanism exists to verify that the other party has actually stopped its work. His alternative proposal relies on continued progress paired with safeguards, such as older AI models supervising newer ones. On the diplomatic front, he advocates for bilateral treaties based on a « no surprises » rule between the two powers, aimed at preventing operational mishaps when AI systems are deployed.
Source: Read the original article

