Stuart Russell, a professor at UC Berkeley, slammed the idea that merely slowing AI development is enough to prevent existential threats. He compared the task of controlling superintelligence to a chimpanzee trying to grasp nuclear weapons—an impossible leap for a creature lacking abstract reasoning. NEWPAR Russell cited Dario Amodei’s essay "We Must Pace The Frontier" to argue that a hard barrier—akin to safety tests for aircraft—must be mandatory before any AI system is released. "If a plane can’t meet safety standards, it can’t carry passengers," he said, drawing a parallel to the need for proven safety properties in AI. NEWPAR The professor highlighted that over 1,000 researchers at leading AI firms signed a letter demanding that governments enforce these safety checks. "Their warnings come from a place of deep concern, not profit," Russell noted, pointing out that early voices like Sam Altman and Elon Musk had raised existential risks long before the current hype. NEWPAR Trump’s recent remarks dismissing AI risk as a hoax and Mark Zuckerberg’s reluctance to endorse strict oversight have intensified the debate. Russell described Zuckerberg’s stance as a "different view" that "doesn’t align with the urgency of the situation." NEWPAR With the Senate AI Committee slated to vote on a safety bill next month, Russell’s appeal is timely. The proposed legislation would require AI developers to demonstrate compliance with a set of safety benchmarks before deployment, mirroring the regulatory framework he likened to aviation safety.