Mallaby: US & China Must Talk AI Safety, Like Nuclear Pacts

Sebastian Mallaby, a Fellow for International Economics, believes the U.S. and China need to discuss AI safety. He recently visited China and argues that current U.S. export controls on AI semiconductors have not worked. Mallaby suggests the two nations should reach an agreement on AI safety, similar to past nuclear weapons treaties. This comes as reports indicate the U.S. and China may begin official talks on establishing AI guardrails. Mallaby's new book, "The Infinity Machine," explores artificial intelligence, inspired by DeepMind's AlphaGo and AlphaFold systems. He sees AI as a powerful tool for making sense of vast permutations. This highlights the growing urgency for international cooperation on advanced technology.