This Hard Fork podcast episode features an interview with Anthropic CEO Dario Amodei about the release of Claude 3.7 Sonnet, a new AI model. The discussion covers the model's improvements in real-world tasks like coding, its "extended thinking mode," and the ongoing debate surrounding AI safety. Amodei expresses concern about the increasing capabilities of AI models and the potential for misuse, particularly in the next three to six months, emphasizing the need for proactive safety measures and international cooperation. He also discusses the rapid pace of AI innovation and the competitive landscape, particularly concerning China's advancements.
