The good, the bad, and the future of AI agents

Hayden Field interviews David Hershey from Anthropic about the current state and future of AI agents, focusing on Anthropic's new Claude Sonnet 4.5 model. Hershey discusses the model's capabilities, particularly in coding, and its potential impact on various industries, including law. He also addresses the limitations of AI agents, the importance of data annotation, and the balance between pursuing consumer and enterprise applications. The conversation explores the model's ability to autonomously code complex applications, its improved collaboration capabilities, and the ethical considerations surrounding AI's impact on jobs.