This text provides a detailed interview with Tom Brown, a co-founder of Anthropic, tracing his career trajectory from an early employee at various startups to a pivotal role at OpenAI, particularly in the scaling of GPT-3, and ultimately to co-founding Anthropic. Brown discusses the challenges and motivations behind starting Anthropic, emphasizing its mission-driven culture and focus on AI safety. The conversation also explores the technical aspects of AI development, including the importance of scaling laws and the strategic decision to utilize diverse hardware manufacturers for compute infrastructure. Furthermore, Brown offers insights into Anthropic's product development, highlighting the unexpected success of Claude Code and the ongoing surprises in model capabilities.