Head of Claude Code on the future of work and productivity

Summary

CNBC's Kate Rooney interviews Boris Cherny, Head of Claude Code at Anthropic, about the company's developer conference, the evolution of AI agents, and their impact on productivity and software development. Cherny describes a transformation in coding where agents now write code, and predicts similar changes for all knowledge work within 6-12 months. He also discusses compute demands, security, and the importance of developers and enterprise focus.

• Anthropic announced new managed agent API and additional compute capacity with Colossus 1 and SpaceX partnership.
• Claude Code has seen explosive growth, outpacing all estimates, with usage extending beyond coding to data analytics and project management.
• Cherny compares the current AI shift to a 'printing press moment', predicting coding becomes a basic skill like literacy.
• Enterprise customers restructuring business processes around Claude are seeing productivity improvements of hundreds of percentage points.
• AI models have reached a phase change in security, now exceptional at finding vulnerabilities; Anthropic prioritizes giving defenders the best model first.
• Cherny advises students to learn AI tools and start startups, calling this the 'golden age' for innovation.
• The company remains focused on developers and enterprise, with no plans to broadly roll out the security tool Mythos.
• Competitors like OpenAI are also moving into this space, but Anthropic's consistent focus on coding and enterprise is highlighted.