Head of Claude Code: What happens after coding is solved | Boris Cherny
TL;DR
Boris Cherny, Head of Claude Code at Anthropic, reveals that AI agents now write 100% of his code and predicts the software engineering profession will be completely transformed within a year, with traditional developer titles disappearing as coding becomes universally accessible through natural language.
🚀 The AI Coding Revolution 3 insights
100% AI-generated code adoption
Boris Cherny has not manually written a line of code since November 2024, instead shipping 10-30 pull requests daily using five simultaneous Claude Code agents.
Massive market penetration
Claude Code now generates 4% of all GitHub commits globally, with internal data suggesting private repository usage is even higher and growth rates accelerating monthly.
Dramatic productivity gains
Engineering productivity has increased 200% as AI handles implementation minutiae while humans focus on architectural decisions and product direction.
🛠️ Origins and Mission 3 insights
Terminal-first prototype strategy
Claude Code began as a terminal-based hack called ClaudeCLI because it was fastest to build solo, later proving terminals could adapt more quickly than IDEs to rapid model improvements.
Mission-driven return to Anthropic
Boris left for Cursor but returned within two weeks because Anthropic's safety-first mission provided essential motivation that product excitement alone could not replace.
Organic growth pattern
Initially receiving only two internal likes when announced, the tool grew through latent demand by meeting engineers in existing terminal workflows rather than forcing new environments.
🔮 The Future of Engineering 3 insights
End of the software engineer title
Boris predicts by year-end the title 'software engineer' will be replaced by 'builder' as coding becomes fully automated and everyone becomes a product manager.
Coding becomes universally solved
Within 1-2 years, manual coding skills will become irrelevant as natural language enables anyone to build software, eliminating the barrier to entry.
Evolution to autonomous coworkers
Claude Code is transitioning from code generation to autonomous agents that analyze telemetry, bug reports, and feedback to independently decide what features to ship.
💡 Innovation Methodology 2 insights
Psychological safety for experimentation
Innovation requires creating space for engineers to fail, with 80% of ideas expected to be bad, while maintaining accountability to cut losses quickly on unsuccessful experiments.
Understanding the layer below
Effective AI product development requires engineers to understand the underlying models through direct experience like post-training work to build systems that leverage model capabilities effectively.
More from Lenny's Podcast
View all
Building the most AI-pilled engineering team in the world | Fiona Fung (Anthropic)
Fiona Fung, leader of Claude Code and Co-work at Anthropic, reveals how her engineers now ship 8x more code than in 2021, fundamentally shifting the engineering bottleneck from writing to verification and requiring new AI-native management techniques to maintain quality at scale.
The hidden pattern behind successful products | Mark Pincus (FarmVille, Words with Friends, & more)
Mark Pincus shares his "Proven Better New" framework for building hit consumer products, arguing that founders should copy proven elements, make incremental improvements that 10/10 users love, and treat truly novel features as high-risk experiments—contrary to the instinct to prioritize innovation over familiarity.
Father of the iPod and iPhone on building taste, judgment, and creativity in the AI era
Tony Fadell shares lessons from building the iPod and iPhone, arguing that creating category-defining products requires resisting AI-driven cognitive surrender, embracing opinion-based decision-making for 1.0 versions, and micromanaging critical details while maintaining ruthless focus on customer pain points and storytelling.
A rational conversation on where AI is actually going | Benedict Evans
AI represents a transformative shift comparable to the internet or mobile, but we remain in its infancy (like 1997), with adoption spread unevenly across industries and the real challenge being integration and judgment rather than mere task automation.