Building the most AI-pilled engineering team in the world | Fiona Fung (Anthropic)
TL;DR
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 Post-Coding Bottleneck 3 insights
Engineering throughput increased 8x since 2021
Anthropic engineers now ship eight times more code per quarter compared to previous years, rendering coding itself no longer the primary constraint in software development.
Verification replaces coding as the bottleneck
As designers, PMs, and engineers all commit code, the critical challenge has shifted to validation and quality control rather than writing speed.
Automated framework validation is essential
Teams must check in detailed specifications and use AI to validate code against these frameworks, ensuring high throughput does not compromise quality.
🤖 AI-Native Management Systems 3 insights
Remote AI sessions enable oversight at scale
Fung maintains visibility into 8x output by running a dedicated Claude Code remote session with access to all repos and Slack channels for real-time monitoring.
Morning routines now automated for feedback synthesis
Claude routines automatically synthesize feedback from internal channels, email, and social media, generating morning summaries and draft PRs for review.
Human reviewers focus on expertise, not syntax
Claude handles framework-based code review against checked-in specs, while humans focus only on deep subject matter expertise, eliminating previous review bottlenecks.
🤝 Culture and Human Connection 3 insights
High agency requires high accountability
Successful engineers demonstrate maximum initiative paired with strict accountability for clear hypotheses and outcomes, not just shipping features.
Combat AI isolation with structured human connection
Teams address loneliness from working primarily with agents through initiatives like pair-wise programming lunches to maintain essential human collaboration.
A widening gap between adopters and resisters
Leaders must address the growing divide between engineers embracing AI tools and those fighting them by encouraging teams to lean into fear rather than resist change.
Bottom Line
Treat coding as a solved problem and immediately implement AI agents for monitoring and feedback synthesis while establishing strict accountability frameworks that redirect human energy toward verification, quality control, and team connection.
More from Lenny's Podcast
View all
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.
The AI paradox: More automation, more humans, more work | Dan Shipper
Dan Shipper argues that AI will not eliminate jobs but instead bifurcate work into delegated agent tasks and deep creative work within AI-native environments, requiring more human oversight, not less.