# Boris Cherny: How We Built Claude Code

Source: https://www.youtube.com/watch?v=PQU9o_5rHC4
Recap page: https://rapidrecap.app/video/PQU9o_5rHC4
Generated: 2026-02-17T15:32:42.703+00:00

---
## Quick Overview

Boris Cherny, Creator & Head of Claude Code at Anthropic, advises founders building on LLMs not to build for the model of today but for the model six months from now, emphasizing that Anthropic's code, like Claude Code, has been constantly rewritten and that relying solely on external tools or highly specialized expertise can be limiting, favoring instead a generalist approach and continuous learning.

**Key Points:**
- Cherny advises founders building on LLMs to build for the model six months in the future, not the current one, as capabilities change rapidly.
- Anthropic's code base, including Claude Code, has been continuously rewritten and iterated upon over the last six months.
- He notes that many early AI applications relied heavily on the terminal (CLI) because building a graphical UI was too time-consuming.
- The development of QuantMD, an internal tool, was an accident driven by the need to debug memory leaks, leading to a useful product.
- He argues against over-specialization in AI development, stating that generalists who understand core principles and debugging (like reading logs) are often more effective than hyper-specialists in areas like formal verification.
- Anthropic's core principle is teaching the model how to use tools rather than hardcoding specific features, which allows the model to adapt better to future capabilities.

![Screenshot at 00:05: Boris Cherny discusses Anthropic's advice to founders to build for the future model, not the current one, as exemplified by the constant iteration of their internal code base.](https://ss.rapidrecap.app/screens/PQU9o_5rHC4/00-00-05.jpg)

**Context:** Boris Cherny, Creator & Head of Claude Code at Anthropic, participates in a panel discussion with other Y Combinator partners (Garry Tan, Diana Hu, Jared Friedman) about building and scaling AI products, specifically focusing on the evolution of their internal coding tools and the philosophy behind developing for rapidly advancing LLM capabilities.

## Detailed Analysis

Boris Cherny, Creator & Head of Claude Code at Anthropic, strongly advises founders building applications on large language models (LLMs) to design for the model six months out, not the current one, due to the rapid pace of AI evolution. He points out that Anthropic's internal code base, including Claude Code, has been constantly rewritten and refined over the past six months, illustrating the need for adaptability. Cherny recounts the accidental creation of QuantMD, an internal debugging tool born from the necessity to fix memory leaks, which proved highly useful for developers. He stresses that relying too heavily on hyper-specialized knowledge or overly complex tooling like formal verification can be limiting; he prefers generalists who understand core principles and debugging techniques. He contrasts this with the past, where early AI tools often relied on the terminal (CLI) because building a full GUI was too slow. A key principle at Anthropic is enabling the model to use tools effectively rather than hardcoding features, which allows the model to remain useful even as capabilities evolve. He notes that Anthropic's core product code is heavily typed (TypeScript), which helps prevent common errors, unlike older, less rigorous systems. He also highlights that their internal tools are designed to be generalist-friendly, contrasting with hyper-specialist tools, and that their productivity has increased significantly, with internal code reviews showing 70% of public commits coming from Anthropic engineers, a figure he finds surprisingly high.

### LLM Building Philosophy

- Build for the model six months from now
- Don't hardcode features, enable tool use
- Generalists often outperform hyper-specialists in debugging/core principles

### Claude Code Evolution

- Constantly rewritten over the last six months
- Core code base is heavily typed (TypeScript) for safety
- Productivity per engineer has increased by 150% since last year

### Internal Tooling (QuantMD)

- Created accidentally to fix memory leaks in the CLI
- Served as Anthropic's initial debugging tool
- Led to insights about what features developers actually need

### Hiring Philosophy

- Value engineers who think scientifically (first principles)
- Prefer generalists who can adapt to new tools and concepts (like new LLM versions) over those reliant on specific, complex toolchains (like Vim).

### AI Safety & Code

- Anthropic engineers are highly safety-conscious, with many contributing to safety-related code reviews and suggesting improvements.

![Screenshot at 00:05: Boris Cherny discussing Anthropic's advice to founders to build for the future model, not the current one, as exemplified by the constant iteration of their internal code base.](https://ss.rapidrecap.app/screens/PQU9o_5rHC4/00-00-05.jpg)
![Screenshot at 00:27: Harj Taggar asks about the advice given to technical founders, prompting a discussion on building for future model capabilities.](https://ss.rapidrecap.app/screens/PQU9o_5rHC4/00-00-27.jpg)
![Screenshot at 00:58: Garry Tan recounting how Boris Cherny's internal tool, QuantMD, was created by accident to debug memory leaks.](https://ss.rapidrecap.app/screens/PQU9o_5rHC4/00-00-58.jpg)
![Screenshot at 02:07: Boris Cherny explaining that the early tools were intentionally kept simple \(like a terminal app\) to focus on core model interaction.](https://ss.rapidrecap.app/screens/PQU9o_5rHC4/00-02-07.jpg)
![Screenshot at 04:46: Boris Cherny noting that Anthropic's programming practices, like recursively written code and high test coverage, are indicative of their focus on long-term code health.](https://ss.rapidrecap.app/screens/PQU9o_5rHC4/00-04-46.jpg)