Three layers of AI safety at Anthropic

Boris Cherny is the creator and head of Claude Code at Anthropic. What began as a simple terminal-based prototype just a year ago has transformed the role of software engineering and is increasingly transforming all professional work.

We discuss:

1. How Claude Code grew from a quick hack to 4% of public GitHub commits, with daily active users doubling last month

2. The counterintuitive product principles that drove Claude Code’s success

3. Why Boris believes coding is “solved”

4. The latent demand that shaped Claude Code and Cowork

5. Practical tips for getting the most out of Claude Code and Cowork

6. How underfunding teams and giving them unlimited tokens leads to better AI products

7. Why Boris briefly left Anthropic for Cursor, then returned after just two weeks

8. Three principles Boris shares with every new team member

—

Brought to you by:

DX—The developer intelligence platform designed by leading researchers: https://getdx.com/lenny

Sentry—Code breaks, fix it faster: https://sentry.io/lenny

Metaview—The AI platform for recruiting: https://metaview.ai/lenny

—

Episode transcript: https://www.lennysnewsletter.com/p/head-of-claude-code-what-happens

—

Archive of all Lenny's Podcast transcripts: https://www.dropbox.com/scl/fo/yxi4s2w998p1gvtpu4193/AMdNPR8AOw0lMklwtnC0TrQ?rlkey=j06x0nipoti519e0xgm23zsn9&st=ahz0fj11&dl=0

—

Where to find Boris Cherny:

• X: https://x.com/bcherny

• LinkedIn: https://www.linkedin.com/in/bcherny

• Website: https://borischerny.com

—

Where to find Lenny:

• Newsletter: https://www.lennysnewsletter.com

• X: https://twitter.com/lennysan

• LinkedIn: https://www.linkedin.com/in/lennyrachitsky/

—

In this episode, we cover:

(00:00) Introduction to Boris and Claude Code

(03:45) Why Boris briefly left Anthropic for Cursor (and what brought him back)

(05:35) One year of Claude Code

(08:41) The origin story of Claude Code

(13:29) How fast AI is transforming software development

(15:01) The importance of experimentation in AI innovation

(16:17) Boris’s current coding workflow (100% AI-written)

(17:32) The next frontier

(22:24) The downside of rapid innovation 

(24:02) Principles for the Claude Code team

(26:48) Why you should give engineers unlimited tokens

(27:55) Will coding skills still matter in the future?

(32:15) The printing press analogy for AI’s impact

(36:01) Which roles will AI transform next?

(40:41) Tips for succeeding in the AI era

(44:37) Poll: Which roles are enjoying their jobs more with AI

(46:32) The principle of latent demand in product development

(51:53) How Cowork was built in just 10 days

(54:04) The three layers of AI safety at Anthropic

(59:35) Anxiety when AI agents aren’t working

(01:02:25) Boris’s Ukrainian roots

(01:03:21) Advice for building AI products

(01:08:38) Pro tips for using Claude Code effectively

(01:11:16) Thoughts on Codex

(01:12:13) Boris’s post-AGI plans

(01:14:02) Lightning round and final thoughts

—

References: https://www.lennysnewsletter.com/p/head-of-claude-code-what-happens

—

Production and marketing by https://penname.co/. For inquiries about sponsoring the podcast, email podcast@lennyrachitsky.com.

—

Lenny may be an investor in the companies discussed.