Podcast insights
What Boris Cherny says about building Claude Code
Boris Cherny created Claude Code inside Anthropic Labs and, by his own account, hasn't written a line of code in his job for months. Across four podcast appearances (WorkOS, Y Combinator, Lenny's Podcast, and Sequoia's Training Data) he keeps returning to the same ideas: models could already do far more than any product had captured, the best dev tools are built for yourself first, and 'late demand' should drive what you build next. Quotes are verbatim; sign in to read the full transcripts.
From 4 episodes on PodLearn
The 'product overhang': models could already do more than any product captured
There's this idea that the model can do all this stuff that no product has yet captured.
Build the tool for yourself first
Build the thing for yourself first and then hopefully it's useful for others.
'Late demand' is the single biggest product principle
Probably the single, for me, biggest principle in product is late demand.
Inside Claude Code With Its Creator Boris Cherny · Y Combinator
The first real use case was writing unit tests — lower risk
The first use case was actually writing unit tests, because it's a little bit lower risk.
Inside Claude Code With Its Creator Boris Cherny · Y Combinator
To do good AI work, you have to understand the model itself
If you're working in AI, you just really have to understand the model to some degree to do good work.
Deliberately under-resource a new product at the start
You want to under resource things a little bit at the start.
Episodes this draws from
- Boris Cherny: Claude Code & the Future of Engineering
WorkOS
- Inside Claude Code With Its Creator Boris Cherny
Y Combinator
- Head of Claude Code: What happens after coding is solved
Lenny's Podcast
- Anthropic's Boris Cherny: Coding's Printing Press Moment
Training Data (Sequoia)
Full transcripts are free to read with a PodLearn account.
FAQ
Who is Boris Cherny?
Boris Cherny is the creator of Claude Code at Anthropic. He started it inside an incubator called Anthropic Labs in late 2024, and has said he hasn't written a line of code in his job for months because Claude Code does it.
Why is Claude Code a terminal (CLI) tool?
Cherny says he built it in a terminal because for the first couple of months it was just him, so the CLI was the easiest thing to build. The team kept it because the model was improving so fast that no heavier form factor could keep up.
What is Boris Cherny's top product principle?
Late demand. He says the single biggest principle in product is building for demand that shows up after you ship something small, and that nearly every part of Claude Code was built that way — CLAUDE.md, for example, came from users writing their own markdown files for the model to read.
What was the first real use case for Claude Code?
Writing unit tests. Cherny says it was an early use case because it was lower risk while the model was still improving, alongside automating Git and bash commands.
Read every word — free
Every quote above links to its full, searchable transcript on PodLearn. Create a free account and read, search, and turn any episode into a lesson.