summarize https://www.youtube.com/watch?v=PQU9o_5rHC4 and pull out interesting Garry Tan quotes
Guests: Boris Cherny (creator of Claude Code at Anthropic) with Garry Tan and the YC team.
The episode is a deep dive into how Claude Code was built, why it works the way it does, and what it signals about the future of software engineering.
"Thanks for creating a thing that has taken away my sleep for about 3 weeks straight. I am very addicted to Claude Code and it feels like rocket boosters."
— Opening the interview, on his personal obsession with the tool.
"It's very kind of contrarian that Claude Code works so well in such an elegant simple form factor… it doesn't feel like working, it just feels fun as a developer."
— On why a CLI — an "old" interface — turns out to be the ideal form factor.
"I love the verbosity because basically sometimes it just goes off the deep end and I'm watching… and I'm like 'Oh no no it's not that.' And then I escape and stop it."
— Why he wants to see the model's raw output — it lets him intervene before an agent goes down the wrong path.
"It's like an auto-bug fixing and test making… making a startup factory."
— Describing the shift from manually debugging to just handing logs to the agent and watching it fix itself.
"Do you think you would ever hire someone based on the Claude Code transcript of them working with the agent? Because we're actively doing that right now."
— On how YC is already evaluating candidates by their ability to "pilot" an AI agent, not just write code themselves.
"It's making me feel creator mode again, sort of founder mode again. It's been an exhilarating three weeks."
— Closing remarks: these tools let founders return to the maker stage, bypassing coordination bottlenecks.
The thread running through all of Garry's comments: He's not talking about productivity as an abstract metric — he's describing a feeling he thought he'd lost. Builder's joy, founder's velocity, creator mode. That's the real sell.