ai-agents code-generation developer-tools y-combinator

YC CEO says he ships 37K LoC AI code per day. A developer looked under the hood

Y Combinator's CEO says he ships 37,000 lines of AI code per day. A developer looked under the hood

fastcompany.com

July 7, 2026

3 min read

🔥🔥🔥🔥🔥

52/100

Summary

Y Combinator CEO Garry Tan claims to deploy 37,000 lines of AI code daily across five projects. He emphasizes the significance of "agentic engineering" in his recent social media posts.

Key Takeaways

Y Combinator CEO Garry Tan claims to deploy 37,000 lines of AI code daily across five projects and is on a 72-day shipping streak.
A developer found significant inefficiencies in Tan's website, including excessive server requests and bloated code, highlighting that quantity of code does not equate to quality.
The website's performance issues included loading 169 server requests totaling 6.42 megabytes, compared to 12 kilobytes for the minimalist Hacker News homepage.
Gregorein's review revealed numerous coding mistakes, such as duplicate content, unnecessary files, and outdated image formats, emphasizing the need for proper scrutiny in AI-generated code.

Read original article

View story →

Coverage & discussions

Hacker News · fastcompany.com

YC CEO says he ships 37K LoC AI code per day. A developer looked under the hood ↗

Community discussions

Hacker News discussion ↗

Historical count: 5 · 5 saved in sample · Historical sample

Analysis ready

Community Sentiment

Mixed

Positives

Some developers are leveraging AI tools to boost productivity, allowing them to produce higher quality software at the same pace as before, which is a game changer.
Despite concerns, there’s an argument that the current state of code quality was already declining before LLMs, so it's not all on AI.

Concerns

The sheer volume of code being produced raises alarms about quality; many fear that bloated and inefficient code is becoming the norm, not the exception.
Critics argue that measuring success by lines of code is misguided, as it often results in unnecessary complexity instead of effective solutions.
Access to AI tools doesn't equate to expertise; many users are churning out code without understanding its quality, which is a recipe for disaster.