Qirui's Blog

My First Vibe Coding Experience: Building lf.nvim

The Itch I use Lingua Franca -- a polyglot coordination language for concurrent and distributed systems. The official tooling lives in VSCode. I live in Neovim. That gap had been bugging me for a while. On November 22, 2025, I sat down and decided to do something about it. Not

Quoting Paul Graham

The reason AI coding works so well is that the source code of the median app was already slop before LLMs. — Paul Graham AI seems surprisingly competent at writing code not because it’s truly “intelligent,” but because the bar it needs to clear — the median human code quality — isn’

Why Fewer MCP Tools (and More CLI) Makes Your LLM Smarter

Every week, I see people installing dozens of Model Context Protocol (MCP) servers into their coding assistants and then wondering why the model feels sluggish, inconsistent, or forgetful. The culprit isn’t the model — it’s context bloat. Context Is Finite An LLM’s context window is like scratch paper.

The state of GPU capabilities in U.S. and China by the end of 2024

Below views are integrated from the sources I listed, which do not guarantee to be facts. Current advancements in AI depend heavily on increased compute density, which fundamentally relates to power density that restricts data center capacity. While nuclear power may be essential in the long term, U.S. hyperscalers

Coming soon

This is Qirui's Blog, a brand new site by Qirui FAN that's just getting started. Things will be up and running here shortly, but you can subscribe in the meantime if you'd like to stay up to date and receive emails when new content

Qirui's Blog © 2026