<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Methodology on Yuanchang&#39;s Blog</title>
    <link>https://yuanchang.org/en/tags/methodology/</link>
    <description>Recent content in Methodology on Yuanchang&#39;s Blog</description>
    <generator>Hugo -- gohugo.io</generator>
    <language>en-us</language>
    <lastBuildDate>Mon, 23 Mar 2026 14:00:00 +0100</lastBuildDate><atom:link href="https://yuanchang.org/en/tags/methodology/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>I Built an AI Personal Evolution System: Voice-Driven, Auto-Distilled, Gets Smarter Over Time</title>
      <link>https://yuanchang.org/en/posts/ai-personal-evolution-system/</link>
      <pubDate>Mon, 23 Mar 2026 14:00:00 +0100</pubDate>
      
      <guid>https://yuanchang.org/en/posts/ai-personal-evolution-system/</guid>
      
      <description>AI models keep getting cheaper and smarter, but if they don&amp;rsquo;t know you, they&amp;rsquo;ll always give average-level answers. I built a system based on Apple&amp;rsquo;s Voice Memos, local transcription, AI auto-classification, and email notifications that continuously accumulates personal context. The longer you use it, the better it understands you. This article breaks down the architecture, cognitive framework, cost model, and remote terminal pitfalls.</description>
      
    </item>
    
    <item>
      <title>Why AI Only Gives You Correct Nonsense, and How to Push It Out of Its Comfort Zone</title>
      <link>https://yuanchang.org/en/posts/yage-context-infrastructure/</link>
      <pubDate>Mon, 16 Mar 2026 16:30:00 +0100</pubDate>
      
      <guid>https://yuanchang.org/en/posts/yage-context-infrastructure/</guid>
      
      <description>An LLM&amp;rsquo;s default output is consensus: correct but mediocre. Deep Research is really Wide Research. Ya Ge found a systematic way to pull LLMs out of consensus using personal cognitive context.</description>
      
    </item>
    
  </channel>
</rss>
