
  <rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
    <channel>
      <title>c4a4d65b&#39;s Blog</title>
      <link>https://c4a4d65b.xyz/blog</link>
      <description>Thoughts, notes, and explorations in tech</description>
      <language>en-us</language>
      <managingEditor>c4a4d65b@gmail.com (c4a4d65b)</managingEditor>
      <webMaster>c4a4d65b@gmail.com (c4a4d65b)</webMaster>
      <lastBuildDate>Thu, 04 Jun 2026 00:00:00 GMT</lastBuildDate>
      <atom:link href="https://c4a4d65b.xyz/tags/random-thoughts/feed.xml" rel="self" type="application/rss+xml"/>
      
  <item>
    <guid>https://c4a4d65b.xyz/blog/ai-equality-illusion</guid>
    <title>AI 平权的幻觉：真正的奇点不是人人变强，而是差距被重新放大</title>
    <link>https://c4a4d65b.xyz/blog/ai-equality-illusion</link>
    <description>AI 圈流行一种诱人的叙事——人人都将拥有超级助手，知识差距会被抹平。但工具普及不等于结果平均。从模型底层机制到社会结构，这篇文章想说明：AI 最深的影响未必是平权，而是重新制造不平等，而真正的机会就藏在这个错位里。</description>
    <pubDate>Thu, 04 Jun 2026 00:00:00 GMT</pubDate>
    <author>c4a4d65b@gmail.com (c4a4d65b)</author>
    <category>random-thoughts</category>
  </item>

  <item>
    <guid>https://c4a4d65b.xyz/blog/hello-world</guid>
    <title>The Great Wave, and the End of a Craft</title>
    <link>https://c4a4d65b.xyz/blog/hello-world</link>
    <description>AI 是否终结了传统漏洞研究？经过将近一年的 AI 审计实践，我的答案是——于我而言，是的。这是一篇关于告别、关于满意法则、关于从 0 到 1 的笔记。</description>
    <pubDate>Tue, 19 May 2026 00:00:00 GMT</pubDate>
    <author>c4a4d65b@gmail.com (c4a4d65b)</author>
    <category>random-thoughts</category>
  </item>

  <item>
    <guid>https://c4a4d65b.xyz/blog/mismatch-as-a-craft</guid>
    <title>从 Skip-Trigram 说到黑客精神</title>
    <link>https://c4a4d65b.xyz/blog/mismatch-as-a-craft</link>
    <description>这篇文章从 Transformer 里的 skip-trigram 现象出发，聊局部判断为什么会带来系统性误报，再把这个问题放到漏洞挖掘、AI 审计和对齐里看。最后我想说的是：真正能留下来的，也许不是某一套逆向技巧，而是黑客那种不信表面、非要把东西拆开看明白的劲头。</description>
    <pubDate>Thu, 25 Jun 2026 00:00:00 GMT</pubDate>
    <author>c4a4d65b@gmail.com (c4a4d65b)</author>
    <category>random-thoughts</category>
  </item>

    </channel>
  </rss>
