<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
<channel>
<title>紫衡 ZiHeng Group — 洞察</title>
<link>https://ziheng-hk.vercel.app/</link>
<description>三個產品,一種做法:AI 給答案,系統留證據。與其信,不如驗。Three products, one way: AI gives the answer, the system keeps the evidence.</description>
<language>zh-Hant</language>
<atom:link href="https://ziheng-hk.vercel.app/feed.xml" rel="self" type="application/rss+xml"/>
<item>
  <title>業績,點樣驗?五條問題分得出真假 AI 選股紀錄</title>
  <link>https://ziheng-hk.vercel.app/insights/verify</link>
  <guid isPermaLink="true">https://ziheng-hk.vercel.app/insights/verify</guid>
  <pubDate>Fri, 24 Oct 2026 12:00:00 +0800</pubDate>
  <description>封存預測、簽名淨值、可回放歷史、第三方對照、認衰邊界——任何 AI 選股工具答唔出呢五條,答案已經喺度。Five questions to verify any AI stock picker's track record.</description>
</item>
<item>
  <title>SFC 講 AI 管治,你啲工具跟唔跟得上?</title>
  <link>https://ziheng-hk.vercel.app/insights/sfc-ai</link>
  <guid isPermaLink="true">https://ziheng-hk.vercel.app/insights/sfc-ai</guid>
  <pubDate>Fri, 24 Oct 2026 12:00:00 +0800</pubDate>
  <description>SFC 對 AI 語言模型嘅四條線:問責、人類監督、能力、通知——同一套「可問責」設計點樣落地。The SFC's four AI expectations mapped to a no-source-no-answer architecture.</description>
</item>
</channel>
</rss>
