<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Consistency on Programmer.ie: Modern AI programming</title>
    <link>http://programmer.ie/tags/consistency/</link>
    <description>Recent content in Consistency on Programmer.ie: Modern AI programming</description>
    <generator>Hugo</generator>
    <language>en-us</language>
    <lastBuildDate>Sun, 30 Aug 2026 10:16:00 +0100</lastBuildDate>
    <atom:link href="http://programmer.ie/tags/consistency/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Beyond Hallucination: Consistency and Sensitivity</title>
      <link>http://programmer.ie/books/hallucination-from-first-principles/09-chapter/</link>
      <pubDate>Sun, 30 Aug 2026 10:16:00 +0100</pubDate>
      <guid>http://programmer.ie/books/hallucination-from-first-principles/09-chapter/</guid>
      <description>&lt;p&gt;Chapter 8 ended with a rule:&lt;/p&gt;&#xA;&lt;blockquote&gt;&#xA;&lt;p&gt;&lt;strong&gt;If a downstream decision depends on a distinction, do not discard that distinction before the decision is made.&lt;/strong&gt;&lt;/p&gt;&lt;/blockquote&gt;&#xA;&lt;p&gt;That rule breaks the idea of one universal hallucination score.&lt;/p&gt;&#xA;&lt;p&gt;A response can be well contained and still reverse a relation. It can preserve every relation and still ignore the decisive facts of the problem. It can be correct once and unstable under a harmless rephrasing. It can be perfectly repeatable and consistently wrong.&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
