<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>LLM on Mr.Yam</title><link>https://mr.nickyam.com/tags/llm/</link><description>Recent content in LLM on Mr.Yam</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Wed, 21 Oct 2026 08:17:00 +0800</lastBuildDate><atom:link href="https://mr.nickyam.com/tags/llm/index.xml" rel="self" type="application/rss+xml"/><item><title>Why AI Output Gets Worse the More You Edit It (and How to Fix It)</title><link>https://mr.nickyam.com/Tech/Why-AI-Output-Gets-Worse-When-Edited/</link><pubDate>Wed, 21 Oct 2026 08:17:00 +0800</pubDate><guid>https://mr.nickyam.com/Tech/Why-AI-Output-Gets-Worse-When-Edited/</guid><description>&lt;p&gt;A common observation: AI-generated content looks decent on the first pass, but ask the model to revise it and the result often turns into a mess. Why?&lt;/p&gt;
&lt;p&gt;Two AI assistants were asked to explain. Their answers are reproduced below.&lt;/p&gt;
&lt;h2 id="deepseeks-explanation-1"&gt;DeepSeek&amp;rsquo;s explanation [1]&lt;/h2&gt;
&lt;p&gt;The phenomenon is widespread enough that the industry even has a phrase for it: &amp;ldquo;the more you edit AI copy, the worse it gets.&amp;rdquo; The cause is not a single factor but a stack of several mechanisms:&lt;/p&gt;</description></item></channel></rss>