<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Ai on Blogarious</title>
    <link>https://blogarious.kairopath.net/categories/ai/</link>
    <description>Recent content in Ai on Blogarious</description>
    <generator>Hugo</generator>
    <language>en_US</language>
    <lastBuildDate>Sun, 03 May 2026 21:31:46 +0800</lastBuildDate>
    <atom:link href="https://blogarious.kairopath.net/categories/ai/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Hallucination</title>
      <link>https://blogarious.kairopath.net/posts/20260503-hallucination/</link>
      <pubDate>Sun, 03 May 2026 21:31:46 +0800</pubDate>
      <guid>https://blogarious.kairopath.net/posts/20260503-hallucination/</guid>
      <description>&lt;p&gt;In my other post, I mentioned the limitations of the AI:&lt;/p&gt;&#xA;&lt;ul&gt;&#xA;&lt;li&gt;AI models will never be truly up-to-date because they are trained on existing knowledge.&lt;/li&gt;&#xA;&lt;li&gt;AI is based on generalization; it can never be truly specific - therefore, its so-called suggestions tend to be general.&lt;/li&gt;&#xA;&lt;li&gt;AI cannot create &amp;ldquo;ex nihilo&amp;rdquo; out of nothing - its so-called novelty mainly comes from different combinations.&lt;/li&gt;&#xA;&lt;li&gt;AI cannot learn from its own generated content - this would lead to model collapse.&lt;/li&gt;&#xA;&lt;/ul&gt;&#xA;&lt;p&gt;But I didn&amp;rsquo;t include another common limitation based on the current technology: &lt;strong&gt;hallucination&lt;/strong&gt;. In my opinion, this will probably be overcome in the future, because technically the chance of hallucination can be reduced.&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
