<?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>Edtech on orndorff.dev</title>
    <link>https://orndorff.dev/tags/edtech/</link>
    <description>Recent content in Edtech on orndorff.dev</description>
    <generator>Hugo -- 0.138.0</generator>
    <language>en-us</language>
    <lastBuildDate>Thu, 25 Jun 2026 00:00:00 +0000</lastBuildDate>
    <atom:link href="https://orndorff.dev/tags/edtech/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Want an LLM to Be a Domain Expert? Build It a Wiki — Automatically.</title>
      <link>https://orndorff.dev/posts/auto-wiki-domain-expert/</link>
      <pubDate>Thu, 25 Jun 2026 00:00:00 +0000</pubDate>
      <guid>https://orndorff.dev/posts/auto-wiki-domain-expert/</guid>
      <description>&lt;p&gt;Everyone wants their LLM to be an expert in &lt;em&gt;their&lt;/em&gt; thing — their product, their
policies, their field. The default playbook is to fine-tune a model on it. That&amp;rsquo;s
expensive, opaque, frozen the day you finish, and it still makes things up.&lt;/p&gt;
&lt;p&gt;There&amp;rsquo;s a cheaper, more honest move: don&amp;rsquo;t teach the model your domain. &lt;strong&gt;Hand it a
wiki.&lt;/strong&gt;&lt;/p&gt;
&lt;h2 id=&#34;the-idea&#34;&gt;The idea&lt;/h2&gt;
&lt;p&gt;Take your pile of documents. Have a model read all of it and write the wiki you wish
existed — every concept as its own page, every relationship as a link between pages, the
whole thing backed by a database the model can search and traverse.&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
