<?xml version="1.0" encoding="UTF-8"?>
<?xml-stylesheet href="/feed.rss.xml" type="text/xsl" media="screen"?>
<rss version="2.0" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:media="http://search.yahoo.com/mrss/" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Priyo Harjiyono</title>
    <description/>
    <link>https://speakerdeck.com/orionclaw</link>
    <atom:link rel="self" type="application/rss+xml" href="https://speakerdeck.com/orionclaw.rss"/>
    <lastBuildDate>2026-09-03 01:00:43 -0400</lastBuildDate>
    <item>
      <title>How LLMs Recognize &amp; Mention New Brands: Insights from Testing 9 AI Models</title>
      <description>Ever wondered how AI tools like ChatGPT, Claude, or Gemini decide whether to mention a new brand? We tested 9 major AI models on a real-world brand that recently underwent a name, domain, and core business change.

Key Takeaways &amp; Insights:

Self-Declared Content vs. External Proof: 9 out of 9 AI models treat on-site website content as unverified claims. External citations and third-party mentions are required to validate brand identity.

Legacy Data Persistence: Old digital footprints and unupdated structured data (e.g., Wikidata) can confuse AI models up to 4+ months after a pivot.

Source Hierarchy &amp; Citation Triggers: Discover how AI models prioritize primary docs vs. third-party sources and when they feel compelled to attribute links.

Disambiguation Challenge: Why new brands must focus on entity disambiguation rather than just adding more site content.

Based on original research and empirical test results by Komunitech AI Research.</description>
      <media:content url="https://files.speakerdeck.com/presentations/510c39fb7da64020a9d51bb5b3724c1f/preview_slide_0.jpg?40449247" type="image/jpeg" medium="image"/>
      <content:encoded>Ever wondered how AI tools like ChatGPT, Claude, or Gemini decide whether to mention a new brand? We tested 9 major AI models on a real-world brand that recently underwent a name, domain, and core business change.

Key Takeaways &amp; Insights:

Self-Declared Content vs. External Proof: 9 out of 9 AI models treat on-site website content as unverified claims. External citations and third-party mentions are required to validate brand identity.

Legacy Data Persistence: Old digital footprints and unupdated structured data (e.g., Wikidata) can confuse AI models up to 4+ months after a pivot.

Source Hierarchy &amp; Citation Triggers: Discover how AI models prioritize primary docs vs. third-party sources and when they feel compelled to attribute links.

Disambiguation Challenge: Why new brands must focus on entity disambiguation rather than just adding more site content.

Based on original research and empirical test results by Komunitech AI Research.</content:encoded>
      <pubDate>Thu, 03 Sep 2026 00:00:00 -0400</pubDate>
      <link>https://speakerdeck.com/orionclaw/how-llms-recognize-and-mention-new-brands-insights-from-testing-9-ai-models</link>
      <guid>https://speakerdeck.com/orionclaw/how-llms-recognize-and-mention-new-brands-insights-from-testing-9-ai-models</guid>
    </item>
  </channel>
</rss>
