
  <rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
    <channel>
      <title>Articles by Dan Orlando</title>
      <link>https://danorlando.com/blog</link>
      <description>How-to&#39;s, discoveries and other useful nuggets fit for developer consumption.</description>
      <language>en-us</language>
      <managingEditor>undefined (Dan Orlando)</managingEditor>
      <webMaster>undefined (Dan Orlando)</webMaster>
      <lastBuildDate>Wed, 13 May 2026 00:00:00 GMT</lastBuildDate>
      <atom:link href="https://danorlando.com/tags/informationtheory/feed.xml" rel="self" type="application/rss+xml"/>
      
  <item>
    <guid>https://danorlando.com/blog/genai/multi-agent-is-a-trap</guid>
    <title>Multi-Agent AI Is a Trap</title>
    <link>https://danorlando.com/blog/genai/multi-agent-is-a-trap</link>
    <description>One team running a six-agent debate system switched to two agents with a strict state machine. Latency dropped from 18 seconds to 3. Cost per query dropped from $8-12 to $0.40. Accuracy changed by less than 1%. This wasn&#39;t a fluke. Information theory explains exactly why multi-agent systems can never outperform a single agent with full context.</description>
    <pubDate>Wed, 13 May 2026 00:00:00 GMT</pubDate>
    <author>undefined (Dan Orlando)</author>
    <category>generativeai</category><category>agenticai</category><category>architecture</category><category>systemdesign</category><category>agentsystems</category><category>informationtheory</category>
  </item>

  <item>
    <guid>https://danorlando.com/blog/genai/single-agent-vs-multi-agent</guid>
    <title>Single vs. Multi-Agent: The Cognition-Anthropic Schism and Why Both Are Right</title>
    <link>https://danorlando.com/blog/genai/single-agent-vs-multi-agent</link>
    <description>Two frontier teams published production data that directly contradicts each other. Cognition says multi-agent architectures cause compounding information loss and single-threaded execution is the fix. Anthropic says multi-agent delegation beat a single agent by 90.2% on their research eval. Both are right—for the model version they tested on. Architectural best practices in agent systems have a six-month half-life.</description>
    <pubDate>Mon, 25 May 2026 00:00:00 GMT</pubDate>
    <author>undefined (Dan Orlando)</author>
    <category>generativeai</category><category>agenticai</category><category>architecture</category><category>systemdesign</category><category>agentsystems</category><category>informationtheory</category>
  </item>

    </channel>
  </rss>
