<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
<channel>
  <title>MATRIX · Dessa</title>
  <link>https://derekwang2002.github.io/ai-intelligence/en/projects/dessa/</link>
  <description>Material changes in AI models, agents, engineering and infrastructure, with primary sources and recommendations.</description>
  <language>en</language>
  <lastBuildDate>Mon, 02 Nov 2026 16:56:00 GMT</lastBuildDate>
  <atom:link href="https://derekwang2002.github.io/ai-intelligence/en/feeds/projects/dessa.xml" rel="self" type="application/rss+xml"/>
  <item>
    <title>Block previews Dessa, a speech-only full-duplex conversation model</title>
    <link>https://derekwang2002.github.io/ai-intelligence/en/events/ev-20261023-06/</link>
    <guid isPermaLink="false">ev-20261023-06-published</guid>
    <pubDate>Sat, 24 Oct 2026 15:59:49 GMT</pubDate>
    <category>New</category>
    <category>WATCH</category>
    <description>&lt;p&gt;Block previewed Dessa, a speech-to-speech model trained from scratch on human conversation without text or a pretrained model. It listens and speaks simultaneously and runs faster than real time, according to the team. The page provides selected audio demonstrations but no weights, code, API, benchmark protocol or independent evaluation.&lt;/p&gt;&lt;p&gt;Recommendation: WATCH&lt;/p&gt;&lt;p&gt;Sources: &lt;a href=&quot;https://block.xyz/dessa&quot;&gt;Dessa: the conversation model trained only on speech&lt;/a&gt;&lt;/p&gt;</description>
  </item>
</channel>
</rss>
