<?xml version="1.0" encoding="UTF-8"?>


<feed xmlns="http://www.w3.org/2005/Atom">
    <id>https://www.julianwyngaard.com/quotations/feed.xml</id>
    <title>Loosely Coupled Thoughts &#x2014; quotations</title>
    <subtitle>A blog about software engineering, AI, and development.</subtitle>
    <link rel="self" type="application/atom+xml" href="https://www.julianwyngaard.com/quotations/feed.xml"/>
    <link rel="alternate" type="text/html" href="https://www.julianwyngaard.com/quotations"/>
    <updated>2026-06-29T08:25:20.106184+12:00</updated>
    <author>
        <name>Julian Wyngaard</name>
    </author>
    <entry>
        <id>urn:uuid:019f0fe8-60c4-7a2a-90d0-ea59c71bdefc</id>
        <title>Quoting Matt Van Horn</title>
        <link rel="alternate" type="text/html" href="https://www.julianwyngaard.com/2026/06/29/matt-van-horn"/>
        <published>2026-06-29T08:25:20.106184+12:00</published>
        <updated>2026-06-29T08:25:20.106184+12:00</updated>
        <category term="ai" label="ai"/>
        <category term="engineering-with-ai" label="engineering-with-ai"/>
        <category term="harness-engineering" label="harness-engineering"/>
        <category term="agentic-workflows" label="agentic-workflows"/>
        <category term="matt-van-horn" label="matt-van-horn"/>
        <category term="agentic-loops" label="agentic-loops"/>
        <content type="html">&lt;blockquote&gt;&lt;p&gt;An open loop that writes code with no feedback is a machine for generating confident mistakes. A loop that writes, runs, reads the result, and corrects is the thing that actually works. The loop is not the magic. The feedback inside it is.&lt;/p&gt;&lt;/blockquote&gt;&lt;p&gt;&#x2014; &lt;a href=&quot;https://x.com/mvanhorn/status/2063865685558903149&quot;&gt;Matt Van Horn&lt;/a&gt;&lt;/p&gt;</content>
    </entry>
    <entry>
        <id>urn:uuid:019daf76-58d6-7745-90bb-d89680fe9636</id>
        <title>Quoting Ryan Lopopolo</title>
        <link rel="alternate" type="text/html" href="https://www.julianwyngaard.com/2026/04/21/ryan-lopopolo"/>
        <published>2026-04-21T21:51:00+12:00</published>
        <updated>2026-04-21T21:54:25.189256+12:00</updated>
        <category term="engineering-with-ai" label="engineering-with-ai"/>
        <category term="harness-engineering" label="harness-engineering"/>
        <category term="ryan-lopopolo" label="ryan-lopopolo"/>
        <content type="html">&lt;blockquote&gt;&lt;p&gt;Every company should have a full stack team of 5 building a product who are banned from directly writing code; they must force the agents to do it. In 2 months they will be your most productive team.&lt;/p&gt;&lt;/blockquote&gt;&lt;p&gt;&#x2014; &lt;a href=&quot;https://x.com/_lopopolo/status/1991024524410958066?s=46&quot;&gt;Ryan Lopopolo&lt;/a&gt;&lt;/p&gt;</content>
    </entry>
    <entry>
        <id>urn:uuid:019da937-d113-760e-8be5-784cbb95b725</id>
        <title>Quoting Devansh</title>
        <link rel="alternate" type="text/html" href="https://www.julianwyngaard.com/2026/04/20/devansh"/>
        <published>2026-04-20T16:48:23.277771+12:00</published>
        <updated>2026-04-20T16:48:23.277771+12:00</updated>
        <category term="ai" label="ai"/>
        <category term="engineering-with-ai" label="engineering-with-ai"/>
        <category term="llms" label="llms"/>
        <content type="html">&lt;blockquote&gt;&lt;p&gt;This is kind of a big deal. The performance difference between Cursor and Claude Code? Or Claude Code and OpenCode? All that comes from the system around the model. That&#x2019;s why Google can have one of the best models in the market, and still produce the Shakespearean tragedy that is the Gemini CLI. When people say &#x201c;Mythos found a zero-day,&#x201d; the truth is more &#x201c;Mythos, orchestrated by a purpose-built vulnerability research pipeline, found a zero-day.&lt;/p&gt;&lt;/blockquote&gt;&lt;p&gt;&#x2014; &lt;a href=&quot;https://substack.com/@chocolatemilkcultleader/note/c-246066432&quot;&gt;Devansh&lt;/a&gt;&lt;/p&gt;</content>
    </entry>
</feed>
