<?xml version="1.0" encoding="utf-8"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Posts · Mesh Refinement</title><link>https://meshrefine.com/en/tags/codex/</link><description>Posts · Mesh Refinement</description><atom:link href="https://meshrefine.com/en/tags/codex/index.xml" rel="self" type="application/rss+xml"/><lastBuildDate>Wed, 04 Jun 2025 10:12:42 +0200</lastBuildDate><item><title>OpenAI Codex Gains Internet Access: First Impressions</title><link>https://meshrefine.com/en/posts/openai-codex-internet-access/</link><description>&lt;h2 id="f05970d8e8e7ace55eca4844553fe66b-what-on-earth-is-codex"&gt;What on Earth is Codex?&lt;/h2&gt;
&lt;p&gt;Good question, right? The thing is, until recently, OpenAI had a model called Codex, which was used as the foundation for autocompletion in GitHub Copilot. Then, OpenAI released a console agent for development, which they named, so no one would get confused, &lt;a href="https://github.com/openai/codex"&gt;Codex&lt;/a&gt;. Everyone had a laugh at OpenAI&amp;rsquo;s naming skills , and life went on. Until the fateful day when a tweet like this appeared from Sam Altman:&lt;/p&gt;</description><pubDate>Wed, 04 Jun 2025 10:12:42 +0200</pubDate><guid>https://meshrefine.com/en/posts/openai-codex-internet-access/</guid></item></channel></rss>