<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Ai on Kevin's blog</title><link>https://blog.kevinroyer.com/tags/ai/</link><description>Recent content in Ai on Kevin's blog</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Sun, 18 Oct 2026 07:36:00 -0700</lastBuildDate><atom:link href="https://blog.kevinroyer.com/tags/ai/index.xml" rel="self" type="application/rss+xml"/><item><title>Quoting Sean Goedecke</title><link>https://blog.kevinroyer.com/blog/quoting-sean-goedecke-writing-proof-of-work/</link><pubDate>Sun, 18 Oct 2026 07:36:00 -0700</pubDate><guid>https://blog.kevinroyer.com/blog/quoting-sean-goedecke-writing-proof-of-work/</guid><description>&lt;blockquote&gt;
&lt;p&gt;In other words: don&amp;rsquo;t trust or waste time poring through LLM-authored documents yourself, but also don&amp;rsquo;t waste time writing documents by hand that are just intended to serve as evidence. Save that effort for documents that are intended to serve as communication.&lt;/p&gt;</description></item><item><title>All You Need Is Prod</title><link>https://blog.kevinroyer.com/blog/all-you-need-is-prod/</link><pubDate>Sat, 17 Oct 2026 18:30:00 -0700</pubDate><guid>https://blog.kevinroyer.com/blog/all-you-need-is-prod/</guid><description>&lt;p&gt;&lt;em&gt;AI for AI quality&lt;/em&gt;&lt;/p&gt;
&lt;p&gt;Evals are only as good as the environment they run in. Production is the only environment that is always accurate: it&amp;rsquo;s where failure modes show up, and it should be where they get fixed. So instead of building synthetic environments, treat a read-only mirror of production as the fixture for experiments.&lt;/p&gt;</description></item><item><title>Loosely quoting Charles Robin</title><link>https://blog.kevinroyer.com/blog/loosely-quoting-charles-robin/</link><pubDate>Sat, 15 Aug 2026 09:23:49 -0700</pubDate><guid>https://blog.kevinroyer.com/blog/loosely-quoting-charles-robin/</guid><description>&lt;h1 id="amino-acids---human---ai"&gt;Amino Acids -&amp;gt; Human -&amp;gt; AI?&lt;/h1&gt;
&lt;p&gt;Inspired by &lt;a href="https://www.le-precepteur.fr/qui-suis-je"&gt;Charles Robin&lt;/a&gt; interviewed in the French &amp;#x1f1eb;&amp;#x1f1f7; podcast &lt;a href="https://open.spotify.com/show/0KK0ym48IBJd37maEjfDkI?si=f3f5b6e8f5c34686"&gt;Continue tu m&amp;rsquo;intéresses&lt;/a&gt;.&lt;/p&gt;
&lt;p&gt;Looking at Kubrick&amp;rsquo;s proposal in 2001: A Space Odyssey (1968, a little bit before &lt;em&gt;our&lt;/em&gt; AIs):&lt;/p&gt;
&lt;ol&gt;
&lt;li&gt;Human gets limited by its body.&lt;/li&gt;
&lt;li&gt;Human makes a tool as an extension of itself, a bone does count here.&lt;/li&gt;
&lt;/ol&gt;
&lt;p&gt;[&amp;hellip;a few moments later&amp;hellip;]&lt;/p&gt;</description></item></channel></rss>