<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Artificial Intelligence on Techiavellian</title><link>https://techiavellian.com/categories/artificial-intelligence/</link><description>Recent content in Artificial Intelligence on Techiavellian</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Sun, 02 Nov 2025 00:00:00 +0000</lastBuildDate><atom:link href="https://techiavellian.com/categories/artificial-intelligence/index.xml" rel="self" type="application/rss+xml"/><item><title>Introducing smoke!</title><link>https://techiavellian.com/introducing-smoke/</link><pubDate>Sun, 02 Nov 2025 00:00:00 +0000</pubDate><guid>https://techiavellian.com/introducing-smoke/</guid><description>&lt;figure class="mono-figure mono-figure--center"&gt;&lt;img src="https://techiavellian.com/introducing-smoke/smoke-screenshot.png"
 alt="Screenshot of smoke" width="900px"&gt;
&lt;/figure&gt;

&lt;p&gt;I&amp;rsquo;ve found myself &lt;del&gt;unemployed&lt;/del&gt; with a bit more free time than usual for the last few months, and I decided that it was
the perfect opportunity to learn a bit about LLM tool use and to try out this whole &lt;a class="mono-link" href="https://x.com/karpathy/status/1886192184808149383"&gt;&amp;ldquo;vibe coding&amp;rdquo;&lt;/a&gt; thing.
Inspired in part by &lt;a class="mono-link" href="https://fly.io/blog/youre-all-nuts/"&gt;one blog post by Thomas Ptacek saying I was nuts not to use LLMs more&lt;/a&gt;
and &lt;a class="mono-link" href="https://ampcode.com/how-to-build-an-agent"&gt;another by Thorsten Ball explaining the nuts and bolts of &amp;ldquo;agents&amp;rdquo;&lt;/a&gt;, I set out to make one with
the right ergonomics and features just for &lt;em&gt;me&lt;/em&gt; - a CLI application I could run in tmux beside my NeoVim pane while
editing code.&lt;/p&gt;</description></item></channel></rss>