<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Agentic Loops on Bruce on AI Engineering</title><link>http://www.heyuan110.com/tags/agentic-loops/</link><description>Recent content in Agentic Loops on Bruce on AI Engineering</description><generator>Hugo</generator><language>en</language><lastBuildDate>Fri, 03 Jul 2026 14:00:00 +0800</lastBuildDate><atom:link href="http://www.heyuan110.com/tags/agentic-loops/index.xml" rel="self" type="application/rss+xml"/><item><title>Agentic Loops 2026: Self-Looping AI Agents Explained</title><link>http://www.heyuan110.com/posts/ai/2026-07-03-agentic-loops/</link><pubDate>Fri, 03 Jul 2026 14:00:00 +0800</pubDate><guid>http://www.heyuan110.com/posts/ai/2026-07-03-agentic-loops/</guid><description>&lt;p&gt;&lt;img src="http://www.heyuan110.com/posts/ai/2026-07-03-agentic-loops/cover.webp"
 alt="Agentic loops and self-looping AI agents in 2026, from ReAct to the Ralph loop"
 
 loading="lazy"
 decoding="async"
 fetchpriority="low"
 width="1200"
 height="630"
/&gt;
&lt;/p&gt;
&lt;p&gt;Here is the one sentence that separates a chatbot from an agent: a chatbot answers one prompt in one pass, an &lt;strong&gt;agent runs a while loop that carries state across turns&lt;/strong&gt;. That loop — reason, act, observe, repeat — is the entire mechanism. Everything else in the 2026 agent stack is decoration on top of it. And in 2026 the loop grew up: agents stopped being limited to a single request-response exchange and started running for minutes, then hours, unattended.&lt;/p&gt;</description></item></channel></rss>