<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Knowledge-Work on Marcos Souza</title>
    <link>https://www.msouza.tech/tags/knowledge-work/</link>
    <description>Recent content in Knowledge-Work on Marcos Souza</description>
    <generator>Hugo -- gohugo.io</generator>
    <language>en</language>
    <lastBuildDate>Tue, 09 Jun 2026 17:04:53 -0300</lastBuildDate>
    <atom:link href="https://www.msouza.tech/tags/knowledge-work/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>AI Workflows That Protect Thinking</title>
      <link>https://www.msouza.tech/posts/tech/ai-workflows-that-protect-thinking/</link>
      <pubDate>Tue, 09 Jun 2026 17:04:53 -0300</pubDate>
      <guid>https://www.msouza.tech/posts/tech/ai-workflows-that-protect-thinking/</guid>
      <description>AI Workflows That Protect Thinking Link to heading Over the last few days, I kept running into the same realization: the most useful AI-assisted workflows are the ones that leave human judgment at the center.&#xA;That sounds obvious until you look at how AI is usually discussed. One side treats it like a universal accelerator. The other treats it like a machine for intellectual decay. Both positions miss the part that interests me most.</description>
    </item>
  </channel>
</rss>
