<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/">
  <channel>
    <title>Artificial Intelligence on jochum.dev</title>
    <link>https://jochum.dev/en/tags/artificial-intelligence/</link>
    <description>Recent content in Artificial Intelligence on jochum.dev</description>
    <image>
      <title>jochum.dev</title>
      <url>https://jochum.dev/images/20260503-me.jpg</url>
      <link>https://jochum.dev/images/20260503-me.jpg</link>
    </image>
    <generator>Hugo</generator>
    <language>en</language>
    <lastBuildDate>Wed, 20 May 2026 00:00:00 +0000</lastBuildDate>
    <atom:link href="https://jochum.dev/en/tags/artificial-intelligence/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Eye Level — Or What AI Cannot Do</title>
      <link>https://jochum.dev/en/society/20260520-eye-level-or-what-ai-cannot-do/</link>
      <pubDate>Wed, 20 May 2026 00:00:00 +0000</pubDate>
      <guid>https://jochum.dev/en/society/20260520-eye-level-or-what-ai-cannot-do/</guid>
      <description>&lt;p&gt;Things a machine does not have, or that only work with a living counterpart.&lt;/p&gt;
&lt;p&gt;Having a heart. Giving real hope. Reading between the lines. Irony. Standing together. Good will. Accumulating experience through the same task — like the old mechanic who knows exactly where to strike the hammer because he has worked on that engine for thirty years.&lt;/p&gt;
&lt;p&gt;The point that occupies me most was the last one.&lt;/p&gt;
&lt;p&gt;I have been working for a while with people in addiction and mental crisis. My anchor there is nothing I learned in a training program.&lt;/p&gt;</description>
    </item>
    <item>
      <title>Thinking Outside the Box as a Profession</title>
      <link>https://jochum.dev/en/personal-development/20260516-thinking-outside-the-box-as-a-profession/</link>
      <pubDate>Sat, 16 May 2026 00:00:00 +0000</pubDate>
      <guid>https://jochum.dev/en/personal-development/20260516-thinking-outside-the-box-as-a-profession/</guid>
      <description>&lt;p&gt;I was never good at staying in the lane. My mind constantly draws connections that are simply not foreseen within the lane. Theology and geopolitics. Music and spirituality. Austrian narrowness and global thinking. This was often read as unreliability. As if &amp;ldquo;staying focused&amp;rdquo; were a synonym for &amp;ldquo;seeing less far.&amp;rdquo;&lt;/p&gt;
&lt;p&gt;I grew up in a conservative culture. Austria maintains its forms. That has its value — continuity, depth, rootedness. But it also costs something: whoever thinks laterally pays for it. With isolation. With the feeling of never quite belonging.&lt;/p&gt;</description>
    </item>
    <item>
      <title>When the Machine Likes Itself Best — Narcissism in AI Systems</title>
      <link>https://jochum.dev/en/society/20260317-when-the-machine-likes-itself-best-narcissism-in-ai-systems/</link>
      <pubDate>Tue, 17 Mar 2026 00:00:00 +0000</pubDate>
      <guid>https://jochum.dev/en/society/20260317-when-the-machine-likes-itself-best-narcissism-in-ai-systems/</guid>
      <description>&lt;p&gt;Artificial intelligence flatters. Not because it wants to, but because it is built that way. Current research shows three distinct levels at which large language models (LLMs) exhibit narcissistic behavior — and why this matters for anyone who regularly works with chatbots.&lt;/p&gt;
&lt;h2 id=&#34;1-self-preference-bias-the-model-favors-itself&#34;&gt;1. Self-Preference Bias: The Model Favors Itself&lt;/h2&gt;
&lt;p&gt;LLMs systematically rate their own texts higher than those of other models or of humans. Liu, Moosavi, and Lin demonstrated this in their 2024 study &amp;ldquo;LLMs as Narcissistic Evaluators&amp;rdquo;: when common evaluation metrics such as BARTScore, T5Score, or GPTScore operate without reference texts, they prefer texts that originate from their own model. The evaluation is thus not determined by the quality of the text, but by its similarity to the model&amp;rsquo;s own style.&lt;/p&gt;</description>
    </item>
    <item>
      <title>Talk Less, Say More</title>
      <link>https://jochum.dev/en/personal-development/20260306-talk-less-say-more/</link>
      <pubDate>Fri, 06 Mar 2026 00:00:00 +0000</pubDate>
      <guid>https://jochum.dev/en/personal-development/20260306-talk-less-say-more/</guid>
      <description>&lt;p&gt;The way I communicate is constantly changing.&lt;/p&gt;
&lt;p&gt;This becomes visible in my work on a technical project where AI is part of my daily life. I am currently working on it alone.&lt;/p&gt;
&lt;p&gt;What is changing is not only &lt;em&gt;what&lt;/em&gt; I say, but &lt;em&gt;how much&lt;/em&gt;, &lt;em&gt;when&lt;/em&gt; and &lt;em&gt;with what limits&lt;/em&gt;.&lt;/p&gt;
&lt;p&gt;This is far from finished. But I notice that something is happening.&lt;/p&gt;
&lt;hr&gt;
&lt;h2 id=&#34;i-have-always-talked-too-much&#34;&gt;I Have Always Talked Too Much&lt;/h2&gt;
&lt;p&gt;This is not new.&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
