<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/">
  <channel>
    <title>emre şahin's digital garden 🍃 - ai</title>
    <link>https://emresahin.net/tags/ai/</link>
    <description>Posts in the ai tag</description>
    <language>en</language>
    <managingEditor>contact@emresahin.net (Emre Şahin)</managingEditor>
    <lastBuildDate>Tue, 15 Sep 2026 19:46:32 +0000</lastBuildDate>
    <atom:link href="https://emresahin.net/tags/ai/rss.xml" rel="self" type="application/rss+xml"/>
    <item>
      <title>bits 17</title>
      <published>2025-09-28T17:20:37+00:00</published>
      <updated>2025-09-28T17:20:37+00:00</updated>
      <author>Emre Şahin</author>
      <pubDate>Sun, 28 Sep 2025 17:20:37 +0000</pubDate>
      <link>https://emresahin.net/bits-17/</link>
      <guid isPermaLink="true">https://emresahin.net/bits-17/</guid>
      <description>I wanted to calculate the number of words added between commits in my blog posts. I asked Gemini for a Nushell pipeline based on git diff : It proposed the following: git diff --word-diff=plain -- "*.md" | rg -o '\{\+(.*?)\+\}' | str replace -r '\{\+|\+\}' '' | str join " " | str words | length H...</description>
      <category>shell</category>
      <category>productivity</category>
      <category>Gemini</category>
      <category>nushell</category>
      <category>git</category>
      <category>regex</category>
      <category>ai</category>
      <content:encoded><![CDATA[<p>I wanted to calculate the number of words added between commits in my blog posts. I asked Gemini for a Nushell pipeline based on <code>git diff</code>:</p>
<p>It proposed the following:</p>
<pre><code class="language-nushell">git diff --word-diff=plain -- "*.md" | rg -o '\{\+(.*?)\+\}' | str replace -r '\{\+|\+\}' '' | str join " " | str words | length
</code></pre>
<p>However, I ended up using this simplified version:</p>
<pre><code class="language-nushell">git diff --word-diff=plain -- "*.md" | rg -o '\{\+(.*?)\+\}' | str replace -a "{+" "" | str replace -a "+}" "" | str replace -a ' ' "\n" | lines | length
</code></pre>
<p>I’ve noticed that LLMs often lean toward overly complex solutions. I’m not sure if this is due to being trained to generate as many tokens as possible, but using regular expressions where a simple string replacement would suffice is often a sign of unnecessary complexity.</p>]]></content:encoded>
    </item>
    <item>
      <title>bits 5</title>
      <published>2023-03-14T11:24:00+00:00</published>
      <updated>2023-03-14T11:24:00+00:00</updated>
      <author>Emre Şahin</author>
      <pubDate>Tue, 14 Mar 2023 11:24:00 +0000</pubDate>
      <link>https://emresahin.net/bits-5/</link>
      <guid isPermaLink="true">https://emresahin.net/bits-5/</guid>
      <description>I found a little tool where you describe a diagram in plain text, and it draws it. It likely uses a ChatGPT-to-Mermaid converter.</description>
      <category>bits</category>
      <category>diagrams</category>
      <category>chatgpt</category>
      <category>mermaid</category>
      <category>tools</category>
      <category>ai</category>
      <content:encoded><![CDATA[<p>I found a <a href="https://diagramify.agiliq.com">little tool</a> where you describe a
diagram in plain text, and it draws it. It likely uses a ChatGPT-to-Mermaid
converter.</p>]]></content:encoded>
    </item>
  </channel>
</rss>
