<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Ai-Tools on The Curtain Must Rise</title><link>https://thecurtainmustrise.com/tags/ai-tools/</link><description>Recent content in Ai-Tools on The Curtain Must Rise</description><generator>Hugo -- gohugo.io</generator><language>en-us</language><lastBuildDate>Thu, 25 Jun 2026 19:30:00 +1200</lastBuildDate><atom:link href="https://thecurtainmustrise.com/tags/ai-tools/index.xml" rel="self" type="application/rss+xml"/><item><title>Handwriting</title><link>https://thecurtainmustrise.com/post/handwriting/</link><pubDate>Thu, 25 Jun 2026 19:30:00 +1200</pubDate><guid>https://thecurtainmustrise.com/post/handwriting/</guid><description>&lt;img src="https://thecurtainmustrise.com/post/handwriting/header.svg" alt="Featured image of post Handwriting" /&gt;&lt;p&gt;I wanted the English notes on the grimmerie to be handwritten, but my early attempts at using AI didn’t work out so well. Most attempts just fell back to use a script font. The image generating AIs are mostly worse. That’s where you get the recognisable garbled text.&lt;/p&gt;
&lt;p&gt;I did stumble upon a site that came close, &lt;a class="link" href="https://calligrapher.ai/" target="_blank" rel="noopener"
 &gt;calligrapher.ai&lt;/a&gt;. I have no idea who runs this site, it’s basically entirely bare, you give it some text, adjust a few knobs, and it spits out an SVG. It doesn’t appear to be an LLM, it looks more like a small set of randomised letters, but the results are nice, certainly better than if I tried to write it myself.&lt;/p&gt;
&lt;hr&gt;
&lt;p&gt;&lt;strong&gt;Edit:&lt;/strong&gt; Ok it turns out it&amp;rsquo;s not randomised letters, but an RNN.&lt;sup id="fnref:1"&gt;&lt;a href="#fn:1" class="footnote-ref" role="doc-noteref"&gt;1&lt;/a&gt;&lt;/sup&gt; Essentially, it was trained on real handwriting samples and uses past pen strokes to be part of the input for the next stroke (that&amp;rsquo;s how the letters always connect appropriately). Hats off (and thank you) to whoever built this handy site.&lt;/p&gt;
&lt;div class="footnotes" role="doc-endnotes"&gt;
&lt;hr&gt;
&lt;ol&gt;
&lt;li id="fn:1"&gt;
&lt;p&gt;&lt;strong&gt;Recurrent Neural Network&lt;/strong&gt; — a type of neural network designed for sequential data. Unlike a regular neural network that processes each input independently, an RNN maintains a hidden &amp;ldquo;state&amp;rdquo; that gets updated at each step, so each output depends on both the current input and what came before.&amp;#160;&lt;a href="#fnref:1" class="footnote-backref" role="doc-backlink"&gt;&amp;#x21a9;&amp;#xfe0e;&lt;/a&gt;&lt;/p&gt;
&lt;/li&gt;
&lt;/ol&gt;
&lt;/div&gt;</description></item></channel></rss>