<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
	<channel>
		<title>Phos on AIBussin — AI applications, systems and books</title>
		<link>https://aibussin.com/tags/phos/</link>
		<description>Recent content in Phos on AIBussin — AI applications, systems and books</description>
		<generator>Hugo</generator>
		<language>en-US</language>
		
		
		
		
			<lastBuildDate>Thu, 09 Oct 2025 00:30:36 +0100</lastBuildDate>
		
			<atom:link href="https://aibussin.com/tags/phos/index.xml" rel="self" type="application/rss+xml" />
			<item>
				<title>🔦 Phōs: Visualizing How AI Learns and How to Build It Yourself</title>
				<link>https://aibussin.com/post/phos/</link>
				<pubDate>Thu, 09 Oct 2025 00:30:36 +0100</pubDate>
				<guid>https://aibussin.com/post/phos/</guid>
				<description>&lt;blockquote&gt;&#xA;&lt;p&gt;“The eye sees only what the mind is prepared to comprehend.” &lt;em&gt;Henri Bergson&lt;/em&gt;&lt;/p&gt;&#xA;&lt;/blockquote&gt;&#xA;&lt;h2 id=&#34;-we-finally-see-learning&#34;&gt;🔍 We Finally See Learning&lt;/h2&gt;&#xA;&lt;p&gt;For decades, we’ve measured artificial intelligence with numbers loss curves, accuracy scores, reward signals.&lt;br&gt;&#xA;We’ve plotted progress, tuned hyperparameters, celebrated benchmarks.&lt;/p&gt;&#xA;&lt;p&gt;But we’ve never actually &lt;em&gt;seen&lt;/em&gt; learning happen.&lt;/p&gt;&#xA;&lt;p&gt;Not really.&lt;/p&gt;&#xA;&lt;p&gt;Sure, we’ve visualized attention maps or gradient flows but those are snapshots, proxies, not processes.&lt;/p&gt;&#xA;&lt;p&gt;What if we could watch understanding emerge not as a number going up, but as a pattern stabilizing across time?&lt;br&gt;&#xA;What if reasoning itself left a visible trace?&lt;/p&gt;</description>
			</item>
	</channel>
</rss>
