<?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>Generative AI on Blogs by Kush</title><link>https://blogsbykush.com/tags/generative-ai/</link><description>Recent content in Generative AI on Blogs by Kush</description><image><title>Blogs by Kush</title><url>https://blogsbykush.com/assets/images/og-default.png</url><link>https://blogsbykush.com/assets/images/og-default.png</link></image><generator>Hugo</generator><language>en</language><lastBuildDate>Sun, 04 Oct 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://blogsbykush.com/tags/generative-ai/feed.xml" rel="self" type="application/rss+xml"/><item><title>Digital Dhaba: how I built an AI-written tech newsletter that can't make up news</title><link>https://blogsbykush.com/build-log/digital-dhaba-ai-newsletter-that-cannot-make-up-news/</link><pubDate>Sun, 04 Oct 2026 00:00:00 +0000</pubDate><guid>https://blogsbykush.com/build-log/digital-dhaba-ai-newsletter-that-cannot-make-up-news/</guid><description>How I built Digital Dhaba, a weekly AI and tech newsletter: code gathers and ranks stories, Claude writes only from them, and every item links to its source.</description></item><item><title>How LLMs Think</title><link>https://blogsbykush.com/concept-breakdown/how-llms-think/</link><pubDate>Sun, 04 Oct 2026 00:00:00 +0000</pubDate><guid>https://blogsbykush.com/concept-breakdown/how-llms-think/</guid><description>How do LLMs like ChatGPT think? A comic breakdown using Antakshari to explain next-token prediction, simply, for product people.</description></item><item><title>Training an AI for one skill quietly changes its other answers</title><link>https://blogsbykush.com/learning-notes/training-an-ai-for-one-skill-changes-its-other-answers/</link><pubDate>Sun, 04 Oct 2026 00:00:00 +0000</pubDate><guid>https://blogsbykush.com/learning-notes/training-an-ai-for-one-skill-changes-its-other-answers/</guid><description>A learning note on the paper &amp;#39;Post-Training Leaves Behavioral Shadows on Unrelated Decisions&amp;#39;: training an AI for one skill shifts its other answers too.</description></item></channel></rss>