<?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>Human Oversight on Barnabé BALP</title><link>https://www.barnabebalp.fr/en/tags/human-oversight/</link><description>Recent content in Human Oversight on Barnabé BALP</description><generator>Hugo</generator><language>en</language><copyright>CC BY 4.0</copyright><lastBuildDate>Mon, 26 Oct 2026 04:14:40 +0000</lastBuildDate><atom:link href="https://www.barnabebalp.fr/en/tags/human-oversight/index.xml" rel="self" type="application/rss+xml"/><item><title>Recursion, Self-Training, and Control</title><link>https://www.barnabebalp.fr/en/posts/ai/recursion-self-training-and-control/</link><pubDate>Wed, 21 Oct 2026 18:40:31 +0200</pubDate><guid>https://www.barnabebalp.fr/en/posts/ai/recursion-self-training-and-control/</guid><description>From recursion to self-training, analysis of a scenario leading to loss of control. A plea for demonstrability in AI.</description></item></channel></rss>