<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Aaronson-Ambainis on BriefTechNews.com</title><link>https://brief-tech-news.com/tags/aaronson-ambainis/</link><description>Recent content in Aaronson-Ambainis on BriefTechNews.com</description><generator>Hugo</generator><language>en</language><copyright>Copyright &amp;copy;2021 &lt;br&gt; Designed &amp;amp; Developed by [Gethugothemes](https://gethugothemes.com/)</copyright><lastBuildDate>Mon, 09 Nov 2026 15:16:51 +0000</lastBuildDate><atom:link href="https://brief-tech-news.com/tags/aaronson-ambainis/index.xml" rel="self" type="application/rss+xml"/><item><title>AI Now Accelerating Its Own Research: METR Says Anthropic Crossed the Threshold</title><link>https://brief-tech-news.com/blogs/2026-11-09-ai/</link><pubDate>Mon, 09 Nov 2026 15:16:51 +0000</pubDate><guid>https://brief-tech-news.com/blogs/2026-11-09-ai/</guid><description>A METR report finds Anthropic&amp;rsquo;s capability progress roughly doubled year-over-year, with most of the acceleration attributed to AI automating AI R&amp;amp;D - and the company&amp;rsquo;s own Responsible Scaling Policy threshold may already be crossed. OpenAI raises $11B at a $1.25T valuation, short of its $30B target, while Anthropic&amp;rsquo;s IPO valuation expectations quietly drop. A Meta research engineer breaks down what RSI work actually looks like day-to-day, and a PSI system running a non-public Anthropic model proves the Aaronson-Ambainis conjecture, open since 2008. Plus: OpenRouter&amp;rsquo;s new Auto Code router, Anthropic&amp;rsquo;s agent memory design, and a white paper showing LLMs that redesign themselves mid-training.</description></item></channel></rss>