<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>PAI on Tai&#39;s Life</title>
    <link>https://taislife.work/tags/pai/</link>
    <description>Recent content in PAI on Tai&#39;s Life</description>
    <generator>Hugo</generator>
    <language>en</language>
    <lastBuildDate>Tue, 25 Nov 2025 09:30:00 +0800</lastBuildDate>
    <atom:link href="https://taislife.work/tags/pai/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>【Deep Analysis】Gemini 3 Counterattack (Part 2): From Digital Brain to Physical AI (PAI) — The Jedi Counterattack of OpenAI and Nvidia</title>
      <link>https://taislife.work/blogs/google-vs-nvidia-proxy-war-with-pai/</link>
      <pubDate>Tue, 25 Nov 2025 09:30:00 +0800</pubDate>
      <guid>https://taislife.work/blogs/google-vs-nvidia-proxy-war-with-pai/</guid>
      <description>&lt;h2 id=&#34;prologue-breaking-out-of-the-brain-in-a-vat-dilemma&#34;&gt;Prologue: Breaking Out of the &amp;ldquo;Brain in a Vat&amp;rdquo; Dilemma&lt;/h2&gt;&#xA;&lt;p&gt;In the previous article, we analyzed how Google used the vertical integration of TPUs and the cost advantage of Gemini 3 to besiege OpenAI on the &amp;ldquo;cloud pure software&amp;rdquo; battlefield. If the future of AI is merely generating text, images, or writing code, then OpenAI will be extremely passive in this price cutthroat war.&lt;/p&gt;&#xA;&lt;p&gt;However, another revolution in Silicon Valley is quietly happening. Current LLMs (Large Language Models) are essentially &lt;strong&gt;&amp;ldquo;Brains in a Vat&amp;rdquo;&lt;/strong&gt;. They are trapped in server rooms, talking to the world through APIs, but are always subject to the terminal high walls built by Google (Android) and Apple (iOS).&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
