<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>模型坍缩 on AI内参</title>
    <link>https://www.neican.ai/tags/%E6%A8%A1%E5%9E%8B%E5%9D%8D%E7%BC%A9/</link>
    <description>Recent content in 模型坍缩 on AI内参</description>
    <generator>Hugo</generator>
    <language>zh-cn</language>
    <lastBuildDate>Sun, 07 Jun 2026 07:40:04 +0800</lastBuildDate>
    <atom:link href="https://www.neican.ai/tags/%E6%A8%A1%E5%9E%8B%E5%9D%8D%E7%BC%A9/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>递归的代价：当AI开始自行构建文明的“脚手架”</title>
      <link>https://www.neican.ai/insights/article-20260607074004468-0/</link>
      <pubDate>Sun, 07 Jun 2026 07:40:04 +0800</pubDate>
      <guid>https://www.neican.ai/insights/article-20260607074004468-0/</guid>
      <description>递归式自我改进（RSI）正推动AI从人工辅助进化为自主研发主体，这一变革将大幅提升科研效率，但也带来了模型坍缩与人类主体性退化的潜在风险。未来的竞争将不再单纯取决于算力储备，而在于谁能构建起高效、鲁棒的自我训练算法闭环。</description>
    </item>
    <item>
      <title>AI早报 2026年06月07日</title>
      <link>https://www.neican.ai/morningnews/2026-06-07-ai-2026-06-07-/</link>
      <pubDate>Sun, 07 Jun 2026 07:31:21 +0800</pubDate>
      <guid>https://www.neican.ai/morningnews/2026-06-07-ai-2026-06-07-/</guid>
      <description>AI研发进入“递归自我改进”新阶段，效率飞跃与失控风险并存。</description>
    </item>
    <item>
      <title>AI替你写了全世界一半的文章，你的大脑开始“摸鱼”了吗？</title>
      <link>https://www.neican.ai/insights/article-20260525081003434-2/</link>
      <pubDate>Mon, 25 May 2026 08:10:03 +0800</pubDate>
      <guid>https://www.neican.ai/insights/article-20260525081003434-2/</guid>
      <description>互联网上AI生成的文章数量已经碾压人类，连词典都用“Slop（泔水）”来形容这股低质内容洪流。更可怕的是，这背后是一个“双输”飞轮：人类写得越少脑子越钝，AI吃不到新鲜“文本饲料”则会“模型坍缩”变智障。在等待超级智能拯救世界之前，人类可能先把自己“外包”成了废物。</description>
    </item>
    <item>
      <title>10-21日报|AI科学家“创世纪”，卡帕西却在耳语：慢下来，看清本质！</title>
      <link>https://www.neican.ai/newspaper/2025-10-21-10-21-ai-/</link>
      <pubDate>Tue, 21 Oct 2025 19:33:01 +0800</pubDate>
      <guid>https://www.neican.ai/newspaper/2025-10-21-10-21-ai-/</guid>
      <description>今天是2025年10月21日。当我们以为AI正在以指数级速度狂飙突进，麻省理工、哈佛、DeepSeek等巨头们用“AI科学家”的创世纪成果不断刷新着我们对智能极限的认知。然而，就在这场沸腾的“智能体盛宴”中，前OpenAI核心成员安德烈·卡帕西的尖锐洞察却如一盆冷水，浇透了狂热的预期，直指大模型固有的缺陷与未来十年AGI的漫长旅程。AI的狂飙是通向下一个文明飞跃，还是将自己困于“模型坍缩”的泥沼？</description>
    </item>
  </channel>
</rss>
