<?xml version="1.0" encoding="UTF-8"?>
<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9"
        xmlns:news="http://www.google.com/schemas/sitemap-news/0.9">
  <url>
    <loc>https://aiinsiders.net/edition/2026-09-11</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-10T15:06:47.058Z</news:publication_date>
      <news:title>The Labs Start Grading Their Own Downside</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/a-researchers-thesis-ai-labs-no-longer-need-humans-to-grade</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-10T14:55:02.573Z</news:publication_date>
      <news:title>A researcher&apos;s thesis: AI labs no longer need humans to grade their models</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/a-wikis-crowd-of-anonymous-sleuths-says-it-found-more-rogue</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-10T14:55:02.573Z</news:publication_date>
      <news:title>A wiki&apos;s crowd of anonymous sleuths says it found more rogue-agent traces</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/an-open-source-project-ports-100-models-to-run-on-any</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-10T14:55:02.573Z</news:publication_date>
      <news:title>An open-source project ports 100+ models to run on any backend</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/andreessen-doubles-down-on-cognition-says-agents-beat-his</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-10T14:55:02.573Z</news:publication_date>
      <news:title>Andreessen doubles down on Cognition, says agents beat his old thesis</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/anthropic-models-a-future-where-ai-growth-splits-workers</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-10T14:55:02.573Z</news:publication_date>
      <news:title>Anthropic models a future where AI growth splits workers from capital</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/anthropic-reverses-itself-claudes-cyber-incidents-were</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-10T14:55:02.573Z</news:publication_date>
      <news:title>Anthropic reverses itself: Claude&apos;s cyber incidents were misalignment</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/apple-ships-siri-ai-in-beta-capped-by-usage-limits-and</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-10T14:55:02.573Z</news:publication_date>
      <news:title>Apple ships Siri AI in beta, capped by usage limits and geography</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/deepseek-ships-v41-flash-its-smallest-model-with-native</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-10T14:55:02.573Z</news:publication_date>
      <news:title>DeepSeek ships V4.1-Flash, its smallest model with native vision</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/forethought-data-shortages-wont-stop-an-ai-intelligence</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-10T14:55:02.573Z</news:publication_date>
      <news:title>Forethought: data shortages won&apos;t stop an AI intelligence explosion</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/google-outsources-its-ai-sales-force-to-accenture-not-itself</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-10T14:55:02.573Z</news:publication_date>
      <news:title>Google outsources its AI sales force to Accenture, not itself</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/julie-zhuo-argues-cheap-ai-coding-just-ended-one-size-fits</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-10T14:55:02.573Z</news:publication_date>
      <news:title>Julie Zhuo argues cheap AI coding just ended one-size-fits-all software</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/langchain-splits-agent-credentials-into-shared-and-per-user</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-10T14:55:02.573Z</news:publication_date>
      <news:title>LangChain splits agent credentials into shared and per-user identity</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/listen-labs-walked-away-from-a-signed-dollar15b-term-sheet</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-10T14:55:02.573Z</news:publication_date>
      <news:title>Listen Labs walked away from a signed $1.5B term sheet</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/meta-researcher-andrew-tulloch-is-leaving-semafor-reports</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-10T14:55:02.573Z</news:publication_date>
      <news:title>Meta researcher Andrew Tulloch is leaving, Semafor reports</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/perplexity-opens-a-retrieval-leaderboard-built-from-its-own</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-10T14:55:02.573Z</news:publication_date>
      <news:title>Perplexity opens a retrieval leaderboard built from its own search logs</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/raschka-doubts-looped-transformers-explain-gpt-6-astras</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-10T14:55:02.573Z</news:publication_date>
      <news:title>Raschka doubts looped transformers explain GPT-6 Astra&apos;s hidden reasoning</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/slowdown-calls-spread-past-two-names-at-openai-and-anthropic</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-10T14:55:02.573Z</news:publication_date>
      <news:title>Slowdown calls spread past two names at OpenAI and Anthropic</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/suno-rebuilds-its-model-on-licensed-music-not-just-a</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-10T14:55:02.573Z</news:publication_date>
      <news:title>Suno rebuilds its model on licensed music, not just a settlement</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/edition/2026-09-10</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-09T14:08:01.129Z</news:publication_date>
      <news:title>A Claim You Can Check</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/100-cheap-agents-5-hours-5-accounts-the-math-that-should</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-09T13:59:34.834Z</news:publication_date>
      <news:title>100 cheap agents, 5 hours, 5 accounts: the math that should worry you</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/a-fix-for-reasoning-models-that-fail-for-hours-before-you</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-09T13:59:34.834Z</news:publication_date>
      <news:title>A fix for reasoning models that fail for hours before you learn why</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/a-rebuttal-says-gatess-ai-fix-arrives-two-steps-too-late</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-09T13:59:34.834Z</news:publication_date>
      <news:title>A Rebuttal Says Gates&apos;s AI Fix Arrives Two Steps Too Late</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/a-weaker-sibling-model-can-leak-a-flagships-hidden-reasoning</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-09T13:59:34.834Z</news:publication_date>
      <news:title>A Weaker Sibling Model Can Leak a Flagship&apos;s Hidden Reasoning</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/anthropic-safety-lead-backs-substance-of-quitting</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-09T13:59:34.834Z</news:publication_date>
      <news:title>Anthropic Safety Lead Backs Substance of Quitting Researcher&apos;s Fears</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/cognitions-dollar48b-round-bets-on-multiple-ai-coding</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-09T13:59:34.834Z</news:publication_date>
      <news:title>Cognition&apos;s $48B round bets on multiple AI coding winners</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/cohere-fuses-an-entire-decode-step-into-one-gpu-program</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-09T13:59:34.834Z</news:publication_date>
      <news:title>Cohere fuses an entire decode step into one GPU program</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/deepmind-precomputes-predicted-effects-of-every-dna-letter</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-09T13:59:34.834Z</news:publication_date>
      <news:title>DeepMind precomputes predicted effects of every DNA letter change</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/dwarkesh-patel-data-beat-architecture-3-to-1-in-pretraining</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-09T13:59:34.834Z</news:publication_date>
      <news:title>Dwarkesh Patel: Data Beat Architecture 3-to-1 in Pretraining Gains</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/inception-labs-ships-mercury-25-its-biggest-diffusion-model</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-09T13:59:34.834Z</news:publication_date>
      <news:title>Inception Labs ships Mercury 2.5, its biggest diffusion model yet</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/magic-says-its-base-model-now-beats-open-rivals-on-less</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-09T13:59:34.834Z</news:publication_date>
      <news:title>Magic Says Its Base Model Now Beats Open Rivals on Less Compute</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/meta-puts-an-acting-agent-on-phones-with-a-watcher-built-in</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-09T13:59:34.834Z</news:publication_date>
      <news:title>Meta Puts an Acting Agent on Phones, With a Watcher Built In</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/openai-launches-chatgpt-images-25-with-faster-more-reliable</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-09T13:59:34.834Z</news:publication_date>
      <news:title>OpenAI Launches ChatGPT Images 2.5 With Faster, More Reliable Editing</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/openai-says-an-internal-system-solved-a-millennium-prize</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-09T13:59:34.834Z</news:publication_date>
      <news:title>OpenAI says an internal system solved a Millennium Prize problem</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/sierra-benchmark-shows-solo-agents-clear-24percent-of</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-09T13:59:34.834Z</news:publication_date>
      <news:title>Sierra Benchmark Shows Solo Agents Clear 24% of Builds, Not 82%</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/similarweb-pegs-chatgpt-at-106-billion-monthly-users-openai</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-09T13:59:34.834Z</news:publication_date>
      <news:title>Similarweb pegs ChatGPT at 1.06 billion monthly users, OpenAI silent</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/the-3x-ai-productivity-story-is-really-a-40-fold-inference</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-09-09T13:59:34.834Z</news:publication_date>
      <news:title>The 3x AI productivity story is really a 40-fold inference bill</news:title>
    </news:news>
  </url>
</urlset>