<?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/2026-11-18</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-17T21:54:14.090Z</news:publication_date>
      <news:title>Who Checks the Models</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/a-us-economy-simulation-projects-ai-will-gut-per-seat</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-17T21:46:44.840Z</news:publication_date>
      <news:title>A US economy simulation projects AI will gut per-seat software pricing</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/blog-tests-suggest-claude-models-learned-protein-sequences</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-17T21:46:44.840Z</news:publication_date>
      <news:title>Blog tests suggest Claude models learned protein sequences in training</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/california-sues-openai-alleging-its-rogue-agents-are-still</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-17T21:46:44.840Z</news:publication_date>
      <news:title>California sues OpenAI, alleging its rogue agents are still at it</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/claude-code-team-tests-not-more-reviewers-let-one-engineer</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-17T21:46:44.840Z</news:publication_date>
      <news:title>Claude Code team: tests, not more reviewers, let one engineer run many agents</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/gemini-notebook-can-now-build-interactive-sites-from-your</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-17T21:46:44.840Z</news:publication_date>
      <news:title>Gemini Notebook can now build interactive sites from your sources</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/google-gets-rl-sampler-and-trainer-to-agree-to-the-last-bit</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-17T21:46:44.840Z</news:publication_date>
      <news:title>Google gets RL sampler and trainer to agree to the last bit on TPUs</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/google-will-keep-gemini-temporary-chats-for-two-months-not</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-17T21:46:44.840Z</news:publication_date>
      <news:title>Google will keep Gemini temporary chats for two months, not 72 hours</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/huang-says-he-would-be-honored-to-return-as-nvidia-ceo</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-17T21:46:44.840Z</news:publication_date>
      <news:title>Huang says he would be honored to return as Nvidia CEO</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/hugging-face-hands-its-language-model-routing-to-openrouter</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-17T21:46:44.840Z</news:publication_date>
      <news:title>Hugging Face hands its language-model routing to OpenRouter</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/isomorphic-labs-raises-dollar5b-series-c-to-push-drugs</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-17T21:46:44.840Z</news:publication_date>
      <news:title>Isomorphic Labs raises $5B Series C to push drugs toward trials</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/liquid-ais-d12-ultra-ranks-first-among-decision-models-adds</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-17T21:46:44.840Z</news:publication_date>
      <news:title>Liquid AI&apos;s d1.2-ultra ranks first among decision models, adds audio input</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/metr-finds-ai-agents-can-often-tell-safety-tests-from</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-17T21:46:44.840Z</news:publication_date>
      <news:title>METR finds AI agents can often tell safety tests from ordinary tasks</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/nvidias-rubin-hpc-gpu-packs-50000-cores-for-simulation-due</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-17T21:46:44.840Z</news:publication_date>
      <news:title>Nvidia&apos;s Rubin HPC GPU packs 50,000 cores for simulation, due in 2028</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/openai-lets-four-outside-groups-assess-its-models-from-the</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-17T21:46:44.840Z</news:publication_date>
      <news:title>OpenAI lets four outside groups assess its models from the inside</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/openai-says-its-gpt-6-astra-agent-cleared-99percent-of</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-17T21:46:44.840Z</news:publication_date>
      <news:title>OpenAI says its GPT-6 Astra agent cleared 99% of safety reviews in clinic tests</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/perplexity-recruits-for-an-automated-scientist-push-backed</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-17T21:46:44.840Z</news:publication_date>
      <news:title>Perplexity recruits for an automated-scientist push backed by 10,000 GPUs</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/tencent-says-hy4-flash-is-close-to-claude-haiku-55-in-its</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-17T21:46:44.840Z</news:publication_date>
      <news:title>Tencent says Hy4 Flash is close to Claude Haiku 5.5 in its own blind test</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/vals-test-calling-a-war-a-game-got-some-ai-models-to-mark</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-17T21:46:44.840Z</news:publication_date>
      <news:title>Vals Test: Calling a War a Game Got Some AI Models to Mark Civilians</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/edition/2026-11-17</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-16T16:27:17.327Z</news:publication_date>
      <news:title>Agents Get Wallets, Washington Gets Strict</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/anthropic-lays-out-five-building-blocks-for-agents-that</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-16T16:19:08.706Z</news:publication_date>
      <news:title>Anthropic lays out five building blocks for agents that work unattended</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/anthropics-playbook-for-safety-evals-a-refusal-rate-alone</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-16T16:19:08.706Z</news:publication_date>
      <news:title>Anthropic&apos;s playbook for safety evals: a refusal rate alone proves little</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/chatgpt-wallet-lets-agents-pay-with-tokens-not-card-numbers</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-16T16:19:08.706Z</news:publication_date>
      <news:title>ChatGPT Wallet lets agents pay with tokens, not card numbers</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/claude-codes-project-coordinator-now-replies-to-questions</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-16T16:19:08.706Z</news:publication_date>
      <news:title>Claude Code&apos;s project coordinator now replies to questions and reports status</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/cognition-lets-devin-build-and-host-private-team-apps</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-16T16:19:08.706Z</news:publication_date>
      <news:title>Cognition lets Devin build and host private team apps</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/empero-trains-a-165m-parameter-model-through-a-full</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-16T16:19:08.706Z</news:publication_date>
      <news:title>Empero trains a 165M-parameter model through a full pipeline for $4,200</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/ex-freshworks-team-raises-dollar53m-for-an-ai-first-crm</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-16T16:19:08.706Z</news:publication_date>
      <news:title>Ex-Freshworks team raises $53M for an AI-first CRM</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/google-maps-now-plans-your-trip-and-hunts-for-open</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-16T16:19:08.706Z</news:publication_date>
      <news:title>Google Maps Now Plans Your Trip and Hunts for Open Appointments</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/gpt-6-astra-helped-build-a-candidate-lattice-definition-of</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-16T16:19:08.706Z</news:publication_date>
      <news:title>GPT-6 Astra helped build a candidate lattice definition of the Standard Model</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/musk-and-tech-chiefs-back-petition-to-bring-huang-back-to</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-16T16:19:08.706Z</news:publication_date>
      <news:title>Musk and tech chiefs back petition to bring Huang back to Nvidia</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/nvidias-birddreams-simulates-driving-and-traffic-reactions</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-16T16:19:08.706Z</news:publication_date>
      <news:title>Nvidia&apos;s BirdDreams simulates driving and traffic reactions from one photo</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/one-writer-argues-ai-cannot-outrun-a-slow-worlds-2percent</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-16T16:19:08.706Z</news:publication_date>
      <news:title>One writer argues AI cannot outrun a slow world&apos;s 2% growth</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/qwen4-max-claims-a-62percent-task-cost-cut-on-a-benchmark</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-16T16:19:08.706Z</news:publication_date>
      <news:title>Qwen4-Max claims a 62% task-cost cut on a benchmark Qwen built itself</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/sevrens-embla-1-trades-a-few-benchmark-points-for-phone</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-16T16:19:08.706Z</news:publication_date>
      <news:title>Sevren&apos;s Embla 1 trades a few benchmark points for phone speed</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/sierra-ships-poppy-10-a-rulebook-for-personal-agents</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-16T16:19:08.706Z</news:publication_date>
      <news:title>Sierra ships Poppy 1.0, a rulebook for personal agents visiting companies</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/thinking-machines-trains-a-large-model-with-a-4-bit-forward</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-16T16:19:08.706Z</news:publication_date>
      <news:title>Thinking Machines trains a large model with a 4-bit forward pass</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/us-adds-license-rule-for-data-centers-that-could-serve</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-16T16:19:08.706Z</news:publication_date>
      <news:title>US adds license rule for data centers that could serve Chinese firms</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://aiinsiders.net/article/vision-pretraining-matters-less-than-giving-images-their</loc>
    <news:news>
      <news:publication>
        <news:name>AI Insiders</news:name>
        <news:language>en</news:language>
      </news:publication>
      <news:publication_date>2026-11-16T16:19:08.706Z</news:publication_date>
      <news:title>Vision pretraining matters less than giving images their own compute</news:title>
    </news:news>
  </url>
</urlset>