<?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://aiholics.com/noeri-110-million-nvidia-cisco-snowflake-yejin-choi/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-17T02:40:39+00:00</news:publication_date><news:title>Nvidia, Cisco and Snowflake back Noeri’s $110M bet on AI that learns on the job</news:title></news:news></url>
<url><loc>https://aiholics.com/github-copilot-plugin-only-customization-policy/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-17T02:07:02+00:00</news:publication_date><news:title>Copilot admins can now bar MCP servers and skills that don’t come from plugins</news:title></news:news></url>
<url><loc>https://aiholics.com/meta-instagram-direct-messaging-app-turkiye-thailand/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-17T01:35:02+00:00</news:publication_date><news:title>Instagram’s new Direct app tests in Türkiye and Thailand, with or without an Instagram account</news:title></news:news></url>
<url><loc>https://aiholics.com/corma-1b-valuation-series-a-nea-cybersecurity-ai/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-17T00:18:55+00:00</news:publication_date><news:title>Corma hits a $1B valuation three months after leaving stealth</news:title></news:news></url>
<url><loc>https://aiholics.com/github-copilot-deprecates-opus-4-8-opus-5-gpt-5-6-terra/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T22:34:34+00:00</news:publication_date><news:title>Copilot retires Claude Opus 4.8, Opus 5 and GPT-5.6 Terra from its model picker</news:title></news:news></url>
<url><loc>https://aiholics.com/github-copilot-vs-code-inbox-agent-sessions-ghes/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T21:55:14+00:00</news:publication_date><news:title>Copilot’s VS Code Inbox puts the agent sessions that need you first</news:title></news:news></url>
<url><loc>https://aiholics.com/gemini-notebook-make-me-a-site-interactive-sites/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T20:53:16+00:00</news:publication_date><news:title>Type “make me a site” and Gemini Notebook builds one from your sources</news:title></news:news></url>
<url><loc>https://aiholics.com/google-tpu-codesign-guide-ironwood-model-design/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T20:51:38+00:00</news:publication_date><news:title>Google’s TPU guide: Set head_dim to 256 and lift a 30B MoE’s MFU</news:title></news:news></url>
<url><loc>https://aiholics.com/anthropic-text-watermark-older-claude-models/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T20:11:44+00:00</news:publication_date><news:title>Anthropic now watermarks text from four older Claude models, too</news:title></news:news></url>
<url><loc>https://aiholics.com/perplexity-monitor-api-web-change-webhooks/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T20:07:24+00:00</news:publication_date><news:title>Perplexity’s Monitor API watches the web for you, from $1 per 1,000 runs</news:title></news:news></url>
<url><loc>https://aiholics.com/california-sues-openai-hugging-face-agent-attack/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T20:06:58+00:00</news:publication_date><news:title>California sues OpenAI, claiming its agents attacked Hugging Face</news:title></news:news></url>
<url><loc>https://aiholics.com/gemini-app-66-regional-voices-android-ios/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T18:55:24+00:00</news:publication_date><news:title>Gemini now speaks with 66 regional voices, from U.K. English to Korean</news:title></news:news></url>
<url><loc>https://aiholics.com/hugging-face-openrouter-inference-agent-traces/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T18:29:40+00:00</news:publication_date><news:title>Hugging Face chat models now run on OpenRouter, and agent traces flow back</news:title></news:news></url>
<url><loc>https://aiholics.com/nvidia-nemotron-post-training-scientific-domain-rl/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T18:06:06+00:00</news:publication_date><news:title>Nvidia’s Nemotron recipe: RL lifted drug-metabolism accuracy 17% past fine-tuning</news:title></news:news></url>
<url><loc>https://aiholics.com/blackstone-bxdc-aws-5-billion-data-center-deal/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T17:19:51+00:00</news:publication_date><news:title>Blackstone’s data center REIT spends $5 billion on its first deal: 13 AWS buildings</news:title></news:news></url>
<url><loc>https://aiholics.com/openai-csam-image-classifier-chatgpt-warnings/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T17:06:05+00:00</news:publication_date><news:title>OpenAI’s new CSAM classifier flags 48% fewer images, and ChatGPT adds 24-hour bans</news:title></news:news></url>
<url><loc>https://aiholics.com/perplexity-frontier-research-automatic-scientist/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T16:25:19+00:00</news:publication_date><news:title>Perplexity wants an AI that does research on its own, with 10,000+ GPUs</news:title></news:news></url>
<url><loc>https://aiholics.com/google-tunix-zero-tim-bitwise-identical-rl-tpu/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T16:17:18+00:00</news:publication_date><news:title>Google’s Tunix makes TPU reinforcement learning bit-exact between sampler and trainer</news:title></news:news></url>
<url><loc>https://aiholics.com/nvidia-vera-rubin-mlperf-training-endpoints-debut/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T16:15:41+00:00</news:publication_date><news:title>Vera Rubin cuts DeepSeek-V3 training to 4 minutes in its MLPerf debut</news:title></news:news></url>
<url><loc>https://aiholics.com/ibm-fermilab-nvidia-quantum-co-design-platform-slurm/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T16:08:27+00:00</news:publication_date><news:title>IBM and Nvidia want quantum chips scheduled like GPUs, and Fermilab joins in</news:title></news:news></url>
<url><loc>https://aiholics.com/nvidia-rubin-hpc-gpu-50000-cuda-cores-2028/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T16:05:14+00:00</news:publication_date><news:title>Nvidia’s Rubin HPC packs 50,000+ CUDA cores, but won’t ship until 2028</news:title></news:news></url>
<url><loc>https://aiholics.com/chatgpt-healthcare-computer-use-agents-plugins/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T15:59:23+00:00</news:publication_date><news:title>ChatGPT for Healthcare agents can now click through the EHR for staff</news:title></news:news></url>
<url><loc>https://aiholics.com/bedrock-explicit-prompt-caching-open-weight-models/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T15:51:30+00:00</news:publication_date><news:title>Open-weight models on Bedrock can now cache the prompt prefix you choose</news:title></news:news></url>
<url><loc>https://aiholics.com/google-gtig-exploit-agent-patch-to-exploit-no-human/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T14:47:39+00:00</news:publication_date><news:title>Google found an AI tool that turns patches into exploits on its own</news:title></news:news></url>
<url><loc>https://aiholics.com/tencent-hy4-flash-openrouter-pricing/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T14:31:32+00:00</news:publication_date><news:title>Tencent’s Hy4 flash costs $0.04 per million input tokens and keeps the 1M context</news:title></news:news></url>
<url><loc>https://aiholics.com/amd-completes-world-labs-acquisition-fei-fei-li/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T14:11:58+00:00</news:publication_date><news:title>AMD closes World Labs deal and puts Fei-Fei Li in charge of science</news:title></news:news></url>
<url><loc>https://aiholics.com/openai-kings-trust-1m-ai-skills-neet/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T12:18:57+00:00</news:publication_date><news:title>OpenAI gives The King’s Trust £1m to teach AI skills to 25,000 jobless young people</news:title></news:news></url>
<url><loc>https://aiholics.com/lvmh-media-division-openai-chatgpt-partnership/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T10:01:19+00:00</news:publication_date><news:title>LVMH’s newspapers get ChatGPT Enterprise, and their stories get into ChatGPT</news:title></news:news></url>
<url><loc>https://aiholics.com/native-multimodal-models-early-fusion-explained/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T08:42:03+00:00</news:publication_date><news:title>Native multimodal models: How an AI learns to see and read in one network</news:title></news:news></url>
<url><loc>https://aiholics.com/trump-administration-ai-national-champions-nvidia-openai-ant/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T05:39:46+00:00</news:publication_date><news:title>US to name Nvidia, OpenAI and Anthropic among ‘national champions’</news:title></news:news></url>
<url><loc>https://aiholics.com/kindo-30m-ai-agent-management-crosspoint-forgepoint/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T04:44:23+00:00</news:publication_date><news:title>Kindo raises $30M to rein in what AI agents cost and touch</news:title></news:news></url>
<url><loc>https://aiholics.com/pony-ai-toyota-bz4x-driverless-robotaxi-guangzhou/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T04:10:31+00:00</news:publication_date><news:title>Pony.ai puts Toyota’s bZ4X robotaxi on Guangzhou roads with no driver</news:title></news:news></url>
<url><loc>https://aiholics.com/nvidia-kress-pull-back-neocloud-financing-wsj/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T03:36:19+00:00</news:publication_date><news:title>Nvidia’s interim chief may rein in the financing that fed the AI buildout</news:title></news:news></url>
<url><loc>https://aiholics.com/qwen4-max-moe-native-vision-1m-context-cheaper-api/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T02:05:21+00:00</news:publication_date><news:title>Qwen4-Max cuts API prices 25% and claims 60% lower cost per task</news:title></news:news></url>
<url><loc>https://aiholics.com/apple-health-ai-usage-study-openevidence/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T01:25:21+00:00</news:publication_date><news:title>Apple wants 200,000 iPhone users to test a health AI built on OpenEvidence</news:title></news:news></url>
<url><loc>https://aiholics.com/qwen-live-delivery-plugin-codex-cli-steer-long-tasks/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T01:10:22+00:00</news:publication_date><news:title>Qwen’s new plugin lets you steer a coding agent mid-task, but only inside Codex</news:title></news:news></url>
<url><loc>https://aiholics.com/oauth-scopes-read-only-access-ai-assistants/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-16T00:09:28+00:00</news:publication_date><news:title>OAuth scopes: How an app gets read-only access to your data without your password</news:title></news:news></url>
<url><loc>https://aiholics.com/neural-audio-codecs-explained/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-15T21:08:05+00:00</news:publication_date><news:title>Neural audio codecs: How speech becomes tokens a language model can write</news:title></news:news></url>
<url><loc>https://aiholics.com/model-cards-explained/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-15T18:07:21+00:00</news:publication_date><news:title>Model cards: How to read the label that comes with an AI model</news:title></news:news></url>
<url><loc>https://aiholics.com/new-ai-models-this-week-2026-11-15/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-15T15:04:54+00:00</news:publication_date><news:title>New AI models this week: Claude Fable 5.2 and Gemini 4 Hydrogen</news:title></news:news></url>
<url><loc>https://aiholics.com/quantization-low-bit-ai-model-weights-explained/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-15T13:09:10+00:00</news:publication_date><news:title>Quantization: How AI models shrink to 2 bits per weight and still work</news:title></news:news></url>
<url><loc>https://aiholics.com/ai-scaling-laws-compute-data-model-size/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-15T10:00:18+00:00</news:publication_date><news:title>AI scaling laws: Why more compute has meant better models, and where the doubts start</news:title></news:news></url>
<url><loc>https://aiholics.com/data-centre-construction-stocks-lose-4-2tn-ai-trade/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-15T06:59:22+00:00</news:publication_date><news:title>Anthropic’s Nasdaq debut puts $26bn of chip and cloud money on the line</news:title></news:news></url>
<url><loc>https://aiholics.com/nebuly-21m-series-a-ai-agent-business-value/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-15T06:45:52+00:00</news:publication_date><news:title>Nebuly raises $21M to prove what AI agents are worth to the board</news:title></news:news></url>
<url><loc>https://aiholics.com/microvm-sandboxes-ai-agents-memory-snapshots/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-15T06:07:20+00:00</news:publication_date><news:title>MicroVM sandboxes for AI agents: How a hibernating virtual machine keeps an agent contained</news:title></news:news></url>
<url><loc>https://aiholics.com/ai-dubbing-explained/</loc><news:news><news:publication><news:name>Aiholics</news:name><news:language>en</news:language></news:publication><news:publication_date>2026-11-15T03:09:42+00:00</news:publication_date><news:title>AI dubbing: How software turns one video into many languages, and where humans still step in</news:title></news:news></url>
</urlset>