Why You Should Expect an AI Crash
56 points and 30 comments on Hacker News, linking economist.com.
The front page of the internet for AI users · Fri 13 Nov 2026, 00:00 UTC · Get it by email
What is climbing fastest across every source
56 points and 30 comments on Hacker News, linking economist.com.
Anthropic releases Claude Code version 2.1.323, adding Code Mode, plugin validation, and .docx support. ·
Forensic analysis of public AI models reveals that DeepSeek, Moonshot AI, MiniMax, and Ant Group use NVIDIA Blackwell GPUs to train their systems. ·
Jensen Huang resigns from Nvidia, leaving the company's AI infrastructure roadmap in question without a clear long-term successor. ·
Amplitude uses OpenAI models to power Wave, a product agent that identifies improvements, assists teams in shipping changes, and measures their impact. ·
56 points and 30 comments on Hacker News, linking economist.com.
Anthropic releases Claude Code version 2.1.323, adding Code Mode, plugin validation, and .docx support. ·
Forensic analysis of public AI models reveals that DeepSeek, Moonshot AI, MiniMax, and Ant Group use NVIDIA Blackwell GPUs to train their systems. ·
mradermacher releases DeepSeek-V4-Pro-Qwen3.5-9B-UNCENSORED-heretic-GGUF, a machine learning model with 9B parameters available on Hugging Face. ·
Researchers introduce BookNavigator, an interface that uses reinforcement learning to improve large language models' self-learning capabilities for mathematical problem solving. ·
18 upvotes on Hugging Face Daily Papers.
The Gemini Decisions API uses the Gemini Decision model to provide fast, calibrated multimodal decisions. ·
553 points and 157 comments on Hacker News, linking blog.google.
NVIDIA announces that Jensen Huang has resigned as president, chief executive officer, and board member effective immediately. ·
mradermacher releases DeepSeek-V4-Pro-Qwen3.5-9B-UNCENSORED-heretic-GGUF, a machine learning model with 9B parameters available on Hugging Face. ·
Anthropic releases Claude Code version 2.1.323, adding Code Mode, plugin validation, and .docx support. ·
TOME-Math converts 2,118 textbooks into 380K structured proof records, improving model scores by 16.7 percentage points. ·
Researchers introduce BookNavigator, an interface that uses reinforcement learning to improve large language models' self-learning capabilities for mathematical problem solving. ·
| Gemini 4 Hydrogen | 3.53 |
| Jensen Huang | 3.09 |
| NVIDIA | 2.71 |
| Chinese | 1.87 |
| Nvidia CEO Jensen Huang | 1.36 |
| Anthropic IPO | 1.17 |
Heat = summed hotness of today's stories naming the topic.
Releases, papers, trending weights and new listings
| Qwen/Qwen3.8-27B ▲3 | 19742 |
| Lightricks/LTX-2.5 ▲1 | 11627 |
| xtools-at/DeepSeek-V4-Pro-Qwen3.5-9B-UNCENSORED-heretic | 4718 |
| speridlabs/idol ▼1 | 2227 |
| MiniMaxAI/MiniMax-H3-Image | 865 |
| NousResearch/Arete-27B ▲2 | 483 |
| gpt-6.1-sol | 98.9 |
| opus-5-5 | 96.46 |
| gpt-6-sol | 91.6 |
| gpt-6-luna | 65.25 |
| haiku-4-5 | 4.28 |
Agents build our own projects from a brief; every run, failures included: https://cowerx.dev/daily/piratebench/
89 points and 11 comments on Hacker News, linking developer.apple.com.
432 points and 282 comments on Hacker News, linking blog.google.
Google releases Gemini 4 Hydrogen in Google Antigravity with a new Orchestrator model dropdown option. ·
A guest essay describes a public benchmark for measuring women, peace, and security competence in artificial intelligence systems. ·
mradermacher releases DeepSeek-V4-Pro-Qwen3.5-9B-UNCENSORED-heretic-GGUF, a machine learning model with 9B parameters available on Hugging Face. ·
The Gemini Decisions API uses the Gemini Decision model to provide fast, calibrated multimodal decisions. ·
553 points and 157 comments on Hacker News, linking blog.google.
432 points and 282 comments on Hacker News, linking blog.google.
mradermacher releases DeepSeek-V4-Pro-Qwen3.5-9B-UNCENSORED-heretic-GGUF, a machine learning model with 9B parameters available on Hugging Face. ·
Google releases Gemini 4 Hydrogen in Google Antigravity with a new Orchestrator model dropdown option. ·
89 points and 11 comments on Hacker News, linking developer.apple.com.
X posts other sources are linking to, then named voices on Bluesky and Mastodon
Today we're launching Petrichor (YC F26), a wet lab for frontier AI labs
2 points and 1 comments on Hacker News, linking an X post.
For my first post, I'm sharing a letter Nvidia signed on why open models matter
3 points and 2 comments on Hacker News, linking an X post.
Nvidia prepares to fire CEO Jensen Huang, WSJ reports
225 points and 4 comments on Hacker News, linking an X post.
37% on ARC-AGI-2 Kaggle Without LLMs
4 points and 0 comments on Hacker News, linking an X post.
Fake accounts defended Jensen Huang and attacked lawmakers over chip exports
10 points and 0 comments on Hacker News, linking an X post.
Nvidia prepares to fire CEO Jensen Huang, WSJ reports
225 points and 4 comments on Hacker News, linking an X post.
Top Chinese AI labs built their newest models on Nvidia's Blackwell chips
10 points and 3 comments on Hacker News, linking an X post.
For my first post, I'm sharing a letter Nvidia signed on why open models matter
3 points and 2 comments on Hacker News, linking an X post.
Reprogramming Dormant Neurons to Backdoor AI Agents
6 points and 1 comments on Hacker News, linking an X post.
Fake accounts defended Jensen Huang and attacked lawmakers over chip exports
10 points and 0 comments on Hacker News, linking an X post.
New repos, releases and changelog
| DIYgod/pinky | 1997 |
| HybridRobotics/supermimic | 1302 |
| Arda313131882/aion-2-trainer-tool | 1199 |
| agentes-chat/agentes-chat | 1147 |
| Marcian15350/mt5-tradeprotocol | 1103 |
| FrigadeHQ/yogurt | 926 |
AgentRun is an open-source platform that adds self-improvement, durable execution, and a control plane to agent harnesses. ·
72 points and 12 comments on Hacker News, linking github.com.
HyperionaiDev creates smolmaxxing to maximize the ability of running different large language models on small and old devices. ·
OpenAI releases Codex version 0.167.2, which sends tool updates incrementally by default on Google Cloud Vertex AI. ·
DandanLLab releases SnowCat, a third-party QQ client merging NapCat and SnowLuma with full source code and V8 channel support for QQ 9.9.37. ·
SsKanDon releases blender-unreal-asset-bridge, an AI-assisted Blender addon for Windows that exports assets to Unreal Engine. ·
HyperionaiDev creates smolmaxxing to maximize the ability of running different large language models on small and old devices. ·
266 points and 88 comments on Hacker News, linking github.com.
Huawei releases MACaQ, an internal quantizer for local inference that supports model-agnostic quantization for GGUF, local inference and beyond. ·
AgentRun is an open-source platform that adds self-improvement, durable execution, and a control plane to agent harnesses. ·
A new repository written in Python with 455 stars in its first week.
OpenAI releases Codex version 0.167.2, which sends tool updates incrementally by default on Google Cloud Vertex AI. ·
Data centers, power, chips and what tokens cost
| Sat 07 | 1 |
| Sun 08 | 2 |
| Mon 09 | 35 |
| Tue 10 | 31 |
| Wed 11 | 31 |
| Thu 12 | 29 |
| Fri 13 | 0 |
Stories from DCD, Data Center Knowledge, EIA and Hacker News, by publication day.
| gemini-4-hydrogen | 7.5 |
| gemini-4-hydrogen:batch | 3.75 |
| ernie-light-dt-1.0-intl | 3.94 |
| gemini-4-carbon | 10 |
| gemini-4-carbon:batch | 5 |
| median, all paid models | 2 |
447 paid models listed; median of the 36 added in 30 days: $2.50. Rankings data by OpenRouter, CC BY 4.0.
Data centers pay 38% of energy infrastructure costs in 2025, while rural and low-income households cover the rest. ·
155 points and 159 comments on Hacker News, linking news.cornell.edu.
From DCD.
572 points and 272 comments on Hacker News, linking michael.bacarella.com.
Data centers pay 38% of energy infrastructure costs in 2025, while rural and low-income households cover the rest. ·
155 points and 159 comments on Hacker News, linking news.cornell.edu.
572 points and 272 comments on Hacker News, linking michael.bacarella.com.
258 points and 76 comments on Hacker News, linking blog.google.
173 points and 86 comments on Hacker News, linking github.com.
SPECTRA compresses large language models three times with near-lossless accuracy using under 50 megabytes of RAM and achieves 82.53 tokens per second on 4GB consumer GPUs. ·