Nvidia CEO Jensen Huang's letter to staff about affair
305 points and 307 comments on Hacker News, linking axios.com.
Top now
6 storiesWhat is climbing fastest across every source
GPT-6.2 Sol8.6xNvidia CEO Jensen Huang7.5xClaude Code6.5xGPUs8.7xMihail Eric5.2xMonaco5.2xStanford Teaches AI-Native Engineering5.2xChinese4.8xautotrust1.6x
Nvidia Board of Directors Is Investigating Jensen Huang
Nvidia's board investigates CEO Jensen Huang after recent reports indicate he was targeted by a foreign intelligence effort. ·
Scantist/GLM-5.3-SLIM-GGUF
Scantist releases GLM-5.3-SLIM-GGUF, a text generation model with 563 billion parameters available on Hugging Face. ·
How Stanford Teaches AI-Native Engineering — Mihail Eric, Monaco
Mihail Eric describes how Stanford teaches AI-native engineering in CS146S and at the Monaco engineering team. ·
[One Million AI Engineer Special] How Stanford Teaches AI-Native Engineering — Mihail Eric, Monaco
Stanford instructor Mihail Eric teaches AI-native engineering in CS146S by having students propose changes to open-source repositories. ·
Nvidia CEO Jensen Huang's letter to staff about affair
305 points and 307 comments on Hacker News, linking axios.com.
Nvidia Board of Directors Is Investigating Jensen Huang
Nvidia's board investigates CEO Jensen Huang after recent reports indicate he was targeted by a foreign intelligence effort. ·
Scantist/GLM-5.3-SLIM-GGUF
Scantist releases GLM-5.3-SLIM-GGUF, a text generation model with 563 billion parameters available on Hugging Face. ·
Outsideclofilter/aion-2-trainer-tool
Outsideclofilter/aion-2-trainer-tool provides a pre-configured environment for Aion 2, an AI-driven LLM prompt engineering resource, on local workstations in 2026. ·
elunboundfiremail/LiveWall---Live-Wallpaper-For-Desktop
A new repository with 311 stars in its first week.
Nvidia CEO Jensen Huang's letter to staff about affair
305 points and 307 comments on Hacker News, linking axios.com.
Goldman Sachs is pulling Claude Code from its bankers and traders
Goldman Sachs removes Claude Code from bankers and traders, returning the AI coding agents to engineers before the company takes Anthropic public. ·
Scantist/GLM-5.3-SLIM-GGUF
Scantist releases GLM-5.3-SLIM-GGUF, a text generation model with 563 billion parameters available on Hugging Face. ·
Outsideclofilter/aion-2-trainer-tool
Outsideclofilter/aion-2-trainer-tool provides a pre-configured environment for Aion 2, an AI-driven LLM prompt engineering resource, on local workstations in 2026. ·
elunboundfiremail/LiveWall---Live-Wallpaper-For-Desktop
A new repository with 311 stars in its first week.
moonshotai/Kimi-K3.1-Pro
Moonshotai releases Kimi-K3.1-Pro, an image-text-to-text transformer model featuring 2.8 trillion parameters available on Hugging Face. ·
Trending topics today: heat
Table
| GPT-6.2 Sol | 1.26 |
| Nvidia CEO Jensen Huang | 1.17 |
| Claude Code | 1.02 |
| GPUs | 0.87 |
| Mihail Eric | 0.52 |
| Monaco | 0.52 |
Heat = summed hotness of today's stories naming the topic.
Models
5 storiesReleases, papers, trending weights and new listings
Top 6 trending on Hugging Face: likes
Table
| Lightricks/LTX-2.5 | 11133 |
| moonshotai/Kimi-K3.1-Pro ▼1 | 1729 |
| speridlabs/idol | 1382 |
| ccore/Qwen3.5-4B-3D-Coder ▼3 | 917 |
| xtools-at/DeepSeek-V4-Pro-Qwen3.5-9B-UNCENSORED-heretic ▲2 | 838 |
| ricursiveAI/Ripple-65B-Preview ▲2 | 403 |
PirateBench: score out of 100
Table
| 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/
Goldman Sachs is pulling Claude Code from its bankers and traders
Goldman Sachs removes Claude Code from bankers and traders, returning the AI coding agents to engineers before the company takes Anthropic public. ·
574 points and 348 comments on Hacker News, linking thestateofai.com.
I'm not a fan of Deno's security model (2024)
82 points and 39 comments on Hacker News, linking goldstein.lol.
Llama-modchip: 284B DeepSeek on 64GB RAM, no GPU
Llama-modchip runs DeepSeek-V4-Flash and 30B MoEs on 64GB RAM without GPUs, achieving up to 111 tokens per second. ·
Claude Code's "Don't Ask Again" exposes passwords in 1,122 public GitHub files
Claude Code exposes passwords in 1,122 public GitHub files. The tool secures coding agents and supply chains. ·
mradermacher/DeepSeek-V4-Pro-Qwen3.5-9B-UNCENSORED-heretic-i1-GGUF
Mradermacher releases the DeepSeek-V4-Pro-Qwen3.5-9B-UNCENSORED-heretic-i1-GGUF model, a machine learning model with 9B parameters available on Hugging Face. ·
Goldman Sachs is pulling Claude Code from its bankers and traders
Goldman Sachs removes Claude Code from bankers and traders, returning the AI coding agents to engineers before the company takes Anthropic public. ·
574 points and 348 comments on Hacker News, linking thestateofai.com.
Llama-modchip: 284B DeepSeek on 64GB RAM, no GPU
Llama-modchip runs DeepSeek-V4-Flash and 30B MoEs on 64GB RAM without GPUs, achieving up to 111 tokens per second. ·
Claude Code's "Don't Ask Again" exposes passwords in 1,122 public GitHub files
Claude Code exposes passwords in 1,122 public GitHub files. The tool secures coding agents and supply chains. ·
I'm not a fan of Deno's security model (2024)
82 points and 39 comments on Hacker News, linking goldstein.lol.
moonshotai/Kimi-K3.1-Pro
Moonshotai releases Kimi-K3.1-Pro, an image-text-to-text transformer model featuring 2.8 trillion parameters available on Hugging Face. ·
X voices
5 storiesX posts other sources are linking to, then named voices on Bluesky and Mastodon
@DeepSeekHarness
DeepSeek Harness: Growing Together with Our Plugin Community
3 points and 0 comments on Hacker News, linking an X post.
@v12sec
Changing one letter bypassed an admin API in HuggingFace
7 points and 0 comments on Hacker News, linking an X post.
@make_claude
Now Anthropic is blocking Russian users
9 points and 6 comments on Hacker News, linking an X post.
@_FORAB
A guy running Claude KYC business to bypass anthropics restrictions
1 points and 0 comments on Hacker News, linking an X post.
@knowerofmarkets
Are You Ready for Superintelligence?
2 points and 0 comments on Hacker News, linking an X post.
@make_claude
Now Anthropic is blocking Russian users
9 points and 6 comments on Hacker News, linking an X post.
@v12sec
Changing one letter bypassed an admin API in HuggingFace
7 points and 0 comments on Hacker News, linking an X post.
@DeepSeekHarness
DeepSeek Harness: Growing Together with Our Plugin Community
3 points and 0 comments on Hacker News, linking an X post.
@1littlecoder
In less than $1, DeepSeek v4.1 Flash 1017 built a Mac App
4 points and 0 comments on Hacker News, linking an X post.
GitHub
5 storiesNew repos, releases and changelog
New AI repos this week: stars
Table
| KKKKhazix/agent-autoloop | 4101 |
| worldonlyu/Google_AI_Pro_2Year_ZeroCost_Auto | 3257 |
| sarpstar/cinecode | 1591 |
| Sandflow-ai/sand3d | 1387 |
| voyager-so/vork | 1118 |
| shanselman/gputaskman | 1005 |
Llama-modchip: 284B DeepSeek on 64GB RAM, no GPU
Llama-modchip runs DeepSeek-V4-Flash and 30B MoEs on 64GB RAM without GPUs, achieving up to 111 tokens per second. ·
355 points and 148 comments on Hacker News, linking github.com.
Outsideclofilter/aion-2-trainer-tool
Outsideclofilter/aion-2-trainer-tool provides a pre-configured environment for Aion 2, an AI-driven LLM prompt engineering resource, on local workstations in 2026. ·
elunboundfiremail/LiveWall---Live-Wallpaper-For-Desktop
A new repository with 311 stars in its first week.
Show HN: Maple – Sub-2-BPW Moe Pruning to Run Gemma 4 26B at 45 TPS on a 6GB GPU
7 points and 3 comments on Hacker News, linking github.com.
drost-ai/tinker-ai-skill
Drost-ai releases tinker-ai-skill, an agent skill for fine-tuning agentic models using Claude Code sessions. It works in Claude Code, Codex, and OpenCode. ·
Llama-modchip: 284B DeepSeek on 64GB RAM, no GPU
Llama-modchip runs DeepSeek-V4-Flash and 30B MoEs on 64GB RAM without GPUs, achieving up to 111 tokens per second. ·
355 points and 148 comments on Hacker News, linking github.com.
Outsideclofilter/aion-2-trainer-tool
Outsideclofilter/aion-2-trainer-tool provides a pre-configured environment for Aion 2, an AI-driven LLM prompt engineering resource, on local workstations in 2026. ·
elunboundfiremail/LiveWall---Live-Wallpaper-For-Desktop
A new repository with 311 stars in its first week.
VictorCabbie/trx-drainer-bot
A new repository with 300 stars in its first week.
lxe/phoneuse
lxe/phoneuse lets users view and control an unmodified iPhone from a Linux PC using AirPlay and Bluetooth, featuring a CLI and agent skill. ·
Infrastructure
5 storiesData centers, power, chips and what tokens cost
Data-center and energy stories per day
Table
| Tue 03 | 38 |
| Wed 04 | 40 |
| Thu 05 | 36 |
| Fri 06 | 25 |
| Sat 07 | 1 |
| Sun 08 | 2 |
| Mon 09 | 1 |
Stories from DCD, Data Center Knowledge, EIA and Hacker News, by publication day.
Output price, $ per million tokens (newest models)
Table
| lfm-3-1.7b | 0.2 |
| gemini-omni-1.2-flash | 9 |
| gemini-omni-1.1-flash | 9 |
| ripple-65b-preview | 0.6 |
| hy4 | 1.2505 |
| median, all paid models | 2 |
435 paid models listed; median of the 29 added in 30 days: $2.50. Rankings data by OpenRouter, CC BY 4.0.
Llama-modchip: 284B DeepSeek on 64GB RAM, no GPU
Llama-modchip runs DeepSeek-V4-Flash and 30B MoEs on 64GB RAM without GPUs, achieving up to 111 tokens per second. ·
355 points and 148 comments on Hacker News, linking github.com.
Show HN: Maple – Sub-2-BPW Moe Pruning to Run Gemma 4 26B at 45 TPS on a 6GB GPU
7 points and 3 comments on Hacker News, linking github.com.
AI:AM: Is THIS the AI We Want? Alarming Acceleration, AI Election Bets, GPU Trading & AI-Native VC
Nathan Labenz and Prakash Narayanan discuss AI acceleration with guests Brian Zhou, Mike San Miguel, and Bryan Bischof. ·
Llama-modchip: 284B DeepSeek on 64GB RAM, no GPU
Llama-modchip runs DeepSeek-V4-Flash and 30B MoEs on 64GB RAM without GPUs, achieving up to 111 tokens per second. ·
355 points and 148 comments on Hacker News, linking github.com.
Rune: A native GPU-rendered Rust UI framework for macOS, Windows, Linux and web
177 points and 131 comments on Hacker News, linking github.com.
Google has acquired Crusoe's data center site in southeast Wyoming
Google acquires Crusoe's southeast Wyoming data center site after the company completes site selection, approvals, and power strategy for the campus. ·
Show HN: Maple – Sub-2-BPW Moe Pruning to Run Gemma 4 26B at 45 TPS on a 6GB GPU
7 points and 3 comments on Hacker News, linking github.com.
Protesters break into Microsoft data center site in Amsterdam
18 points and 5 comments on Hacker News, linking theverge.com.
