[Industry Event] Advancing Data Center Construction East 2026
From Data Center Knowledge.
Top now
6 storiesWhat is climbing fastest across every source
LLM7.3xDeepSeek7.9xMercury 37.1xAnthropic5.2xFLUX-36.9xLong-Horizon Agents5.4xLearning5.2xVideo World5.1xIndiana3.9xPower3.8x
anthropics/claude-code v2.1.304
Anthropic releases Claude Code version 2.1.304, which adds new mod hooks, a guided setup command, and a security comparison tool. ·
Copilot model policy API for enterprise, orgs, and teams
GitHub provides a REST API that lets enterprises, organizations, and teams retrieve and update Copilot model policies through GET and PATCH endpoints. ·
GitHub Copilot models deprecated on October 19, 2026
GitHub deprecates several models across its Copilot experiences on October 19, 2026, affecting chat, inline edits, and code completions. ·
Towards Company World Models
The text describes AI as a strategic partner for essential businesses, where it learns from owner-operators to propose, implement, and measure better decisions. ·
New paper: Emerging trends in AI agent evaluation
A new paper examines emerging trends in AI agent evaluation, detailing the methods, questions, and ecosystem used to assess these systems. ·
anthropics/claude-code v2.1.304
Anthropic releases Claude Code version 2.1.304, which adds new mod hooks, a guided setup command, and a security comparison tool. ·
New release v2.1.304 of anthropics/claude-code on GitHub.
Mercury 3: Diffusion at scale
Mercury 3 is a diffusion language model that outperforms GPT-6 Luna and Claude Haiku 5.5 while running at 1,000+ tokens per second. ·
AI companies are coming for the plumbers
20 points and 17 comments on Hacker News, linking tomshardware.com.
AndreSzymanski15/ai-jailbreaks
A new repository with 237 stars in its first week.
unslothai/unsloth DeepSeek-V4.1 & FLUX-3
Unsloth releases DeepSeek-V4.1-Flash-1017 and FLUX.3 Image Klein for local use, adding image and video chat features, multi-GPU training, and one-command MCP setup for coding agents. ·
finereli/whatsapp-mcp
A read-only MCP server lets AI agents search and read personal WhatsApp messages while running on Cloudflare Workers. ·
finereli/whatsapp-mcp
A read-only MCP server lets AI agents search and read personal WhatsApp messages while running on Cloudflare Workers. ·
A new repository written in TypeScript with 285 stars in its first week.
anthropics/claude-code v2.1.304
Anthropic releases Claude Code version 2.1.304, which adds new mod hooks, a guided setup command, and a security comparison tool. ·
deepseek-ai/DeepSeek-V4.1-Flash-1017
A image text to text model from deepseek-ai with 1,684 likes and 17.9k downloads on Hugging Face.
AI companies are coming for the plumbers
20 points and 17 comments on Hacker News, linking tomshardware.com.
AndreSzymanski15/ai-jailbreaks
A new repository with 237 stars in its first week.
unslothai/unsloth DeepSeek-V4.1 & FLUX-3
Unsloth releases DeepSeek-V4.1-Flash-1017 and FLUX.3 Image Klein for local use, adding image and video chat features, multi-GPU training, and one-command MCP setup for coding agents. ·
Trending topics today: heat
Table
| LLM | 2.16 |
| DeepSeek | 1.7 |
| Mercury 3 | 1.53 |
| Anthropic | 1.42 |
| FLUX-3 | 0.75 |
| Long-Horizon Agents | 0.54 |
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
| google/embeddinggemma-2 ▼2 | 2725 |
| jialinyyzz/humanizer ▼2 | 1986 |
| deepseek-ai/DeepSeek-V4.1-Flash-1017 | 1684 |
| openbmb/MiniCPM-V-4.7-35B-A3B | 1063 |
| black-forest-labs/flux-3-image-klein-comfy ▲3 | 962 |
| fishaudio/drama-3 ▲1 | 769 |
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/
unslothai/unsloth DeepSeek-V4.1 & FLUX-3
Unsloth releases DeepSeek-V4.1-Flash-1017 and FLUX.3 Image Klein for local use, adding image and video chat features, multi-GPU training, and one-command MCP setup for coding agents. ·
New release v0.1.909-beta of unslothai/unsloth on GitHub.
CERN and Mistral sign a statement of intent to establish a strategic partnership
308 points and 57 comments on Hacker News, linking home.cern.
Show HN: Qwen3-4B with the AI writing tics trained out
A developer fine-tunes Qwen3-4B on a MacBook Air to remove specific writing patterns, sharing answers on forty held-out prompts and training details. ·
Launch HN: Laminar (YC S24) – RL-trained model that finds errors in agent traces
44 points and 22 comments on Hacker News, linking news.ycombinator.com.
Maxwell: A new kind of model that sits between code and hardware
Maxwell is a model that improves code during execution by adapting to hardware and data, with each change formally proven equivalent to the original. ·
deepseek-ai/DeepSeek-V4.1-Flash-1017
A image text to text model from deepseek-ai with 1,684 likes and 17.9k downloads on Hugging Face.
unslothai/unsloth DeepSeek-V4.1 & FLUX-3
Unsloth releases DeepSeek-V4.1-Flash-1017 and FLUX.3 Image Klein for local use, adding image and video chat features, multi-GPU training, and one-command MCP setup for coding agents. ·
CERN and Mistral sign a statement of intent to establish a strategic partnership
308 points and 57 comments on Hacker News, linking home.cern.
Show HN: Qwen3-4B with the AI writing tics trained out
A developer fine-tunes Qwen3-4B on a MacBook Air to remove specific writing patterns, sharing answers on forty held-out prompts and training details. ·
Maxwell: A new kind of model that sits between code and hardware
Maxwell is a model that improves code during execution by adapting to hardware and data, with each change formally proven equivalent to the original. ·
X voices
5 storiesX posts other sources are linking to, then named voices on Bluesky and Mastodon
@AndrewCurran_
Homeland Security is requiring all civilian federal agencies to remove DeepSeek
5 points and 1 comments on Hacker News, linking an X post.
@brandonstirrup
Xbox Business Strategy Manager: "we use AI, just as all companies do in 2026"
2 points and 1 comments on Hacker News, linking an X post.
@AskPerplexity
Perplexity offers free reverse engineering credits
8 points and 2 comments on Hacker News, linking an X post.
@nichochar
Using AI as a tutor to do math with pen and paper
5 points and 1 comments on Hacker News, linking an X post.
@worldnetwork
Muse, dots, Instinct, and the question every agent will be asked
1 points and 0 comments on Hacker News, linking an X post.
@AndrewCurran_
Homeland Security is requiring all civilian federal agencies to remove DeepSeek
5 points and 1 comments on Hacker News, linking an X post.
@brandonstirrup
Xbox Business Strategy Manager: "we use AI, just as all companies do in 2026"
2 points and 1 comments on Hacker News, linking an X post.
@AskPerplexity
Perplexity offers free reverse engineering credits
8 points and 2 comments on Hacker News, linking an X post.
@nichochar
Using AI as a tutor to do math with pen and paper
5 points and 1 comments on Hacker News, linking an X post.
@anemll
Qwen3.8-27B on the iPhone 18 Neural Engine at 15–70 tok/s (1.5 bpw)
30 points and 12 comments on Hacker News, linking an X post.
GitHub
5 storiesNew repos, releases and changelog
New AI repos this week: stars
Table
| feder-cr/git-ai | 2450 |
| Leonxlnx/earshot | 1512 |
| aicodeking/leanrea | 990 |
| LiCinCan/blender-ai-mcp-addon | 956 |
| outcrawlai/scout | 950 |
| OpenMOSS/MOSS-Image | 944 |
AndreSzymanski15/ai-jailbreaks
A new repository with 237 stars in its first week.
unslothai/unsloth DeepSeek-V4.1 & FLUX-3
Unsloth releases DeepSeek-V4.1-Flash-1017 and FLUX.3 Image Klein for local use, adding image and video chat features, multi-GPU training, and one-command MCP setup for coding agents. ·
finereli/whatsapp-mcp
A read-only MCP server lets AI agents search and read personal WhatsApp messages while running on Cloudflare Workers. ·
Harness Engineering 101: The LLM is the brain, the harness is the body
A guide explains how to construct AI agents by treating the large language model as the brain and the surrounding system as the body. ·
Harness2 by Google Research – Recursive Agent Harnessing
54 points and 13 comments on Hacker News, linking github.com.
finereli/whatsapp-mcp
A read-only MCP server lets AI agents search and read personal WhatsApp messages while running on Cloudflare Workers. ·
A new repository written in TypeScript with 285 stars in its first week.
AndreSzymanski15/ai-jailbreaks
A new repository with 237 stars in its first week.
unslothai/unsloth DeepSeek-V4.1 & FLUX-3
Unsloth releases DeepSeek-V4.1-Flash-1017 and FLUX.3 Image Klein for local use, adding image and video chat features, multi-GPU training, and one-command MCP setup for coding agents. ·
Harness Engineering 101: The LLM is the brain, the harness is the body
A guide explains how to construct AI agents by treating the large language model as the brain and the surrounding system as the body. ·
Harness2 by Google Research – Recursive Agent Harnessing
54 points and 13 comments on Hacker News, linking github.com.
Infrastructure
5 storiesData centers, power, chips and what tokens cost
Data-center and energy stories per day
Table
| Wed 14 | 34 |
| Thu 15 | 41 |
| Fri 16 | 35 |
| Sat 17 | 5 |
| Sun 18 | 2 |
| Mon 19 | 30 |
| Tue 20 | 0 |
Stories from DCD, Data Center Knowledge, EIA and Hacker News, by publication day.
Median output price of models added in the last 30 days, $ per million tokens
One point per OpenRouter reading, last 8 days. Data by OpenRouter, CC BY 4.0.
Output price, $ per million tokens (newest models)
Table
| deepseek-v4.1-flash-1017 | 0.6 |
| glimpse | 3 |
| beam-1.0-501b-a23b | 1 |
| mercury-3 | 0.35 |
| moondream3-preview | 2.5 |
| median, all paid models | 2 |
441 paid models listed; median of the 50 added in 30 days: $3.20. Rankings data by OpenRouter, CC BY 4.0.
Detecting model weight tampering in GPU memory (Rome attacks)
Stealthium detects ROME attacks, where malicious Python modules rewrite model weights in GPU memory, by linking CUDA kernel launches to their originating code. ·
15 points and 1 comments on Hacker News, linking stealthium.io.
Belief-Trajectory Energy: Measuring the Path to a Prediction
Researchers introduce Belief-Trajectory Energy, a metric that tracks how large language models revise predictions across layers. The method achieves 0.998 macro-AUROC for detecting human-written text. ·
Detecting model weight tampering in GPU memory (Rome attacks)
Stealthium detects ROME attacks, where malicious Python modules rewrite model weights in GPU memory, by linking CUDA kernel launches to their originating code. ·
15 points and 1 comments on Hacker News, linking stealthium.io.
Belief-Trajectory Energy: Measuring the Path to a Prediction
Researchers introduce Belief-Trajectory Energy, a metric that tracks how large language models revise predictions across layers. The method achieves 0.998 macro-AUROC for detecting human-written text. ·