3031
Discourse Is Not Going Closed Source
Cal.com just closed their source code, arguing AI has made open source too dangerous. After 13 years of building Discourse in public, we're staying open. Here's why.
3032
The Q, K, V Matrices
At the core of the attention mechanism in LLMs are three matrices: Query, Key, and Value. These matrices are how transformers actually pay attention to different parts of the input. In this write-up, we will go through the construction of these matrices from the ground up.
3033
Every single board computer I tested in 2025
I benchmarked 15 SBCs released in 2025 across every price point. Here's what stood out, what disappointed, and what the year looked like for single board computers.
3034
How my e-reader lost its stripes
I’ve been a lifelong reader, which has proved a fragile habit in our era of endless trivial distraction. While books and e-ink readers solve the problem of competition for attention, they can’t compete with a phone for convenience. That tiny device in my pocket has made reading both easier and more vulnerable to…
3035
BMW Group to deploy humanoid robots in production in Germany for the first time
+++ BMW Group bringing Physical AI to Europe +++ Pilot project at BMW Group Plant Leipzig +++ New “Center of Competence for Physical AI in Production” accelerates global integration of AI and robotics in production +++ First pilot deployment of humanoid robots successfully completed at BMW Group Plant Spartanburg, USA…
3036
Fable and the end of the free lunch
Fable’s price ended the free lunch of ever-cheaper frontier models. Now we have to think about which model gets which work.
3037
How Hollywood stopped making movies in Hollywood
The economics behind where movies are filmed, why Hollywood left Los Angeles, and whether audiences can tell the difference.
3038
sp.h: Fixing C by giving it a high quality, ultra portable standard library
Over the past year, I’ve been working on fixing C by giving it a high quality, ultra portable standard library. It is not a simple wrapper on top of libc; it doesn’t depend on libc except when required to by the platform. To my knowledge, there is nothing like it.
3039
Union types in C# 15
C# 15 introduces union types — declare a closed set of case types with implicit conversions and exhaustive pattern matching. Try unions in preview today and see the broader exhaustiveness roadmap.
3040
It's not empowering to hand off the details
I think behind a lot of the enthusiasm around AI is a dream of being able to manifest things into reality without having to get into the details.
3041
3042
GLM 5.2 is nearly as accurate as a human book keeper
A real quarter of UK SME books, 59 transactions, 68 minutes, $2.73 of tokens — and a VAT return within 7 pence of the human-prepared one.
3043
The “mechanical miracle” that ruined Mark Twain’s life
Learning from the automation dreams of the first Machine Age
3044
3045
Court Filings: ICE App Identifies Protesters; Global Entry, PreCheck Get Revoked
ICE is using a smartphone app called “Mobile Fortify” to scan faces and capture contactless fingerprints, instantly pulling back names and biographical data — and court filings say the same encounters are being followed by revocations of Global Entry and TSA PreCheck. That turns “trusted traveler” into chilling of…
3046
P99 0 ms* autocomplete for 240M domain names
How Wirewiki's domain-name autocomplete renders suggestions next frame for 99% of keystrokes.
3047
Knuth reward check
Twenty years ago I got a reward check from Donald Knuth.
3048
Don't Trust the Salt: AI Summarization, Multilingual Safety, and LLM Guardrails
“The devil is in the details,” they say.
3049
My DIY FPGA board can run Quake II
kind of a blog
3050
Ternary Bonsai: Top Intelligence at 1.58 Bits
Today, we’re announcing Ternary Bonsai, a new family of 1.58-bit language models designed to balance strict memory constraints with high accuracy requirements.
3051
Building an FPGA 3dfx Voodoo with Modern RTL Tools
This frame of Screamer 2 was rendered not by an original 3dfx card and not by an emulator, but by an FPGA reimplementation of the Voodoo 1 that I wrote in SpinalHDL. Available on GitHub .
3052
The React2Shell Story
The story of CVE-2025-55182 (React2Shell)
3053
Clojure 1.13 adds support for checked keys
Clojure 1.13.0-alpha1 is now available! Find download and usage information on the Downloads page.
3054
We might all be AI engineers now
The models are good now. But most people still miss the point.
3055
3056
What I'm Hearing About Cognitive Debt (So Far)
A week ago, I wrote about how Generative and Agentic AI may be amplifying what I’ve been calling cognitive debt: the accumulated gap between a system’s evolving structure and a team’s shared unders...
3057
Rebasing in Magit
I read Ian Whitlock’s article on why he can’t quit Magit and it inspired me to share more about Magit from my perspective. This article will focus on rebasing.
3058
Wolfram Language and Mathematica version 15
AI-related upgrades like the built-in AI Assistant, extended computation-augmented generation and the Wolfram Agent Tools framework. Advances in TimeSeries & Tabular. New categorical data, ModelFit automated model selection, symbolic music. Efficient handling of gigabyte-sized notebooks. Gains in the already-robust…
3059
AI for American-produced cement and concrete
Meta is continuing its long-term roadmap to help the construction industry leverage AI to produce high-quality and more sustainable concrete mixes, as well as those exclusively produced in the Unit…
3060
China DRAM Maker CXMT Targets $4.2B IPO as It Takes on Samsung, SK Hynix, Micron
Share Latest News Of Ic&Pcb-Ic&Pcb Directory Of…2026224 points115 commentsTech Industry & Business details
Chinese memory chipmaker ChangXin Memory Technologies (CXMT) has filed for an IPO on Shanghai's STAR Market, seeking to raise up to CNY 295 billion (about USD 42.2 billion) to upgrade DRAM manufacturing, advance next-generation memory technologies, and strengthen China’s position in the global memory supply chain.