Featured post what's this? ✨ The fastest double-to-string algorithm you’ve never heard of The fastest double-to-string algorithm you’ve never heard of
My Ten Years in No Man’s Sky - NMSspot Today marks the 10-year anniversary of the launch of Hello Games’ space exploration survival game, No Man’s Sky. On August 9th, 2016 the game landed on the Playstation 4 (with the PC version arriving a few days later). I remember getting home from work th...
How I Built a SaaS Business From Hostels and Coffee Shops | Pierre-Henry Soria – CTO Insights, Software Architecture & Product Leadership Imagine working at 3 a.m. in a coffee shop filled with students and night owls, or coding in a cramped hostel dorm, squeezed between triple bunk beds. For over 14 years, this was my reality—buildin...
Ben Godfrey When I was a student, I used to regularly drive right across Scotland, from the west coast to the east. Every time, I would get to Glasgow and be faced with a decision: I could either go north from Glasgow over the Kincardine Bridge, or I could head east ...
The case for disaggregated LLM serving | Doubleword Three checkable conditions under which disaggregated prefill is never worse than aggregated serving, and the reasons it's usually better. Plus an inventory-theory model of traffic drift: cold starts, safety stock, and a map of the shocks a deployment can ...
We're entering the human era While the investors, VCs, governments, and every influencer who can read news are all heralding the oncoming age of AI, I believe that the opposite is happening. We're entering the human era. The complexity of human mind has always been underestimated. W...
Introducing Grok 4.6 Grok 4.6 builds on Grok 4.5 with a particular focus on long-running agents and more ambitious interactive and visual work.
Grok 4.6 returns SpaceXAI to the intelligence frontier and leads on cost efficiency SpaceXAI returns to the intelligence frontier with strengths in agentic performance and cost efficiency
Qwen/Qwen3.8-2.4T-A95B · Hugging Face We’re on a journey to advance and democratize artificial intelligence through open source and open science.
Caveman Promises 65% Fewer Tokens. My Bill Didn't Move. Caveman, a wildly popular Claude Code plugin, promises to cut your AI coding tokens by 65% by making the model drop the filler and grunt. It works — the model really does write less. But those words are only ~0.6% of a real session's tokens and ~20% of it...
Introducing Delta - Zed Blog From the Zed Blog: A multiplayer environment for coding with agents, from the creators of Zed.
Purpose and intention — var0.xyz Good code doesn't only describe what it does. It communicates why it exists. A look at how intent shapes names, abstractions, comments, and commit messages.
"The Search Results Are Poisoned" Today I noticed that AI agents are detecting and rejecting content that they suspect is AI-written.
I never changed careers, only systems What a bass guitar, a Commodore 64, and a mixing desk taught me about becoming a CTO. A few weeks ago I was cleaning up my Google Drive, which is the digital equivalent of finally opening that one drawer everyone has at home - the one with cables for devi...
Demographic Dividend, my ass; overpopulation sucks I find it striking how the world often swings between two extremes. Look at American politics. It is hard to believe that only a few years ago the CEO of Twitter had to apologize for [buying a sandwich from Chick-fil-A](https://www.usatoday.com/story/tec...
How Tailscale helped find the SQLite WAL-Reset bug Tailscale and SQLite developers traced maddening corruption incidents to find the WAL-Reset data race, then uncovered a second stale expression index bug.
License Plate Reader Searches Should Require a Warrant So while I work with police departments regularly, I think it is critically important that technology be used reasonably. While this may be off-putting to some of my clients, I worked with the Inst…
DeepSeek V4 Pro 0813 - API Pricing & Benchmarks DeepSeek V4 Pro 0813 is a large-scale mixture-of-experts model from DeepSeek. $0.435 per million input tokens, $0.87 per million output tokens. 1,048,576 token context window, maximum output of 384,000 tokens. Includes independent benchmarks from Artifici...
AI is removing the middle class of software engineering AI makes projects with weak engineering culture fail much faster.
Where Did the Productivity Gains Go? There is a machine called Productivity, and on this machine there is a knob and a button. The knob increases productivity, and the button is labeled "reset."
