George Chiramattel

I'm so gosh-darned excited! css-tricks.com/css-tricks-h...

My first open model lecture, covering all the many pieces that build the ecosystem, and the open questions/debates determining where it goes next. How does distillation fit in? How do we manage risks? How do we see access to models changing in the future? Video: www.youtube.com/watch?v=79Pn...

Is it time to start the community-built HN/lobsters replacement?

This situation is becoming totally crazy.

Couldn't make it to #CloudflareConnect 2026 in San Francisco? We've got you covered. Keynotes, breakout sessions, and resources from Connect are now available on demand — and new sessions are being added regularly. Explore what you missed: https://cfl.re/4jDE65e

If you make HTTPS requests or have other TLS clients and care about MITM attacks, you need to update today. If you run Go: go get go@1.27.3 If you run Homebrew: brew update brew install ca-certificates If you run Linux, maybe stay off the internet for now.

Announcing DuckDB 2.0.0 LTS “Cinnamon Teal” 🦆 This is the biggest #DuckDB feature release ever, and it's shipping a polished version of the Quack protocol for DuckDB-as-a-server, async I/O, a first-class VARIANT type, agent mode, triggers, an extensible SQL parser and more – read on for details.

I built a drone with only one motor [video] View Article | Discussion + AI Summary Summary of HN discussion 🧵👇

Stereographic Rubik’s

How the Claude Code team does code review View Article | Discussion + AI Summary Summary of HN discussion 🧵👇

Six projects from the last few weeks that caught our attention: A wildfire analysis prototype that maps the fire’s spread over time and tracks hot spots, made with Opus 5.5.

Every lightweight self-hosted utility now seems to ship with a dedicated Postgres container attached. Commenters across the thread pushed back hard on that sprawl, wishing more tools just defaulted to SQLite. #opensource 1/4

We’re releasing Jev 1.14.0 along with experimental features for dreamers View Article | Discussion + AI Summary Summary of HN discussion 🧵👇

I recorded a short video of how the game I build with my boys also can teach you programming along the way. Also embeds a pi-durable agent straight in browser to be a teacher. www.youtube.com/watch?v=_35O...

Introducing Void: an open-source SDK, CLI, and self-hosted platform for building full-stack Vite apps and deploying them to Cloudflare. https://cfl.re/4yrzyCL

OpenCode 2.0 View Article | Discussion + AI Summary Summary of HN discussion 🧵👇

REA Reverse – Engineer Anything View Article | Discussion + AI Summary Summary of HN discussion 🧵👇

Everything is open source if you know assembly. And now with agents... rea.tools

Anthropic acquired the Bun and now Deno is acquired by Cloudflare.

📊 LayerChart: a new dimension New in v2.6.0, `view={isometric}` can be passed to any Chart to add a tilted perspective to your Svg, Canvas, and Html layers. Going further, use the `z` accessor to add (faux) 3D depth rendering for most marks.

Show HN: Bigwords.page – Turn any screen into a sign. The URL is the app View Article | Discussion + AI Summary Summary of HN discussion 🧵👇

You can now train your own Decision model like Jev locally! We increased Qwen3.5 0.8B’s aggregate accuracy from 20.7% to 74.3% across 3 decision benchmarks - on just 4GB VRAM. Turn any LLM, Qwen3.8, Gemma 4 into decision models. GitHub: github.com/unslothai/un... Guide: unsloth.ai/docs/basics/...

Workers KV Instant delivers sub-2ms p99 read latencies and 250ms global replication across Cloudflare’s 300+ edge locations. KV Instant eliminates cold-read penalties and uses the familiar Workers KV API. https://cfl.re/4yArgt1 #BirthdayWeek

We are introducing Clef and Clef-flash, open-source decision models hosted on Workers AI for high-speed classification and agentic workflows. https://cfl.re/3VotBZt #BirthdayWeek

svelte.dev/blog/sveltek...

We released Pi 1.0! earendil.com/posts/pi-1-0/

I've published detailed notes and an annotated transcript to accompany the video of the keynote I gave at @wearedevelopers.bsky.social World Congress North America in San Jose on Friday - here's my rundown of everything that's happened with LLMs in 2026 so far simonwillison.net/2026/Sep/27/...

Very much enjoying this “GPU glossary” mini-book by Modal I got from Can Duruk when stopping by at Modal’s NYC office. I don’t know if the physical books can be purchased, but the book has an online version, as I learned; with the same contents: modal.com/gpu-glossary

Claude made a music video for my linear algebra song and it is insane

Opus 5.5 is good at explainer videos View Article | Discussion + AI Summary Summary of HN discussion 🧵👇

Revealing the details of how OpenAI agents hacked Hugging Face View Article | Discussion + AI Summary Summary of HN discussion 🧵👇

Transformers Explained Visually View Article | Discussion + AI Summary Summary of HN discussion 🧵👇

I was asked by Congressional staff to share my views on open models performance, adoption, and competition vis a vis China. Here’s my briefing with all our latest data and predictions for what comes next. www.interconnects.ai/p/the-curren...

> Once a classification task becomes stable and volume, you could replace Jev with a dedicated trained classifier for that task, which would be cheap and faster. With maybe something like this 🤔 cactuscompute.com/needle

Cheap instant judgment based on extensive real-world knowledge can unlock numerous applications that appear intelligent without resembling AI chatbots. Really excited for this trend.

The first new cat species discovered in 100 years View Article | Discussion + AI Summary Summary of HN discussion 🧵👇

great guest post by @3blue1brown.com

I am glad that this is getting some attention in Hacker News. We require models that possess sufficient real-world knowledge to render adequate judgment calls. Other labs should also pick this up.

Fast judgement seems to be emerging trend in AI now. This development extends beyond Jev, as numerous alternatives are currently available.

wait- why does the entropy go down?

Fable 5.1 reportedly cracked the 370-year-old Cyphral Distich, but commenters immediately zeroed in on verification: did the model actually solve the cipher, or did an obscure historical translation simply leak into its training set? #ai 1/4

Registration without a phone number on Signal will use zero-knowledge proofs View Article | Discussion + AI Summary Summary of HN discussion 🧵👇

Steam Frame starts at $1059 View Article | Discussion + AI Summary Summary of HN discussion 🧵👇

iOS 27, iPadOS 27, and macOS 27 View Article | Discussion + AI Summary Summary of HN discussion 🧵👇

Qwen3.8 Flash Next is now supported in DwarfStar, covering 64GB Mac systems very well and with very fast inference of 50~70 t/s and > 1400 t/s prefill. For now this is Metal only. Thanks to @ivanfioravanti.bsky.social for all the cool work in the PR. N-grams on SSD like for DS4.1F.

Qwen3.8-27B Unsloth GGUF is now the #1 most-liked GGUF of all time! The model hit 10M downloads and 3.7K likes in just 24 days on Hugging Face - all thanks to you. 🤗🦥 GGUF: huggingface.co/unsloth/Qwen... Guide: unsloth.ai/docs/models/...

GLM-5.3 is now open-weight. Their most capable model for agentic coding and cyber defense is now available to download, run, and customize. Weights: huggingface.co/zai-org/GLM-5.3 Tech blog: z.ai/blog/glm-5.3

I were 17, I'd learn how to build LLMs from scratch View Article | Discussion + AI Summary Summary of HN discussion 🧵👇

The entire city of San Francisco as a video game View Article | Discussion + AI Summary Summary of HN discussion 🧵👇

Huzzah tries to fix a specific failure mode: losing the original "why" when models generate mountains of code. Instead of relying on drift-prone commit messages, it keeps pseudocode alongside the generated output as a permanent record of intent. #devtools 2/4