I’m behind on writing about Qwen 3.5, a truly remarkable family of open weight models released by Alibaba’s Qwen team over the past few weeks. I’m hoping that the 3.5 …
What's Next at Bluesky - Bluesky
As we head into 2026, we're entering a new phase for the Bluesky app. Last year was about scaling through rapid growth and getting the fundamentals in place. This year is about leaning into what's working and investing more intentionally in the things that make Bluesky different.
Don't Take the Black Pill - Andrew Kelley | SSW 2026
How Much Money Can We Make Gold Hunting?
CodePen 2.0
Noting perhaps my largest personal career accomplishment, which is launching CodePen 2.0. Far more work, believe it or not, than the entire creation of the original CodePen. This isn’t the pl…
You have to beat the models at something
In 2025, I wrote that software engineers ought to be assessed by “value over replacement”: not how much money they made for their company, but how much they…
Atelier — Your work, in your hands
A transparent foundation preview for an open productivity suite spanning notes, mail, calendars, and tasks, powered by ATProto.
Are we evolving from pages to apps to agents?
Recently, Vercel came to London and teamed up with Anthropic’s local chapter to host a skills event.
The Story Behind our Brand Illustration
Our brand illustration shows three layers: the foundation being built, the valley we're stuck in, and the world waiting above. Here's what it all means.
The future is low-tech: lessons from the early 2000s ft. Jésabel DC | Config 2026
I’m BEGGING You To Make Zines
ATmosphereConf 2026 - Sooraj's Scribbles
My thoughts and experiences as I volunteered at ATmosphereConf 2026!
How StrongDM’s AI team build serious software without even looking at the code
Last week I hinted at a demo I had seen from a team implementing what Dan Shapiro called the Dark Factory level of AI adoption, where no human even looks …
China Just Built What NVIDIA Feared Most. I Saw It
Dear Manager, Again - Ewan’s Blog
Five months later, I have received essentially the exact same email.
Qwen3.6 now runs 2x faster with MTP GGUFs! Run locally on just 18GB RAM. ⚡️ MTP enables Qwen3.6 to generate ~1.4–2.2× faster with no accuracy change. Qwen3.6-27B MTP runs at 160 tokens/s. 35B-A3B reaches 240 t/s. GGUFs: huggingface.co/unsloth/Qwen3.6-27B-MTP-GGUF Guide: unsloth.ai/docs/models/qwen3.6#mtp-guide
