Out now!
A Year of Mornings
A collection of fifty love poems that follows a young heart as it finds love, finds the strength to be in love and finally, finds the strength to let go.
Nab your copy:
I am a platform engineer and a writer based in Finland.
I am the author of A Year of Mornings, a collection of poems for young adults.
NordLetter
I send out a newsletter once a week about living in Finland + five interesting things I've found on the open web.
Stream
Cyborgs vs rooms, two visions for the future of computing by
Loosely I can see two visions for the future of how we interact with computers: cyborgs and rooms.
Cyborgs is us being augmented with implants or devices we wear. Rooms is about ambient computing - we can walk into or away from a computer, which is more interesting to me too.
Apple adds a new toggle to make Liquid Glass less glassy by Jay Peters
The option is available now in the iOS 26.1 developer beta that Apple launched on Monday. You can access it from Settings > Display & Brightness > Liquid Glass, where you can choose between “Clear” and “Tinted.” The latest iPadOS 26.1 and macOS 26.1 developer betas also let you tint Liquid Glass
I like Liquid Glass generally, but there are a few places where I have come across a few bugs and legibility issues.
This is good I guess, but I could not figure out any differences in the screenshots in the article.
Claude Code for web—a new asynchronous coding agent from Anthropic by Simon Willison
Anthropic launched Claude Code for web this morning. It's an asynchronous coding agent - their answer to OpenAI's Codex Cloud and Google's Jules, and has a very similar shape. I had preview access over the weekend and I've already seen some very promising results from it.
This was the one missing feature I cared about, that was available on ChatGPT and not on Claude.
Now that it’s there, I can go subscribe to Claude.
Better than average
Claude Skills are awesome, maybe a bigger deal than MCP by Simon Willison
Skills are Markdown with a tiny bit of YAML metadata and some optional scripts in whatever you can make executable in the environment. They feel a lot closer to the spirit of LLMs - throw in some text and let the model figure it out.
Here’s the learn documentation.
It’s a simple feature. The bunch of things Anthropic has announced since the past one month while I’ve been on the ChatGPT train is staggering.
I am tempted to go back. The only thing ChatGPT has right now is the web ui for codex. This was ot would be free till 20th, after which the web usage will count toward the overall limits.
I think it makes sense to me to go back to Claude and try that out.
