WTF? Írja mikor buildel. 😂

Megrántotta és kitört?

Minek neki az aranyered?

…

🕺

MP dugott! Azt a kurva eget! Ez volt O1G Mihály atombombája?

Finally, some good news!

I like this guy chilling on the front page of Claude.ai waiting for Sonnet 5 to be released.

Check the full post here: decoding.io/2026/01/las...

The other thread was a PKM audit. My old setup had overlap everywhere, so I simplified: each tool does one thing. Drafts captures, Tinderbox processes, and the output splits between Zettelkasten, OmniFocus, and DEVONthink.

This one started with a video about M-shaped careers and turned into three connected notes about how AI changes what "depth" means. I don't write code anymore, I review it while Claude writes and I direct.

I'm trying out a new format: writing about what's happening in my Zettelkasten every couple of weeks. We'll see if it sticks.

Ne tedd túl magasra a lécet a második évaddal kapcsolatban. Vissza fog mászni a From.

Yay! Ment az (RSS) subscribe like!

Wrote up how I built this with OmniFocus, DEVONthink, and Hookmark: decoding.io/2025/12/my-...

The key is that Claude loads all the project context (notes, transcripts, task history) so I don't have to explain everything each time. It reasons from what's already there.

For knowledge work, AI is the copilot and I do the thinking. For life admin, I flipped it. Claude does the thinking (loads context, generates call scripts, drafts emails) and I become the hands and feet that execute.

I've been experimenting with delegating the thinking part of life admin to Claude Code. Not the actions themselves, but the cognitive overhead of figuring out where I left off each time I return to a bureaucratic project.

It also searches my zettelkasten and links back to things I've forgotten.

Wrote about the workflow and recorded a demo: decoding.io/2025/12/thi...

No reorganizing. The thread keeps expanding. This preserves how ideas actually develop instead of forcing structure upfront.

I added an AI layer that watches for questions in my outlines and responds inline as blockquotes. The agent writes like margin notes, not chat.

I've been looking for a tool that supports threaded thinking. Distill by @TGUPJ nails the concept but it's pricey, hosted, and not native. I want local files, so I can use git. So I built something similar with Bike.

The core idea is append-only threading. New thoughts go at the end.

Jesus Fucking Christ!

Érzem a képen a pálinkás TSZ lábszagot.

I am eagerly waiting for the API/MCP integration. Any chance to get into the beta?

Despite its bugs and questionable choices, I’m starting to like Liquid Glass. I really hope Apple gets their act together and iterates heavily on it because there are good ideas here, but it needs a lot of refinement.

I can drag the window through the app menu…

NOBODY QA-D THIS SHIT?

WWDC 2001 111 Make It Aqua Adopting the Mac OS X User Experience

And I'm sure there are more examples like this in Liquid Glass.

I miss the times when we had long WWDC sessions talking about pixel level spacing and shaming weird UIs.
www.youtube.com/watch?v=Elr...

I understand that Alan Dye wants to capsulate the whole world, but should we clip scrollbars?

Checkbox labels should not contain explination text. The best: the same settings window contains the proper example. 🙃

Since when do we indent preferences like this?

Why do we even have the HIG if junior designers at Apple create stuff like this?