Hello 👋, I'm Patrik

Latest thoughts and articles

Latest Articles

I built pbp to solve a problem I was having with my growing collection of side projects. You probably have the same problem.

Revolting against the Spotify algorithm

Changing how I consume music in 2025 to rely less on automated recommendations and get a deeper appreciation of it

How I use Obsidian to get work done

In this follow up to my post about how I set goals that align with my values we dig into how I use Obsidian to actually do all the things I talked about in the previous post.

How I get work done

All the way from my core values down to what I decide to do day by day. We go through weekly and daily planning based on values and how to figure out how to bucket time by mapping out an ideal week.

Generative AI and creativity

Will generative AI take over what we used to believe to be fundamentally human - creativity?

The definitive guide to productivity

This is how you get off the couch and achieve everything in life. A step by step guide.

Why write when AI can write better?

As AI becomes better at writing and creating, is there intrinsic value in human created content?

Temporary is permanent

What is recency bias in LLMs

Prioritizing recent inputs over older ones

Established vs new knowledge

How do we separate knowledge? How do we know that anything is correct?

Reflecting on information

Pondering a framework for how to take in new information.

An informative guide to managing time effectively using the methodology of 'forwards thinking - backwards planning', especially for parents.

Consistency

Consistency is king - but is it?

Multi-tasking is bad and does not work

Exploring the misconception of multitasking, its cognitive costs, overconfidence in abilities, and negative health implications.

An exploration into the harmful effects of overworking and constant productivity, offering insights on how to maintain a healthy work-life balance.

Latest Posts

Every time I use gpt-5 I get so underwhelmed and go back to ... any other model.

It's interesting the way cli coding tools read files like 'Read src/services/content_manager.py (limit=20, offset=95)' might mean that splitting things into short, readable chunks is primarly for the benefit of humans, AI doesn't seem to care.

1password - still figthing me at every turn. Why. I just want to use you.

I use a tiny CLI called bs to get thoughts out fast. It reads from args, stdin, or an editor, then posts to Bluesky and Mastodon. No ceremony, just words out.

super+b in hyprland config opens a small TUI, which launches nvim on a temp file. I write, :wq, and bs ships it. If it's long, Bluesky gets a threaded post with proper linking; Mastodon gets the same content.

bs updated: removed like mode and URL parsing. Posting only (Bluesky + Mastodon). Simplified CLI and code structure. This is a test of bs posting.

The recent spam is from my lovely 'bs' program that comes as part of my omarchy-pbjorklund project. https://github.com/pbjorklund/omarchy-pbjorklund/blob/master/bin/bs

New zen-browser update added folders and having used them for 5 minutes I realize I have missed them for as long as I have been using this wonderful piece of software.

But I think it's time for the project to start thinking more seriously about the role of bookmarks vs tabs. Being able to sync bookmarks but them being hidden away is strange, but I get it.

Testing the fixed threading implementation! This message should be long enough to split into two parts and create a proper thread on Bluesky. The first part posts normally, and the second part should reply to the first post using the correct ReplyRef structure with parent and root

references. This should fix the threading issue we had before.

My god getting kanata up with a second layer letting me user altgr+['; to get åäö actually had me learn som lisp.

Cool, got my own mastodon server up. Not sure why, but choosing a server gave me to much analysis paralysis.

The key skill in vibe coding is noticing when the LLM starts struggling and starts looping or just going the wrong way or where it obviously didn't have stellar training data.

Ok revanced for android is pretty cool.

Ok zellij is kinda nice. Tmux out, zellij in.

My dotfiles are now complete and perfect - said noone ever that has successfully quit vim at least once.

Every time I use gpt-5 I get so underwhelmed and go back to ... any other model.

It's interesting the way cli coding tools read files like 'Read src/services/content_manager.py (limit=20, offset=95)' might mean that splitting things into short, readable chunks is primarly for the benefit of humans, AI doesn't seem to care.

1password - still figthing me at every turn. Why. I just want to use you.

I use a tiny CLI called bs to get thoughts out fast. It reads from args, stdin, or an editor, then posts to Bluesky and Mastodon. No ceremony, just words out.

super+b in hyprland config opens a small TUI, which launches nvim on a temp file. I write, :wq, and bs ships it. If it's long, Bluesky gets a threaded post with proper linking; Mastodon gets the same content.

bs updated: removed like mode and URL parsing. Posting only (Bluesky + Mastodon). Simplified CLI and code structure. This is a test of bs posting.

The recent spam is from my lovely 'bs' program that comes as part of my omarchy-pbjorklund project. https://github.com/pbjorklund/omarchy-pbjorklund/blob/master/bin/bs

New zen-browser update added folders and having used them for 5 minutes I realize I have missed them for as long as I have been using this wonderful piece of software.

But I think it's time for the project to start thinking more seriously about the role of bookmarks vs tabs. Being able to sync bookmarks but them being hidden away is strange, but I get it.

Testing the fixed threading implementation! This message should be long enough to split into two parts and create a proper thread on Bluesky. The first part posts normally, and the second part should reply to the first post using the correct ReplyRef structure with parent and root

references. This should fix the threading issue we had before.

My god getting kanata up with a second layer letting me user altgr+['; to get åäö actually had me learn som lisp.

Cool, got my own mastodon server up. Not sure why, but choosing a server gave me to much analysis paralysis.

The key skill in vibe coding is noticing when the LLM starts struggling and starts looping or just going the wrong way or where it obviously didn't have stellar training data.

Ok revanced for android is pretty cool.

Ok zellij is kinda nice. Tmux out, zellij in.

My dotfiles are now complete and perfect - said noone ever that has successfully quit vim at least once.