I’m not sure “deworming” is the best semantic fit, but it was first to come to mind.
Other apps will take the place of numerous browser functions, of which these four are primary:
- Reading List
- Bookmarks
- Password Manager
- History
As it is, the reading list and bookmarks are inadequate. The reading list lacks categorization and search. Bookmarks lack the ability to archive the website and highlighting/remembering salient points. Neither of them have great UX. I’ve managed to somewhat rig up a solution using Raindrop and Zotero, but I’m not quite happy yet. I believe these to be the best options at the moment due to the SaaS pricing and lack of local data storage for many alternatives. Of the two, I’m most happy with Zotero due to its local storage, website screenshot tool, and annotation tool. It handles all types of sources from PDFs to epubs to websites. Unfortunately, Zotero is not great for real-time sharing, but its open-source nature is a good foundation. I’d imagine one last important thing is a knowledge retrieval system.
I tried to see if using ChatGPT to auto-categorize my articles would be useful, but it seems that it can never exactly represent things how I want them. The mental delineations I have are sharper and more salient; ChatGPT feels like a blob. Another question I had was whether ChatGPT could do “pre-reading” of articles for me, finding necessary vocabulary beforehand. This is impossible as in order for it to highlight vocabulary that is significant to me, it would have to create a representation of my own mind and knowledge. At best, it can create some flashcards. Even then, the flashcards don’t differ that much even if you change the settings in the app I linked. However, it is a unique and different way of reading an article, which helps understanding and perception. I’m positive on ChatGPT for retrieval, but I think cognition, organization, learning, and the consequences/significance have to be the act of the human, even if the last is hard to do for humans.
Oh, I should also write a post about why Notion is poorly suited for these tasks. But that will be for another day.
Fake Productivity and Overcategorization
Overall, I would say the “second brain” idea is quite promising and a solution is in sight. This solution does not involve overcategorization, as many popular YouTubers do it. I never categorize my assignments by class anymore but rather temporally: by semester. All assignments for the current semester go into a folder inside the Downloads folder.
Overcategorization is a sure guarantee that you will never use the information you learned, and these “Notion/Obsidian” brain links will require significant active organization. Maybe it works for some people, but not for me. That’s not to wholly push against the idea of Obsidian’s graph tool. It’s quite useful for inductive reasoning (or maybe abductive, as some would call it) to discern underlying principles, such as I’ve done for a list of bad decisions. From the many come few.
Also of useful distinction is between shared files and variable files. The Linux Filesystem Hierarchy brought this up.
I think that categorization and grouping is an inevitable human construct, but the best categorizations create lines where they matter. Hmm, writing this has me suddenly interested in set theory and group theory, but idk how relevant they are.
I’ve seen some examples online at sharing the links people read. (note, just saw these in my twitter feed)
If you’re interested in password security, here is a good blog post. Password managers such as 1Password, Bitwarden, etc. replace this part of the browser. KeepassXC and MacPass are also good. But the big issue is poor browser integration/phishing due to pasting passwords in.
I’ve not seen a convincing history tool out there yet, but there is a need for local storage and better search tools. I tried both BrowserParrot and HistoryHound.
I get the feeling someone is fake productive when all they have are fancy new tools: Arc browser, Raycast, all while their menu bar is the default menu bar. I get the feeling someone who rolls their entire system may be not productive enough due to being distracted by personal technical system setup (me included).
Articles about learning systems
- https://writings.stephenwolfram.com/2019/02/seeking-the-productive-life-some-details-of-my-personal-infrastructure/
- https://reasonabledeviations.com/2023/02/05/gpt-for-second-brain/
- https://every.to/superorganizers/how-to-build-a-learning-machine-299655 (this strikes me as less “intense” as the other ones)
- https://publish.obsidian.md/history-notes/01+Notetaking+for+Historians
- https://news.ycombinator.com/item?id=22106992