2
1 Comment

Why bookmarking is broken.

We all stumble across interesting links whether they are news articles, research papers, blog posts, etc. Naturally, you want to ensure that you keep track of these interesting things, so, you add them to your bookmarks.

We've all been down this road and I have hundreds of links saved for a time when I want to read and study them. But rarely do I actually revisit them.

This creates the bookmark graveyard.

I've tried apps like Pocket, Raindrop, etc. but still, I end up with a bookmark graveyard. Why is that?

The Collector's Fallacy

Most of the time, bookmarking is just engaging in what is known as the "Collector's Fallacy" - the act of collecting items to remove the fear of losing them. However, "until we merge the contents ... into our own knowledge, we haven’t really learned a thing."

The apps I mentioned before allow me to create folders, tag, and search through my bookmarks, but that doesn't actually change anything. In fact, it can make the collector's fallacy worse.

By creating folders and adding tags I'm merely creating the illusion of productive information processing. Nothings changed. Again, it is only when you've wrestled and merged this information into your knowledge are you being productive.

So what is the point of bookmarking?

Useful bookmarking

You already know this. Bookmarking is useful for capturing the moments when you find something interesting and don't have the time to wrestle with it.

Fundamentally, you're taking the internet and making it more personal.

But bookmark managers don't understand this. They're just making feeds and folders while hoping you come back and look at the links.

Eventually, you will have a reason to come back, but unless you spent all that time carefully organizing your bookmarks you're just looking at a pile of links. Full-text search helps, but unless the perfect link shows up at the top of the results you're not that much better off.

How to fix bookmarking

In an ideal world, you would have a librarian who has already read and organized these links for you to simply walk into the library and have a shelf of useful information on the topic you want.

As someone who has studied ML, worked as an ML engineer, and taught others about ML, I know that we don't have AI that can meaningfully understand concepts as humans do. However, we have ML that can find structure across massive datasets and do it pretty damn well.

There's a reason why youtube has a rabbit hole. Its recommendation system knows the structure of the content it contains and can keep showing you items that are of interest. So why don't your bookmarks have a recommendation system?

With one, you just need one idea to pop into your head and this ML librarian can take the first full-text search result and keep guiding you through all the relevant items in your library like the ideal shelf. Or, you can jump into a random bookmark and be shown an interesting shelf of links you had forgotten about. If you want to find sources for a research project, you could just give it one and it will return a list of more for you to use.

This is my solution to bookmarking. Let the machines organize it and come back later when you want it. You'll have a shelf of links to chose from without wasting time organizing it.


Thanks for taking the time to read through this. I of course have begun building this better system of bookmarking over at lxi.ai.
It would mean a ton if you signed up and let me know if you are getting useful results. Or give me a follow on twitter @lxi_ai for more hot takes on organization.

posted to
Data Science
on March 19, 2021
  1. 1

    Final thought:

    Notetaking applications like Notion, Obsidian, and Roam all help with the other side of bookmarking. The side when you wrestle with the information and embed it into your own knowledge base. These second brains are becoming more and more useful so good bookmarking should help feed these brains.

Trending on Indie Hackers
I've built a 2300$ a month SaaS out of a simple problem. 22 comments Where can I buy newsletter ad promos? 12 comments How would you monetize my project colorsandfonts? 8 comments How I built my SaaS in 2 weeks using NextJS and Supabase 6 comments Key takeaways growing MRR from $6.5k to $20k for my design studio 6 comments Tips on starting a startup 5 comments