IntelliRead

This page is a collection of my experiments and thoughts on personal knowledge management, note capturing systems, bookmarks, and the evolution of tools related to these from the early days of the web.

I was a voracious reader even as a child. It was the mid 90s and we didn’t have access to the Internet. Our only source of information was newspapers, magazines and books. I used to bring books from our school library, read them, and then take notes about what interested me. I would also write down my own thoughts. I also saw myself as an author and would write short books on different subjects.

The Value of Information

And then in 1999, I came across an article in a magazine which describes the problem of Information Overload and how it has become severe due to the Internet. As per the article, until the advent of the Internet, getting information was a problem, but now that’s no longer the case. The problem now is of being drowned with so much information; it’s more of finding the right information at the right time.

As an engineering student (early 2000s), I felt the pain that a lot of students go through. Until our 12th, we have a fixed curriculum and about a year to study them. We have good faculty who understand these subjects with years of teaching. But in engineering, we may have as many as 6 subjects per semester and there is no limit to the depth of knowledge that we can acquire in each subject. And even that is not sufficient. We have to learn several technologies, tools, and frameworks that our curriculum doesn’t cover. Further, we have to work on several projects both mandated by the college and self-initiated ones. We cannot rely on our faculty to guide us in these areas as many are as new to the subjects as you are. So the onus of learning falls on us.

Here are the problems I faced with information and knowledge:

1. Collecting information

The way I value information has changed over the last 3 decades. In the early 2000s I found a lot of value in “hoarding” information. Although the Internet was around, it was not as pervasive as it is now. We would have to go to Internet cafes to access the Internet, pay an hourly fee, search for the right information, download it, split it to fit them into 1.44MB floppy disks, bring them home and then hope that they would reassemble once we get back home.

2. Handling the never ending flood of incoming information

As we learn, we discover new things - new concepts, new tools, new techniques. Some of them are directly relevant to our interests, some are not. This is a never ending process. How do I stop myself from the distraction of new information and put my existing knowledge to use? Where do I draw the line between learning and doing?

3. Digesting information

How does a 600 page book of information become knowledge as we read and assimilate it? How does it change from information to knowledge? What makes one book easier to digest than another? Can this conversion be accelerated? How do I retain more of what I read?

There is quite a lot to learn in quite a short time

4. Organizing information

As I started collecting information that I thought was useful, I realized that it is important to also organize it well. But how do I know what is a good way to organize information when I am new to the subject myself?

The Wonder that was Web 2.0

Year 2005, and I am close to graduating from engineering. I secured an internship in IBM India Software Labs. I had better access to the Internet. And this was the time when the web was undergoing a major revolution - Web 2.0!

What was the big deal?

This article from Paul Graham comes to mind.

People were teasing with ideas like these, that will make you instantly fall in love with that era:

Building web “applications” instead of just web “pages”

These applications mimicked desktop applications in terms of interactivity and responsiveness. The main benefit of this is the ability to access the application any time, from any where without having to “install” the application in every system.

User-generated content (blogging, wikis etc)

Anyone could now create content and share it with the world. The web is no longer about just consuming content written by a few experts.

Connecting with people from just about anywhere in the world, sharing information, collaborating on projects was much easier now!

Tagging, folksonomies

“Crowdsourcing” knowledge organization using simple keywords or phrases, and then enabling search and discovery over this.

Feeds

Ability to publish and subscribe to information sources of our choice and getting updates delivered to our “readers”, so that we don’t have to visit each of these sites, rather we can assemble everything in one application!

Mashups, Read/Write Web, Programmable Web and APIs

This was perhaps the coolest thing about Web 2.0. The web is now a platform for building applications that can interact with each other.

Yet to be completed Here is a dump of the rest of the ideas that I have been thinking about (yet to be expanded):

From Information to Knowledge

There is information out there. This has to first come into our radar. But there are a few issues here:

Not every information that we are interested in comes into our radar.
Not every information that comes into our radar is relevant.
There is a lot of relevant information that comes into our radar and this is only getting worser over time.
The quality of the relevant information that comes into our radar is not the same - some have the depth, some are very superficial. Some are terse, some difficult to understand, some extremely easy to digest.

Forms of information

People publish information in various forms:

Text content (eg: blogs, communities, shorter bits like posts in X etc).
Audio and Video content

Sources of information

How did we come across these pieces of information?

Social media
RSS Feeds
Other recommendations

Let us analyze each of them individually.

Here, I am mostly talking about online social networks and not offline (in-person conversations). Social networks were one of the celebrated features of Web 2.0. For the first time, it was possible to connect with like minded people from anywhere in the world and interact with them. The way of interaction was much quicker than having to send emails.

When people published information, they would tag it. These tags would help people discover each other. If I am interested in “machine-learning” and follow posts related to this, I can then look at people who post messages or blogs with these tags and if I find them to be relevant, I can start following them.

Social networks used to be quite open back in the Web 2.0 days. The APIs allowed us to do just about anything. We could fetch real time information, post information and provide integrations within our applications. It is this openness that made it possible for someone to build an interface like TweetDeck. It was this openness that allowed me to build an app called “Twitter Trending Topics” that is similar to the “Trending” tab of today’s X.

TODO: Mention Flickr, del.icio.us, Blogger, YouTube, digg, Technorati, Bloglines, FeedBurner, Wikipedia, Google+, Google Wave, Flock Browser, the ability to build apps using Facebook’s social network.

Social networks then started providing automatic recommendations based on our reading patterns. This was a really cool feature initially, because we didn’t need to follow “tags” or try to discover people ourselves, but the social networks could recommend posts and people based on our interests.

However, social networks saw this as a massive opportunity to monetize their user bases. Why recommend a post about machine learning when they know that there are higher chances that you will get hooked to some rage bait or political news? So, even if you have never seen anything related to these, social networks started pushing information that was completely irrelevant to our (implicitly stated) interests. This has led to such enshittification that I don’t trust most social networks anymore.

While X in the last month has been trying to improve its post recommendations in the “For You” tab, let’s not forget that “Boosting” a post is a paid feature. This means that the network is still biased towards whether a user pays when compared to whether the post itself is of value.

The only other networks that still have the old spirit of Web 2.0 are the ones that are part of the Fediverse, sites like Mastodon and BlueSky. I especially love the Fediverse, not only because we are getting our old social network concepts back, but it’s built to be decentralized, which means no single entity controls our network or the algorithm or our data. However, they have still not reached a critical mass so discovering information or people in a long tail of interests is still a dream.

RSS Feeds

For a long time, during the Web 2.0 era, feeds meant RSS Feeds and feed readers were quite popular among techies. Sites used to prominently display the RSS subscribe icon in their sites, and also a badge displaying the total number of readers. Browsers would auto-detect feeds in the page and would provide an option for a one-click subscribe. Having a blog and not having an RSS feed was akin to having a blog and not having an X account mentioned in it. The first generation of “microblogging” sites including Twitter supported RSS feeds. Plus the openness of social networks meant that if I found something useful, I could share it from within my reader at the click of a button!

Social networks redefined what feeds are. Today, feeds are synonymous to social media feeds and people hardly know about RSS feeds. The popularity of RSS feeds started reducing with the rapid proliferation of social networks like Facebook and perhaps one significant event that dealt a massive death blow was Google killing Google Reader in 2013.

I hope that the current generation of techies learn about RSS feeds. While it looks unlikely, I hope that they continue to support RSS feeds. If you intend to host a blog, ensure that it supports feeds. We need it now more than ever.

Other Recommendations

This includes sources like Google Chrome’s Discover, YouTube recommendations etc. It can even be recommendations within specific sites like Medium or Substack.

Organizing discovered information

There is a difference in time between when we discover information and when we want to use it. How do we ensure that we can resurface the right information at the right time?

Should we spend a lot of effort to put the information in the right place so that when we need it we dont’ have to guess where the information is?

But do we have enough time, knowledge and energy to know where to put the information when we capture it? Could the same information be relevant in multiple contexts at multiple points in our future?

So what are the ways in which we can organize information so that it is easy to resurface when needed?

We associate tags/keywords with the information so that we can search for it later
We categorize the information into folders/categories so that we can browse to the right category/folder when needed
We create links between related pieces of information so that when we are looking at one piece of information, we can navigate to related pieces of information
We can associate the information with entities/metadata so that we can filter based on these entities/metadata
We can cluster related pieces of information together so that when we are looking at one piece of information, we can see other related pieces of information
We can index the information so that we can search for it later

Resurfacing new information

What is the purpose of any information at all?

We want to use it as a part of a project that we are working on - use it as a fodder for our thoughts, or to include it in a presentation or a document etc
We want to share it with others

What are the ways we can look for information?

Search - this may include keyword search or semantic search
Tags
Categories/Folders

The Trilogy - Notes, Feeds and Bookmarks

Note Taking Systems

Requirements for a good note taking system

Must be scalable
Must support interlinking and support backlinks

Purpose of Notes

Who is the note for?

Many of them are for personal consumption
But some may be shared with a limited group
Some may be shared publicly

Why I liked Zim Wiki?

The format is text based and can be hand-coded

Recategorization is easy

As my knowledge evolves, so should my organization of the information

Capturing Intent

What I capture in Zim wiki

Information that I generate - my thoughts, my understanding of something, details of the projects that I am working on, tasks, daily/weekly/monthly/yearly reviews etc
Bookmarks

How I use Zim wiki

How I capture bookmarks

How I capture thoughts

Paper or Digital?

TBD

My Thoughts on the Notion Data Model

Things that I find uncomfortable with Notion’s data model:

Why is a Collapsible Heading a different type of block from a regular heading?
- A Heading 1 inside a Collapsible Heading 3
- Export this and your hierarchy is interpreted differently
When a parent note is deleted, its children are automatically deleted. What if there is a link to a child note?
No red links.
A parent must mention the child in its body.
Renaming the link therefore renames the page itself!
So if I have 2 projects each with a sub-page listing all the tasks, I cannot rename the link to something like “Tasks for Project A” and “Tasks for Project B” without renaming the page itself.
Moving a note from one place in the hierarchy to another place will just append it to the end of the note. So if I have a section in the page listing all the children, it won’t be added to that. I will have to manually then move it to that location.
If a parent links to the child, it doesn’t show up as a backlink but an ancestor does.
The child pages are not subpaths of a parent page.
Export of the pages have Notion identifiers in them.
Minor concerns like I want backlinks to be a panel on the side instead of a small link on top of the heading.
The block model causes lot of issues with copy/paste. I have seen that if I create some regular text followed by a block of code and want to select the text and a portion of the code block, it is not possible to do that. I have to select the entire text and the code block. The mobile app has even more issues with this.
Can any block be converted to any other block?
- If I change a block of code to a heading, what does it mean?
- A heading cannot be on multiple lines so the export breaks.
Convert a Page to a Heading and the page disappears!
- All the text and sub-pages are merged into the parent page.
Turning a heading into a page - I would expect the sub-headings and text get transferred along with it. Notion does not maintain this hierarchy consistently.
What is the purpose of an indentation?
- Copy pasting indented text does not retain the indentation unless you use bullet markdown

Bookmarking Systems

What is a bookmark? Why do we bookmark something?

How do we surface a bookmark when needed?

Search v/s tagging v/s categorizing

One tool or many tools?

Why multiple tools

because they have unique abilities (eg: mails can be sent and received via SMTP)
Spreadsheets are great at capturing structured information
Database backed applications are even better for information that adheres to specific schema

Can the tools talk to each other?

How do they talk?

Via services or import/export of data

Is there manual work involved

Are there triggers or auto sync?

The term “microblogging”
Audio and Video
Mobile was still new
Yahoo Pipes
Ideas were important, monetization came later - people tried new ideas because it was novel - build it, get traffic and we’ll see how to monetize it later was the mantra

Facebook’s F8 platform was a dream!

Monetization issues and why many of these ideas failed

Web 3.0 tried something else for monetization with token economy