Share your insights: (beta)

I’ve made to share aha! moments. In 3 words, “Twitter meets Wikipedia”.


  1. Writing articles hurts: research, collect thoughts, organize, filter the best, and write. This takes 20+ hours, and most articles languish half-done.

  2. People don’t share “Ah, I get it!” moments on Wikipedia — that’s not its goal. Yet those personal insights are what make ideas click.

My goals:

  • Collect insights as simple as tweets (“Imaginary numbers are regular numbers, rotated”)
  • Categorize them by Wikipedia topic
  • Use voting/comments to select the best insights to develop further (hacker news/reddit/stack overflow)
  • Grow insights organically — an optional “details” section that can expand into an article (Wikipedia)
  • Browse any topic or url and see the best insights for it

Distributed Learning

My friends suggested the term “distributed learning”. While understanding a topic is a personal journey, we can share what really, truly helped us “get” an idea.

For example: there was an excellent article on the Fourier Transform. One of the best parts:

Fourier Transform Explained

He color coded the equation and made it a sentence. Why, oh why doesn’t every wiki math article have this? (Because it’s a reference, not a tutorial).

There are dozens of awesome techniques to help ideas click, and they need to be 1) exposed 2) filtered and 3) applied to topics.


I’m experimenting with ways to share insights. Some ideas:

Hacker-news like homepage to vote up popular insights:

Popular insight view

While reading a page, keep insights on the “margin”:

Insights on the margin

While writing, have live-previews, TeX support, and edit multiple insights at once:

Live previews

The ultimate dream:

  • Small: Anyone can blast out insights without 90% being forgotten in text files (my personal use case)
  • Medium: Insights in-progress can be developed / commented / ranked
  • Big: Visit a topic like “imaginary numbers” and get an organized view of the best insights, diagrams, analogies and examples (not another tutorial walking through the definitions that didn’t work the first time)

This isn’t a replacement for Q&A sites, wikis, or blogs — it’s a repository of insights that helped speed up the learning process from those other sources.


Let’s explore ways to reduce the friction while learning: make it easy to share ideas, and filter the ones that work best.

I’ll be writing as normal, but instead of starting with private notes that wither away, I’ll keep ideas in a public arena. The best insights can be combined and developed into full-on articles.

The project is brand-new, evolving, and feedback is welcome! Feel free to kick the tires and add your own ahas to the mix :).

Questions & Contributions


  1. This is a great idea! Allowing cognitive surplus to take the form of distributed learning like this could really help solve the problem of finding the most intuitive ways to share information.

    Can’t wait to see this in action!

  2. Wow! I’ve been thinking along similar lines, but this was FAR beyond anything I would be able to do! Great initiative!

  3. @Joe: Thanks! I hope to import the best insights from articles & comments and add them to existing ones (so at the bottom of each article, you can vote for what key analogies/explanations helped you the most). I’m really looking forward to using this more.

    @Carln: Thank you! This is an open initiative (all content is creative commons) and I’m definitely open to suggestions / brainstorming. The ultimate goal is to have a clean, simple way to focus on the key parts when learning a topic.

  4. Great Idea!

    I feel excited about this idea and would like to be a part of it, so, go ahead buddy…

  5. I think I will start using this. I do feel like only tagging with wikipedia topics is a bit too restrictive, though!

  6. I am glad someone is willing to put the time and effort into doing something I’ve always wanted to do. I’m so in.

    Having a plugin for chrome where people can write useful insights, links to other tutorials/guides/explanations, attempt better explanation, etc. would be incredibly useful. I also like the idea of having a central site where people can vote on things for people to write about, vote on explanations, possibly even show multiple ways to understand things (different perspectives). The sky is the limit, and you’re definitely breaking new ground for future (and current!) scientists/engineers.

  7. @Harish: Thanks!

    @abramdemski: Great feedback — it can actually allow any topic (shh!), but I think Wikipedia is a good starting point for categorization. Over time we’ll see if new types of topics emerge… but I can’t think of anything that isn’t in Wikipedia already :). There may be meta-categories like Diagram:, Analogy:, etc.

    @Mark: Awesome, thanks for the support. And what a great idea for the Chrome extension! I’d like to make simple bookmarklets/browser plugins to make it easy to browse and contribute.

    As you say, I think there is some magic in seeing different ways to explain things… I’m always reminded of the many ways to add the numbers 1-100 (… that first got me thinking about having different analogies for the same concept.

  8. Hey! Great idea! Can’t wait for the RSS for the service so I can subscribe and use it as a constant source of the insights.

  9. I very much enjoy this idea, so I please hope you figure out a solution to ensure that better explanations rank higher than tricks. Tricks are just fine, but they require no comprehension of how they work, so they would probably be the most popular posts on the site(people looking for homework help with no desire that a better explanation is within their grasp, despite their position)

    I wouldn’t feel comfortable coming up with a comprehensive definition for neither a better explanation(nor trick), so I will defer to the story of Feynman vs the Abacus:

    To put it succinctly, the abacus salesman only has tricks, whereas Feynman has deep understanding and as a supplemental to that, tricks. This allows him to apply those tricks cleverly when he needs to, whereas the abacus salesman is nothing more than a computer blindly applying rules which it cannot comprehend.

    Will betterexplained encourage little Feynmans, or little abacus salesmen?

  10. @none: Great point, thanks for the commet! I love that story, and it captures the essence of truly understanding an idea.

    Currently I’m not sure of the best way to ensure such insights but here are some ideas:

    – offer templates for certain insights (ie every math equation should have a colorized sentence like the fourier transform, above)

    – curation / moderation to highlight the posts that capture the spirit of that real understanding

    – over time, a community-developed sense of what is a “trick” vs a deeper insight may emerge and shape what posts get highlighted

    It’s still in the very early stages but I think with some good examples the insights can be shaped in the right direction :).

  11. Kalid,
    Has anybody submitted a guest post? How to go about it?
    Can I choose a topic myself? How long the article should be?
    What are the do’s&dont’s?

  12. @NANDEESH: Yep, you can create an account on the beta site and write/submit any articles you like :).

    No major do’s/don’ts, but I’d suggest keeping the insights focused on insights/explanations you’ve come across (to keep it consistent with the other articles).

  13. Also, Kalid, where is it that you get this fantastic intuition from? I’ve been reading some of your other articles and the concepts are so well put! Are there any textbooks that you’re drawing your knowledge from in particular, or are you using your undergrad notes. Thanks!

  14. Thanks David! I have many old, half-finished notes from previous classes, and scattered text files, etc. I try to think about a topic I’m interested in, jotting down notes about what’s confusing, then search for analogies which help clarify [some work, some don’t… the ones what work, I try to refine until they can explain more and more of the concept]. I find textbooks, Wikipedia, etc. decent for checking a result found via intuition (is this analogy actually working?), but usually they aren’t good for getting insights directly, as they’re often written in a very dry / rigorous style :).

Your feedback is welcome -- leave a reply!

Your email address will not be published.

LaTeX: $ $e=mc^2$$