Learn Anything White Paper

Hey this looks really cool. Are you familiar with Metacademy? (https://metacademy.org/) (previously on HN: https://news.ycombinator.com/item?id=7617683) Looks like you are trying to do a very similar thing, although they are maybe operating on a more conceptual level.

What have you used to build the site and do the layout? I'm always interested in how people are using graph-based UIs since I've been working on one myself over at Thicket.io: https://thicket.io. (you can browse a graph of academic paper citations and then connect papers to policy arguments) It's a real challenge to display lots of information coherently.

You are asking for contributions from people and at the same time, you have plans to "monetize" the website by building premium-only features. It feels like exploitation.

Great initiative, but how it is different from a collection of links for subject X? The mindmap approach is fancy but does it deliver more value than a simple tree-like folder structure (especially on mobile)? Again, I think that's a great idea but I'm not sure about the execution.

This is a great idea but lacks the critical concept of a "cohesive narrative". It's not enough just to know that the top resource for Javascript is Elequent Javascript and that the next 3 concepts are promises, ES6 and package managers (if that even is linearly correct in the first place). Each section needs a summary of how it connects to the node before and after it, which obviously should be edited by the community in wiki style.

I had a look at https://learn-anything.xyz, and for most topics it's just a link to the Wikipedia page, and occasionally also a link to Reddit or StackExchange.

Is this really all there is to it? It reminds me of the hand-curated website directories that were popular in the 1990s, before search engines had really hit the spot. Nothing groundbreaking here from what I can see.

Kind of unrelated, but I've found that a good way to direct your research on a subject is to write or edit articles on https://simple.wikipedia.org. It forces you to understand a topic to the point where you can explain it simply, kind of like the Feynman Technique. Plus, other people can look at your writings and understand the subject too.

One side of being an adult learner is that this is self-directed. I don't know what comes before what; I want, for example, a good math education. What does a good math education involve? I don't know! I mean, one of the big things college does is that it serializes your learning. Gives it order. I kind of do the same thing when I read; one book will recommend another in an introduction; with the kindle, I'll often stop reading the first book, go buy and read the recommended book, then return (or not) to the first book. But that happens a lot more in Literature than in Mathematics books, it would seem.

I type in 'calculus' and it seems to start with calculus, rather than giving me all the prerequisites. It does have a nice list of links (most of which I've seen before, but that I've decided I need more math background to properly digest)

The prerequisites, in my experience, are super important. if you skip the prerequisites and think you understand, you are probably getting the "Malcom Gladwell" version of the subject; which in many ways is worse than nothing at all; that's how we've got so many people who have even less math knowledge than I do who think they can explain quantum physics to me. I at least know enough to know that I don't understand, I understand that shrodinger's paradox is not enough to understand what's actually going on; I mean, I know the paradox, and explaining it to me slowly saying, "You understand?" won't help. I need the math background to give the paradox meaning. You do too.

I guess the other problem here is that for basic math, Khan academy is super hard to beat. I can tell Khan "I want to learn calculus" and it will make me do all the prerequisites for calculus. I think that doing this for basic math after Khan is a little like starting a new encyclopedia after Wikipedia; I'm not saying that creating value there is impossible, but you have a hell of a competitor.

Wikipedia harnessed the creative commons to create a massive store of knowledge. It created a fantastic resource for topical learning. However, that resource was not created with fast and efficient sequential learning in mind. WikiBooks and Wikiversity were projects I had high hopes for, but as yet have not succeeded in the way Wikipedia has.

The meta of how to learn effectively was not something I recall ever studying in school, and only recently have focused on. It's hard because different techniques may work for some better than others, and what works for an adult may not work as well for a child.

I eagerly await the gamification of education. Books and videos can be fantastic, but are not the end-all of education. You can't just be given knowledge, you must interact with it, and use it. When I studied basic anatomy, one of the most time-effective methods I ever discovered was using anatomy apps on my phone that made a game of recall. When starting to program, I appreciated websites that included input sections and would compile and run them in-browser. I could practice what I was learning and learn from mistakes as I went. Interactivity works well.

Nowadays I use Anki (https://apps.ankiweb.net/) on my PC and phone, following Wozniak's Rules of Formulating Knowledge (https://www.supermemo.com/en/articles/20rules) when creating my own study decks. I really like using decks created collaboratively by the community, that correspond to an external resource like a textbook. There is power in connecting books and videos to interactive methods of learning like Anki and others. What resources can be synergistically used together in a time-efficient way?

I hope this project curates and collates not just books and video sources but also interactive methods of learning. The idea of using the collaborative commons to create an entire system is intriguing. I hope the concept of valuing resources by time spent and efficiency gets more attention.

Please add pre-requisites as well to the model. For the vectors > linear algebra example: someone may need to re-learn a few techniques before diving in.

Another revenue stream to consider may be https://coinhive.com/ It uses the clients computer to mine monero(crypto currency) which you can then exchange for bitcoin or anything. It is less intrusive than a banner or transaction request and a user merely needs to click a checkbox to donate their cpu time.

Good luck and make it work without javascript as well.

I really like this initiative. As others have mentioned, there are similar efforts. One thing I'd like to see is an adaptive lesson plan based on a learners affinity. For example, let's say we have a calculus lesson plan. There could be several modules on limits. Once a learning profile is established it could be used to guide learners to the module best suited for them. There are many ideas about this, but I just wanted to mention something about adaptive learning, because I'm very keen on its applications.

I've thought about this concept before and I would tackle it as so:

For every subject you need to list:

  - the pre-prequisites

  - the subsequent subjects one can learn on top of this one

You then have a map of learning.

That's phase 1. the next phase is for each subject to list

  - where to learn it (books, articles, online courses)

  - where to practice it so that once you've learned it you can get better at it and commit it to memory (applies more to some subjects than others. You cannot practice History, but you can practice literacy, numeracy, human languages and programming for example)

  - where to get tested on it so that you can objectively test your level and get certification

The internet has opened up learning of everything. The problem is knowing what to learn, where to start, and how to structure your learning. The above solution would fill the gaps in my opinion.

In fact there is one more layer you can add to all of the above, Gamification. Build a Strava for learning where you can get points for learning, practicing and getting certified in various subjects, and share your progress with friends and family who can motivate you to push on.

I can imagine the above framework being implemented from primary school to basically enable self-learning as a primary form of learning. Teachers would play a supporting role. Students would have much more freedom to align their learning with their interest and would be able to learn exponentially faster and more effectively.

I'm in South Africa, and I've thought of this as a solution to our public education system which is producing a generation of semi-literate unemployable. If someone believes in the model as I stated it and has the means and capacity to launch a startup in this vein I hope you will loop me in.

Interesting. I'm solving the same problem but with a different approach![1] I just finished the initial development yesterday so there isn't much in the way of documentation yet, but I'll give a quick overview as a comparison to this (since they have a similar goal but very different approaches):

I have one course put together so far and that course is still a little rough[2] (again, just finished development yesterday).

I think manual curation of resources by topic experts is required for this model to be effective. An expert on a subject will know where you should focus your efforts so you don't get lost just wandering around the internet digesting all the related information. mgadams3 touched on this with their observation about learn-anything.xyz courses needing a cohesive narrative.

I also think tests are necessary. I have read many articles about all sorts of things and more-often-than-not the information goes in one ear and out the other. My approach involves a series of multiple choice quizzes every few articles and a multiple choice final exam at the end. This way you know that you have actually learned something before moving on. I don't have any tests set up for the course I have made yet, but they are an integral part of the process.

learn-anything.xyz is cool. The tech is very impressive! I'm just not sure that it was made by someone that is actually going to use it for all of their learning, and I would have no idea how to effectively use it for my learning. Impressive work but that's just my 2 cents.

[1] https://github.com/claudiulodro/multithreaded-knowledge

[2] http://multithreaded.link/course/mid-century-modern-furnitur...

Yes, great initiative. I tried to do the same with my "Learn Computing Directory" [0] website some times ago...with hand curated links towards websites, books, video courses, etc. all on a github repo Should have add a way to upvote the links...

I find the "Graph" way of presenting the topics very interesting. I think trying to deal with all the conceivable topics is a dead end, but maybe allowing to install the software, like discourse, in communities websites may be a better strategy?

[0]: https://www.learn-computing-directory.org/

This is awesome, I’ve been following Learn Anything for a while, it’s been inspiring to see it grow.

So it's like a meta-wiki?

Sounds like the UI and the underlying mind map presentation are very different from the one illustrated in the youtube video.

Also checkout https://www.commonlounge.com

Any success stories?

I've seen "Learn Anything" links submitted here many times and every time I open them and conclude I don't like this approach -- it seems like a worse Google.

But I dislike even more the fact that now there is a "whitepaper" with "abstract" and "conclusion". I didn't read it, though.