Ask HN: Why don't we have a best-of-class open self-learning platform?

7 points by gettodachoppa ↗ HN
About 15 years ago, when some universities started uploading videos of their classes, Khan Academy got popular, and there was news of professors quitting their university jobs to create online classes to reach more people, I thought a new era of learning was upon us.

I honestly thought that by now, a poor kid in a remote village in Africa would have access to a guided education track that could give him the same knowledge as an Ivy League student. The way I imagined it, the student would connect to an open education portal, managed by a community of volunteers like Wikipedia, which would include top professors contributing their knowledge to create the best material, open textbooks to follow, free text/audio presentations/lectures teaching the textbook. It would cover all levels of education, allowing poor and rich alike to study all they want in a portal as ubiquitous as Wikipedia. I also thought if certifications/degrees are needed for self-taught students, they could pay a small fee to do an exam in-person at small centers in their area.

Fast-forward to 2023 and I feel like things haven't really evolved beyond what we already had back then. There's more content, sure. But it's random bits here and there, spread all over the place, and I would imagine 99% of people don't know about 99% of the content available. Khan Academy comes closest as a popular education portal, but it's just Youtube videos, which is only good as an education supplement, not a primary method of education (plus Khan Academy stops at the high school/college freshman level).

Did anyone else expect us to have something better by now?

(Note: I know there is an irreplaceable social aspect to attending school in person so you can socialize with other kids of your age, but that's not incompatible with the future I expected.)

8 comments

[ 3.4 ms ] story [ 32.1 ms ] thread
Sigh

This is one area where I feel like programs like ChatGPT can be helpful in the future. My first experience with ChatGPT was generating a few mock syllabi (recommended textbooks and week-by-week course outlines) and it turned out to be pretty good. This is just a low-end example, but there are probably a lot of other ways that large language models can be helpful in this field. One of the greatest concerns to self-learning is probably self-management.

Your second paragraph describes a beautiful concept that I think takes a tremendous amount of effort which could probably be offloaded to AI. I can’t speak to the quality of this outcome, but in a best(best,best)-case scenario I think it would be splendid. There would still have to be the right amount of human involvement though.

People have different amounts of motivation, different things that motivate them, make different mistakes given the same material...
> it's just Youtube videos

No it's not. Khan A. has a huge structure associated with it. There's tracts. It's constantly growing and improving. There's assisting other students. Maybe it's not as intimate as a class that's paid for. And, I can't help but respond angrily, you've denigrated quite a stunning institution. Maybe sign up at Khan in a new subject and see how much of it isn't just videos. Sheesh, man, the nerve.

>Did anyone else expect us to have something better by now?

Yes, and we have it. Its called general internet. You can pretty much learn advanced topics with the source material online. ChatGPT has made it easier to research. Even back in early 2010s when I was finishing my masters, I would often just google things online and find random webpages that explain the concepts better than professors could.

The biggest issue with learning is motivation, which you cannot put into people. If someone is motivated to do something, he/she will learn. Handholding someone doesn't really do anything in the long run.

Searching the internet for decent material is a pain point. Wikipedia gets too complex, stackoverflow like forums don't exist for all fields, books are not freely available. Great courses from standard/mit are also not always available. Many other courses are behind paywalls. We can certainly aspire to better the status quo for open sourcing education at least to a basic graduation level of all major subjects.
>Wikipedia gets too complex

And yet again, this proves my point.

If wikipedia is too complex, that means you are not motivated to actually learn. The way to learn is experimentation. Modern day computers are great for this, from computer science to physics. If you read a wikipedia article and have a lot of questions, you re-read it, stop at the first thing that doesn't make sense, then go research that, or take something like a program, or a formula, implement it on the computer, and start trying it out.

What people want is handholding, i.e they see certain words, hear certain sounds, perform certain predefined set of actions, and somehow magically gain the knowledge. Thats not how learning actually works.

>I honestly thought that by now, a poor kid in a remote village in Africa would have access to a guided education track that could give him the same knowledge as an Ivy League student

Khan Academy "courses" + MIT OCW gets anyone in the world from pre-elementary through an undergraduate engineering degree at a school arguably superior to the Ivy League.

Content and courses are already there but that's not enough. The missing important pieces are : 1. certification & credibility: In an ideal world , this should not be important. But in reality, this is the number one reason people go to schools. Unless Khan academy or similar org can give credibility, schools will still be important.

2. peers : if you're an adult, it might work for you by learning alone. But for teenagers or even adults, the experience of learning togeher with your peers is irreplaceble.

3. motivation : it takes some superhuman to be able to become self motivated to consistently keep learning day in and day out without someone else (teacher) monitoring you.