Khan Academy.

[drego642]

About a year or so ago, I started watching some lecture videos on physics and astronomy by Salmon Khan and found them really informative and interesting. With those videos, I was able to not just learn more, but also get a better understanding of the subjects than I had in any school course I'd ever taken. It's not all science and math either; right now, I'm about to watch one of his videos on the healthcare system - one of many current event issues of which I have a regrettable lack of understanding.

Last year, Khan gave a speech at TED about his approach to teaching, followed by a short, live interview with Bill Gates. Khan, along with his development team, has created a virtual school using lecture videos as assignments and tutoring by teachers in the classroom, which inverts the traditional classroom dynamic. I was wondering if anyone else has ever watched his lecture videos, seen his channel or even heard of him or his ideas for improving the modern education system. What are everyone else's opinions?

Here are some links:

The Khan Academy website
Khan Academy YouTube channel's YouTube channel containing the library of lecture videos
Khan's TED speech and subsequent interview.

[Bakemaster]

Khan Academy is quite popular. There are more traditional lectures available from recordings on sites like http://www.academicearth.org and an increasing number of universities are providing "open" online course materials and/or videos. See MIT OCW, Open Yale Courses and Stanford Engineering Everywhere for examples.

At the community college where I used to work (and take classes), Khan Academy was very well-known. A decent number of math instructors would recommend it to their students who might be struggling with a particular concept. Its math content is certainly more high-profile than the rest of its catalog.

It's impossible to say exactly what impact KA is going to have in the long run on the education system, but it certainly seems as though it's going to be a large one... eventually. We may have to wait for a lot of educators and administrators to age out of the system (at least, here in the US) before programs like KA can really take off. There's a lot of resistance to change in the system.

Personally, I really liked Sal's videos explaining the financial bailouts. I've also watched a few calculus videos, I think, from when I was taking multivariate and ODEs. I would highly recommend Stanford Engineering Everywhere's CS106A for anyone who has dabbled in coding and is wondering about formal study.

[Zcorp]

Sal has done a great service to the world of education. That said, he got extremely lucky and it is very unfortunate he decided not to make Khan academy a distribution platform where anyone, or a vetted group of people, could submit videos, track the related data and build tools to assist in furthering that understanding.

His great service is making the idea of online learning a common one to teachers, students, school administrators and parents. Additionally his videos and website are good enough to greatly help a lot of people as he is better than average teachers.

Khan Academy and online learning in general could be much greater than it is. If he had spent the last year or two building tools to make it a distribution platform for all teachers and gave incentives to contribute to the project he could of rebuilt the entire education system from the ground up. Tracking personality, aspects of intelligence, helping teachers build tools to recognize and then cater to these differences and then of course work toward building a sane accreditation process for students.


We should be building an interactive and scaffolded wikipedia.

But hey its a start, would just be nice of Gates and Google decided to put money into education besides popular people and trends.


Also check out Udcaity and Ted Ed in addition to Bakemasters' links.

[Bakemaster]

Khan Academy and online learning in general could be much greater than it is. If he had spent the last year or two building tools to make it a distribution platform for all teachers and gave incentives to contribute to the project he could of rebuilt the entire education system from the ground up. Tracking personality, aspects of intelligence, helping teachers build tools to recognize and then cater to these differences and then of course work toward building a sane accreditation process for students.

I can certainly see how, looking at the drastic and sudden changes technology has brought to many aspects of the world, you might say that Khan Academy should be doing more. However, history will readily demonstrate that not all of those sudden changes have been for the better; or perhaps I should say, many of those really important and beneficial changes have occurred so rapidly that they brought along with them a lot of really big problems we're now struggling to repair (e.g., toxic and GHG emissions from combustion engines). Ounce of prevention, pound of cure, etc.

I think there's a great deal of merit in someone like Sal (who is fairly young and has practically no experience with our present system from the instructional and administrative side of things), with a project like KA, focusing down on some significant incremental change rather than opening the revolutionary floodgates and trying to start over from square one, in a system that can't stop running to restructure itself and is notoriously resistant to large-scale reforms. It means a great deal in terms of the extent to which he can observe and control what happens, and that's extremely important both from a scientific point of view and from an entrepreneurial point of view. You don't necessarily want to be taking gobs and gobs of someone's money, and some portion of their faith, and scatter it to the wind in hopes that your innovation will catch on organically.

And youtube still exists; other open video hosts and aggregators exist. Frameworks for true crowd-sourced education efforts are already out there, whether you want them to be completely open or have some kind of filtering/certification process, it's totally do-able. I don't think that's really what Khan Academy has ever tried to do. From what I've seen of interviews with Sal and coverage of the program, he's more focused on challenging the status quo in terms of how and when students demonstrate mastery of material and move on to the next topic, and how and when students do classwork and receive instruction.

[Zcorp]

I can certainly see how, looking at the drastic and sudden changes technology has brought to many aspects of the world, you might say that Khan Academy should be doing more.

I'm really saying that the world of online learning should be doing more, he just happened to be the one lucky enough to get early funding and take off.

And youtube still exists; other open video hosts and aggregators exist. Frameworks for true crowd-sourced education efforts are already out there, whether you want them to be completely open or have some kind of filtering/certification process, it's totally do-able. I don't think that's really what Khan Academy has ever tried to do. From what I've seen of interviews with Sal and coverage of the program, he's more focused on challenging the status quo in terms of how and when students demonstrate mastery of material and move on to the next topic, and how and when students do classwork and receive instruction.

All of this is true. My criticism is that he could be doing something bigger and chose not to. He is in a entirely unique position right now to create this change. He has funding and brand recognition to an extent that no one else does.

Building something like I'm speaking of is inevitable, and it will be superior to KA, we could just be there or nearly there already if he had decided to go this route.

TED Ed might turn into this, they are years behind and aren't sure what they are yet. There are trying to be educational but haven't actually built a system with a goal or cohesion, just random smattering of ideas. A "How things work" in animated video really.

[Catmando]

Sal has done a great service to the world of education. That said, he got extremely lucky and it is very unfortunate he decided not to make Khan academy a distribution platform where anyone, or a vetted group of people, could submit videos, track the related data and build tools to assist in furthering that understanding.

I think that at this point it really is a whole team of them, a "vetted group of people," who do the videos. Sal handles math, finance, science etc., but Drs. Harris and Zucker have art history videos featured on there, Vi Hart's "fun math" videos are there, and Brit Cruise has cryptography videos and, soon, information theory lessons too. It's not as nearly open-ended as you probably wish it was, but I think it's doing what it's doing well.

[Bakemaster]

Building something like I'm speaking of is inevitable, and it will be superior to KA, we could just be there or nearly there already if he had decided to go this route.

It may be inevitable, but I don't think he could have done it in just a few years, even with the funding and exposure he's had.

[Zcorp]

It's not as nearly open-ended as you probably wish it was, but I think it's doing what it's doing well.

It also isn't as open datawise and the entire system isn't built by people who understand educational systems. But completely agree that it is doing what it is doing well or at least leaps and bounds better than anyone else at the moment. We just could be doing better.

It may be inevitable, but I don't think he could have done it in just a few years, even with the funding and exposure he's had.

He could of built a framework in a couple years that would then be continually improved upon. It would likely have less content today if he had done so but once a functioning infrastructure was in place it would grow incredibly quickly and be far more powerful.


While I'm criticizing him for what was a missed opportunity let me state that it is still the best option for individual learning and other applications out there for k-12 education and possibly more. His math area specifically is a great start and example of things being done well, breaking stuff down into various core concepts and focusing on them individually.

Investors reward action and while something like this KA isn't a new idea and various superior ideas have been proposed in education circles no one took action to create them.

Use KA and tell other people to do so, I certainly tell the teachers I work with to use it. However, KA is only a mediocre beginning of what online educational tools will be.

[Tirian]

Khan Academy and online learning in general could be much greater than it is. If he had spent the last year or two building tools to make it a distribution platform for all teachers and gave incentives to contribute to the project he could of rebuilt the entire education system from the ground up. Tracking personality, aspects of intelligence, helping teachers build tools to recognize and then cater to these differences and then of course work toward building a sane accreditation process for students.

I can understand both why he didn't want to do that and why that probably wouldn't work out as nicely as you think. Webquests, for instance, are a useful concept that are available for any teacher and student to contribute to, and it's a hot damn mess over there. I'm sure that there are a thousand great projects there, but they're buried under a million crappy projects because they decided against a meritocracy. There are no barriers to entry with independent learning -- if you want to join the Khan Academy faculty, then make a playlist of fifty videos covering a topic and post them to YouTube. If you're good, you'll gain a following and word will filter up to Sal and he'll decide whether to adopt you. Even if he doesn't, you still have a popular YouTube channel.

And I can't envision KA ever becoming formally accredited. There's just no way that Sal is ever going to want the infrastructure involved in that. Giving course credit and degrees to independent learners will be the domain of your local public colleges, so lean on them. That's not to say that Sal is a dilettante, because providing a suite of intuition-oriented unscripted lectures with an unapologetic rejection of showmanship fills an important need in independent learning.

[Zcorp]

I can understand both why he didn't want to do that and why that probably wouldn't work out as nicely as you think. Webquests, for instance, are a useful concept that are available for any teacher and student to contribute to, and it's a hot damn mess over there. I'm sure that there are a thousand great projects there, but they're buried under a million crappy projects because they decided against a meritocracy.

No need to straw-man me. There should be a vetting process, as mentioned, for delivered material through lots of different means. Just to start with there could also be rankings from KA themselves, other students that share your personality and intelligence characteristics, your friends, your teachers, objective measurements based on how other people like you found a lesson to be useful and assisted them in gaining proficiency in a concept. As more material becomes available having easy ways to sort through it is an obvious aspect of making it functional.

Additionally opening up math classes for more people to contribute will result in a superior and more diverse classes than Sal's. Teachers will create general lessons as well as niche lessons based on demands of type of intelligence, learning styles and interests of students. Some approaches to learning disengage some students while maintain or increase the interests of others. A simple example being some small changes in rhetoric, do you teach the concept first or the term? A cartesian coordinate system is ... vs ... is what we call the cartesian coordinate system. Small changes in approach like this can have a rather significant impact on student learning.

Webquests has problems far more severe than there open aspects. The largest thing hurting them is there terrible terrible web design and UI design.

There are no barriers to entry with independent learning -- if you want to join the Khan Academy faculty, then make a playlist of fifty videos covering a topic and post them to YouTube. If you're good, you'll gain a following and word will filter up to Sal and he'll decide whether to adopt you. Even if he doesn't, you still have a popular YouTube channel.

That isn't my goal I'm also not a classroom teacher and it is entirely besides the point that I'm making. Also take note that I've continually praised the site even if I also express criticism for some of their choices. I'd love to have the funds and access to talent to build what online learning will become.

And I can't envision KA ever becoming formally accredited. There's just no way that Sal is ever going to want the infrastructure involved in that. Giving course credit and degrees to independent learners will be the domain of your local public colleges, so lean on them. That's not to say that Sal is a dilettante, because providing a suite of intuition-oriented unscripted lectures with an unapologetic rejection of showmanship fills an important need in independent learning.

Missing the point again and I could of been a bit more clear myself. They should build their own accreditation process for students. What we have now is corrupt and bad. There is great opportunity to challenge what exists now, and if they do a good enough job at actually measuring student understanding (something we are really poor at right now) it will become the new standard. Then there are all sorts of really interesting experiments to be doing locally and cross culturally improving educational standards and pushing for greater human learning potential.

[Adacore]

I agree with a lot of what you're saying, Zcorp, but I don't think Khan Academy would've got the same publicity and support if he'd focused on developing a more solid infrastructure with user-contributed lessons and stuff first. A large part of the 'wow' factor for all the media pieces I've seen on it have been about this one, solitary guy doing awesome things for education and humanity. The media (and I guess by extension consumers of media) love a hero, and without that it might not have taken off. Plus, if he'd been working on a better infrastructure instead of content, there obviously wouldn't have been the same draw content-wise.

At this point, though, now that the site is known, I think it would be sensible to allow submissions from others, or at least (if he wants to run it this way), to recruit some other tutors to make videos.

[Zcorp]

I agree with a lot of what you're saying, Zcorp, but I don't think Khan Academy would've got the same publicity and support if he'd focused on developing a more solid infrastructure with user-contributed lessons and stuff first.

Maybe, but he got most of his funding a while back from google and gates. I could be wrong but I don't think he has gotten much of his funding recently. So there was less need for that recent publicity to raise funds. If it took 2 more years to create something of greater value and get publicity for it then it could very well of been worth it.

A large part of the 'wow' factor for all the media pieces I've seen on it have been about this one, solitary guy doing awesome things for education and humanity. The media (and I guess by extension consumers of media) love a hero, and without that it might not have taken off. Plus, if he'd been working on a better infrastructure instead of content, there obviously wouldn't have been the same draw content-wise.

Quite possibly. I like to hope we are better than that and would value the creation of the infrastructure from 'one guy' as much as we value the content.

At this point, though, now that the site is known, I think it would be sensible to allow submissions from others, or at least (if he wants to run it this way), to recruit some other tutors to make videos.

I'm skeptical of the utility of this without a proper data mining infrastructure set up. But it might be in development now. He's had a lot of open positions recently, maybe this is something they are working on.

[Zcorp]

Bit of a double post but figured it would be worthwhile to provide this article.

Khan Academy: The hype and the reality

I'm not posting this because I agree with it entirely, just to add to the discussion. There is much Khan Academy could be instead of what it is, if Sal had taken a different approach to spending his money and engaged people who have been trying to solve education problems for while.

[doogly]

I haven't looked into it too too much myself, but I definitely recommend it to my students looking for extra specific algorithmic help. Seems to be pretty good, and they tend to enjoy it.

I tell them to check it out as an alternative to their textbook, which is just as step-by-step, algorithmic and spoon feedy (ugggggh.) Certainly not an alternative to lectures with an actual math person, with some give and take, exploration, activity, etc.

[Tirian]

He is an "actual math person". He's got a BS in Math from MIT, along with a BS/MS in EECS from MIT and a Harvard MBA. What he lacks is formal training in pedagogy, but his philosophy is that a casual approach is more accessible to struggling students than a shiny formal program.

Sal has argued that there is some place for video lectures in classes. A student can watch a lecture and take notes at his or her own pace, without having to embarrass oneself by asking the professor to go over something again in just replaying the video would suffice. And once that's happened, it frees up the classroom time for the inquiry-based dialogue and activities that form the largest part of driving individual comprehension. I think it's fair to ask whether KA is better than a good textbook or someone else's video lectures. Personally, I think it's hit and miss. I love the statistics and core finance playlists, and I'm less enthusiastic about differential equations, linear algebra, and trigonometry, and haven't looked closely enough at calculus to say one way or the other. I would say, as you seem to, that for an arbitrary student and an arbitrary subject that it's worth investigating for twenty minutes to see if it's a good fit.

[eSOANEM]

Personally, I think it's hit and miss. I love the statistics and core finance playlists, and I'm less enthusiastic about differential equations, linear algebra, and trigonometry, and haven't looked closely enough at calculus to say one way or the other. I would say, as you seem to, that for an arbitrary student and an arbitrary subject that it's worth investigating for twenty minutes to see if it's a good fit.

I've had the opposite experience. I didn't like the physics videos much but thought the DE ones were excellent, likewise the calculus ones I watched.

I think KA is definitely a good thing to try if normal classroom methods aren't working, if you only need a bit of polish or if you want a little bit of stretching. I do not think it would work so well as a standalone resource but then, I don't think this is quite what Sal is aiming for.

[Tirian]

I've had the opposite experience. I didn't like the physics videos much but thought the DE ones were excellent, likewise the calculus ones I watched.

To each his own. Except that I'll insist that it's a big deal that Sal spends the whole DE playlist pronouncing "homogeneous" as if the word was "homogenous".

[eSOANEM]

Homogenous is usually listed as an alternative form of homogeneous and, in my part of the UK at least, is more common. So him talking about homogenous equations rather than homogeneous equations was one of the things I liked because to me the word "homogeneous" always sounds a bit weird and foreign.

[Bakemaster]

-genous, root "gene" like genetics, from mid-1800s
-geneous, root is something Latin, from some prior century
They don't mean the same thing but although using them interchangeably is technically incorrect, it rarely results in any mis-communication. In the end only chemists and biologists, or students or teachers of same, really give a damn.

[eSOANEM]

Both come from Greek actually, from γενεα (for homogenous via gene) and γενος (for homogeneous), both clearly from the ultimate root γεν.

Furthermore, "homogenous" has two definitions listed on wiktionary, one is a biological one (fair enough if in some technical contexts the two are distinct), the other is an alternative form of "homogeneous" (it notes that "some educators or other authorities recommend against" this usage). It lists no other definitions and so, outside of biology, whilst some people may prefer one, the two are not distinct and the only interpretation of the disfavoured one is as "homogeneous".

Not only this, but the Bloomsbury English dictionary (which is usually more precise and accurate in technical stuff than the OED in my experience) lists "homogenous" only as an alternative spelling of "homogeneous", without even a note about it being proscribed. The OED does likewise, without a note about any technical ambiguity either.

Most interesting though, is that it notes that the pronunciation /həˈmɒdʒɪnəs/ (as of homogenous) is more common than the spelling homogenous implying that many people pronounce it as if it were "homogenous" but spell it "homogeneous". It is possible Sal is one of these people.

Anyway, the point is, in the DEs (and their most common applications, physics and engineering), the two spellings are not distinct so to say that, in that context, it is technically incorrect to use one for the other is an unnecessary and arbitrary proscription and so a thoroughly unscientific notion.

(a note, you're perfectly free to say it annoys you that he uses one form rather than the other, but, in this context at least, if anyone claims he is wrong to do so then that I am not ok with)

[Bakemaster]

Hey, all I know is that when I Googled it, I saw Latin before Greek. I know practically nothing of either language, myself.

Unless Tirian says otherwise I'd imagine the comment was tongue-in-cheek.

[Tirian]

It actually did disturb me on two points, although this conversation is making me reassess.

- The first point is that I thought it was intellectually lazy to (mostly) write it one way and pronounce it the other. Evidently, though, that isn't specifically Sal's fault. I need to be upset at society for not looking at words carefully enough to pronounce all the syllables and the prescription vs. description debate that endorses swallowing those syllables if a majority want to. Don't worry, I was already upset with them anyways.

- The other was that the one time Sal brought this up in a video he confessed that he had no idea what the connection was between a class of differential equations and milk that both should be homogenous. Again, this struck me as lazy, since the way I worked it out in my mind (probably with Webster's Tenth by my side) is that milk is homogenous (uniform structure) while linear ODEs with constant coefficients were homogeneous (same structure -- as the polynomial whose roots you need to compute). Well, the unabridged and technical dictionaries are convincing me that English is not cooperating with my belief that these are two concepts that are important enough to deserve their own separate long words.

Oh well, so much for a good intellectual superiority rant.

[doogly]

He is an "actual math person".

He considers* math to be about following algorithms that a math person has told you to do. This is not math.

*or at least, he teaches math by pretending to think this way. Maybe he knows better but things the way to make math "accessible" is to turn it into step by step things for the small minds to digest. Whatever, it's wrong.

[Tirian]

How would you teach the true math through videos?

[Bakemaster]

No actual math person takes his porridge with sugar, apparently.

[doogly]

If you don't like "treats math as the following of algorithms" as a reasonable or good criteria for "actual math person," we can go with the more exacting standard of "produced original math."

[w.eckhardt]

I am currently working on a project similar to Khan Academy (only in Dutch)

I think the concept of KA is great, but also lacks in some places.
Since we only have 3 brands of textbooks for physics and science, I try to couple all of my video's to the textbooks.
Also, I am open to other people creating video's. If I stumble upon something really good, I include it.

I see KA as an perfect supplement to the current programs, but I don't see it as a replacement.
At least, you know you get a semi-consistent quality in the video's and everything is easy to find.

The whole idea of Webquests and Hot Potato lessons suffer from the overkill on material on the internet: You just don't know the quality of the product... That's the main reason why I refuse to use them in my classroom. It's just too much work to find something really useful, instead of something someone created when drunk on saturdaynight 03.00...

But that's why I am building something similar in Dutch, and hope to have the search-engine, book-coupling and subject/year-coupling done by christmas.

[doogly]

Interest article on how KA does programming.
http://worrydream.com/LearnableProgramming/

[Tounx]

How much better the world would be if more places like this existed.

[Whizzkid1024]

Love his lessons on Laplace transformations, really intuitive, like a lot of the math videos.