Khan of Khan Academy at TED (video)

I think what could eventually be even more impactful than the videos themselves is the statistics tracking he showed. With that kind of data across thousands or millions of people, you would see patterns of people who had the same problems and questions in the same places, and you could redo the videos to be more clear in those areas and answer those questions preemptively. With a web-scale audience you could do A-B testing experiments and optimize the performance of your teaching material. I think a curriculum optimized in this way could eventually be dramatically better than even the best traditional education has to offer today.

Today teaching is an art; this could turn it into a science.

The Khan Academy is not the future of education. Don't get me wrong, I think they're doing a superb job and I'd encourage people to contribute to their efforts. Khan is obviously a great presenter with a remarkable ability to explain complex concepts using a new medium. However, the future of education is closer to hacker culture than it is about new and better content delivery mechanisms. Let me explain.

First, I totally buy the argument that a YouTube video can be a better method of delivering content than a book or even, as Kahn says, a live teacher. However, learning is much more than that. Learning is about imagining, playing, creating, sharing and reflecting (Resnick 2007, PDF: http://j.mp/gqnc6f).

To say that learning math is about watching a video and drilling on some problem set, is like saying that learning Python is about doing a tutorial and solving some exercises. Hackers know this well. You don't really learn a programming language until you build some project, especially a personally meaningful one. Creating something not only helps you learn the language but also its community, culture, practices, how to ask questions and get feedback on your work.

Think about it, would you rather hire a developer that has finished some tutorials or one that has actually hacked some projects? The same logic applies to most other fields. Artists and scientists know this as well: you learn by doing in the social context of a community.

Unfortunately, schooling has taken the "hacking" away from learning and has focused on improving content delivery, artificial problem-solving and test-taking. Often using technology. I liked when he mentioned that the teachers have more time to mentor students. However, their mentoring is most likely still focused on the artificial and decontextualized approach of traditional schooling.

Initiatives like the Khan Academy will probably play a role in the future but to really improve education the system and the approaches should change. One should look at exemplar learning environments like the hacker communities or even the Brazilian Samba schools. Settings that are "real, socially cohesive and where experts and novices are learning" (Papert 1980, Google Books: http://j.mp/dLNbGG).

In 1971, Ivan Illich outlined an inspirational alternative educational environment where "the child grows up in a world of things, surrounded by people who serve as models for skills and values. He finds peers who challenge him to argue, to compete, to cooperate, and to understand; and if the child is lucky, he is exposed to confrontation or criticism by an experienced elder who really cares."

Illich said that the four resources learners need are "things, models, peers, and elders". In his 1971 vision, Illich said that "the operation of a peer-matching network would be simple. The user would identify himself by name and address and describe the activity for which he sought a peer. A computer would send him back the names and addresses of all those who had inserted the same description. It is amazing that such a simple utility has never been used on a broad scale for publicly valued activity"(http://j.mp/hNUVV8). Sounds familiar? When I read Illich, I think of HN, StackExchange, and the larger ecology of spaces that hackers have created around social learning. My hope is that these and other similar approaches scale up and get adopted more broadly along with resources like the Khan Academy, OCW and others.

I too am curious about this. It didn't look like prezi, but it certainly had zooming and embedded live video and some other neat things.

I think you are missing the point here.

Learning math is not like learning programming. Math has a quantitative evaluation process.

The kind of scenarios you are talking about both have a quantitative and a qualitative evaluation process.

I.e. you can be creative in your programming and the "answers" but not so much in math.

Also I don't think Kahn Academy is in any opposition to Seymore Paperts "Mindstorms". In fact you could imagine the videos being used to teach children to get started before they go on to hack together their little programs.

What they have done is for the first time made a system that quantifies the learning process so that it benefits the student rather than the school.

Pretty significant if you ask me.

For me, the presentation is just better. There's something about 10-12 minute lessons, and obviously the thought that goes into distilling it to that size, that makes it digestible. (I'm really curious how much time goes into making one 10 minute lesson.) It sticks.

One caveat: I've been exposed to this material before in other contexts, so it's more like a phenomenal refresher. I can't know if this is a better way to teach from scratch. But, frankly, it's been long enough, or I was just that bad at it the first time, that it sometimes feels like I never learned it in the first place. This gives me some degree of confidence that it's at least one right way.