This is similar to the optional programming question at ai-class.org. NLP.

Stanford experimented and seemed to have found a right balance for effective online-learning, now MIT is fine tuning OCW. Exciting news.

  Keys needed:

  * quality teaching and teachers
  * highly relevant courses
  * learners in batch
  * online office hour weekly
  * feedback mechanism, sense of connection
  * discussion forums - e.g. reddit

Image MIT's 10-year OCW content.

e.g. Engineering and Computer Science: http://ocw.mit.edu/courses/#electrical-engineering-and-compu...

Awesome.

Having non-assessment quizzes interleaved with the lectures was also fantastic.

Other key needed:

* A way to pay for it.

To a certain extent, a program like this can pay for itself. A school's influence greatly affects the attention (read: grants and contracts) that labs and faculty enjoy basking in. Conversely, competitors theories and techniques are left clawing their way out of obscurity for lucrative funding. Patents and spinoffs make it big business, and all is fair...

Reading also this one now: "My experience as the first employee of a Y Combinator startup"

http://nathanmarz.com/blog/my-experience-as-the-first-employ...

I should have said, 'what is more important is what it tries to deliver', which is different.

But yeah, its UI feel and some layouts take inspiration from 37s more closely.

For me, this app though more than technology implementation choices is an example of business ingenuity.

Good concept, good market aim, running early, clear message and value.

It can stay alive through its analysis skills. Analysis of how to deliver clearly and enjoyably valauable information through good design and clever choices of ideas (not merely technologies).

What is important is what it tries to deliver.

pg is an enlightened man

seem to be.

sim, tokyo