Last Fall Stanford took the initiative to organize live online classes. I took "Introduction of DB" and got a certificate from the professor. Some students complained that Stanford name was not on the certificate. Stanford apparently did not agree to distribute free certificates to online students. When you have a few CS students paying a high price to receive education, it's hard to give an official document for free. coursera.com was created by Stanford teachers who thought they would be better in a startup mode.
What makes MIT initiative unique is that its certificates will be MIT branded ("MITx"). This is a winning strategy, as more people will turn to those classes. Some may continue by applying to MIT itself. It will also help MIT brand reach. Note those certificates will only be free for this pilot course.
You can argue that having a certificate is something that matters or not vs actually learn something. I think a lot of people around the world will feel very proud to receive an official document that proved they know those skills.
Their plan in (near)future is to charge a small fee and conduct tests at authorized testing centres(ETS centres for example) where ones identity can be confirmed. Since that way they can confirm that their is no foul play involved, the student will also be given course credits from MITx(Discussed here: http://web.mit.edu/newsoffice/2011/mitx-faq-1219.html ). That for me is really exciting.
Furthermore, this will annihilate the monopoly that sub-par institutions currently have on conferring degrees and other certifications. Students who have MIT level ability will no longer have to settle for second best. They can prove themselves at the top institutions without having to worry about the 10% acceptance rate.
If executed properly, this will revolutionize post-secondary education and save students and taxpayers billions of dollars. It has long since been proven that the majority of post-secondary institutions are nothing but glorified testing centers. Why not cut out all the excess, and let MIT do the teaching while a test center does the testing?
Also, this will dismantle the broken R&D incentive structure that is publicly funded academic research via tenured professors. The research scientist will be a dedicated profession unto itself, as will the post-secondary professor. Public funding of scientific research will no longer involve the ridiculous process of journal publication, and will instead focus on delivering results to the taxpayer based upon a research contract (much like privately funded research).
Structural inefficiency in academic research has caused an immeasurable slowdown in scientific progress over the past few decades. With one fell swoop, initiatives such as MITx have the ability to rectify this gross misallocation of resources.
By allowing people to take things at their own pace, MITx will perhaps be more optimal for learning the material, but will not provide the same intense environment in which "hardk0re" MIT students are forged (for better or worse).
You could perhaps try to imitate this by taking a soul-crushing courseload from MITx and having a support network of others doing the same. It also makes a big difference whether your support network aspires to get certified so they can get a comfortable job, or aspires to (or actually does) build brain sensors or self-driving cars or musical Tesla-coil hats that play the Mortal Kombat theme (http://www.youtube.com/watch?v=aEgaI6WouQ0).
The reality is that employers can choose to be picky within the boundaries of the marketplace. While I do believe MITx is revolutionary in that it is pursuing a global, open-access education available to anyone with a high-quality brand behind it to boot, I do not think it will change much in terms of employers and hiring.
Employers will continue to hire MIT/Stanford/Harvard graduates because they simply want "the best and brightest." I do not see MITx certificates replacing this recruitment pipeline. MITx credentials are probably more appropriately regarded as something of a wildcard -- an extra edge for a job applicant over more people without evidence of continuing education or certifications.
Since MITx is not in the business of conferring full-blown degrees (yet?), the "sub-par" institutions can rest assured that their students are not going anywhere. And even if MITx did offer full degrees, as does Harvard Extension School, there is the whole other issue of how the marketplace values prestige and brand.
At my local university a few grad students are trying this model: Everyone in a small group is taking either nlp-class,ml-class,or pgm-class. We give the others in the group an overview of what we're learning. To make this worthy of grad credits we're also reading and presenting papers during the semester, and each coming up with a project based on the material which should aim to yield publishable results.
It's a little unfortunate that coursera is taking their time right now, but so far it looks to be promising, and hopefully we'll get a few paper out of it in the end.
Instead of offering separate certificates for courses, why don't they make this an open platform? Say I might be registered at a community college, and instead of choosing to register for "CSXX - Artificial Intelligence" I register for ai-class.com. ai-class/udacity/mitx validates my creds against my university's database and whatever grades I get in quizs, mid terms and finals are pushed from ai-class/udacity/mitx-s to my school's transcript generation system. If ai-class/udacity/mitx-s give an API - kids who maintain schools website should be able to integrate this within a semester.
How this'll benefit the current status quo:
For the students:
Not everybody can go to Standford, MIT for various reasons - but to have a number of courses in your transcript from such prestigious institutions will open doors for him - not only intellectually but also professionally. When he registers for courses, at the beginning of the semester - the registration software shows him the current offering which is a Plain Old Lecture center class by an inhouse faculty and an online class from Prof. Ng. He decides whichever works for him - or may be register for both and see which one suits him and drops the other one.
I strongly believe in the phrase "You become average of the 5 people you hang out with" - think if your classmates are as smart as people in hackernews. [Gawd - no curves in that case!]
They'll be disrupted. I mean, right now the challenge a professor faces is only when there is another professor in the same department offering the same course that same semester. So they tend to use the same slides from the book, same projects (this makes me mad) for generations really. The content is stale compared to what's out there in the real world. Obviously, students paying tuition would want value for their money and for many, going to a good school is gateway to a great job - only to find what he learned in class has been deprecated by industry standard. Students seeing more value in an online course would ditch the stale professors. Furthermore, many hotshot professors really don't give a shit about teaching - they really want to do research, but the department wants them to teach as there's no substitute for that course. If they adopt ai-class/udacity/mitx-s, burden on research profs can be lightened a bit. But bottom line, Professors have to be innovators to bring the students to his class. And if you really enjoy teaching like Prof. Walter Lewin - the world will be tuning in and you'll get paid too (read next point).
Oh you feudal system, you need to innovate to survive. Let me show you why this will make your profits soar:
(1) by allowing students the flexibility to choose from you'll invite more clients (er, students)
(2) Even if you can't afford superstar faculty like Prof. Thrun or Norvig, you can still pay udacity to use their course and get them in your portfolio. Think of the savings!
(3) If you have superstar profs - you can do this too! Just allow other universities to use your courses and charge them for that! Then pay a cut to the professors for doing such a wonderful job.
Edit: add MITx and some coursera courses as well... this semester's been gentle so far, why not liven it up?
I would really like to do it and could most likely also take the time to do so but for me there is a possibility of something more important coming up in the future resulting in me having to cut back on the curse. If something like: "If you want to enter the next course/a different one and your record shows that you began a course which you did not finish, then you have to do X/will be given a lower priority etc." applies, I would reconsider enrolling.
Does anyone have any further information on that?
Gerald Sussman is one of the professors (coauthor of SICP and father of Scheme). If the first statement is true, this is quite an opportunity.
I'm kind of sad, however, that they cancelled the graduate-level sequel to SICP that he was going to teach this semester and that I was going to take. I had even paid my tuition of $7696 for the class (MIT ain't cheap!). I'm one of the first victims of MITx! I still think it's a great thing anyway.
I believe portfolios will become increasingly important in differentiatiing candidates in the future.
As I take these courses, my goal will be to build up a strong enough github profile, and project list to get past any online course discrimination.
Amazon notes "Only 17 left in stock--order soon (more on the way)." Such "free" courses will have a very interesting impact on book sales.
ETA: Now 15.
Amazon now states ships in 1 to 3 weeks
24 new from $80.00 / 29 used from $51.49
so it won't be hard to get a copy before coursework starts in March.
This has a lot of value for someone thrown in a third world country with no financial possibility to join a decent University in the developed world.
There are so many offerings, even since I last made a little list of them for myself ( http://gergely.imreh.net/blog/2012/01/adventures-into-online... ) that it's incredible.
Let's get down to knowledge...