MIT To Expand Online Learning and Offer Certificates 96
mikejuk writes "MIT has announced an online learning initiative that will offer its courses through a new interactive learning platform that will enable students to participate in simulated labs, interact with professors and other students and earn certificates. Is this just a reaction to the Stanford experiment in running courses complete with exams and informal statements of accomplishment? (The first AI course has just finished and the exam results are in.) If so let's hope it spurs other educational establishments to do the same!"
This will get lecture book publishers crying (Score:4, Insightful)
Book publishers are going to be crying about online learning and courses if they can't get their books required for them. They are already doing all kinds of shady monopoly deals and trying to hinder reselling of books by updating their course material almost every year, resulting in incompatible books for classes. I'm sure that if they cannot get their books forced in other ways, they're going to be doing some suing or forcing schools to shut down these online learning courses.
I'm not sure why people cry so much about RIAA and MPAA when there is such an assholish industry preventing people from learning. That has real results on whole advancement of humankind.
interesting times (Score:5, Insightful)
Since my day job is CS professor, these kinds of things aren't in my personal interest (unless I land a tenured job at MIT, which is unlikely :P), but I think they have considerable merit. CS, compared to other fields, is already a little bit ambivalent about degrees, and you can get some kinds of jobs by having alternate demonstrations of knowledge, like your Github "resume", or track record of participation in open-source projects. But a lot of companies worry that without a degree you'll lack some theoretical knowledge that will eventually bite you in the ass, because you didn't realize that something was a well-studied problem with an off-the-shelf solution you could've pulled out of one of Knuth's books and implemented, instead of rolling your own buggier, worse one (sometimes this is a founded fear, other times not).
But the bar in many cases is not that high. Even when I've looked for people to work with on, say, a machine-learning project, what I want to know is that they're familiar with the basics of statistics, common techniques and gotchas, correct and incorrect methods of data analysis, etc. This is more likely if they have a degree with some statistics and/or ML courses, but I could see a certificate from a respected course of online instruction being enough to convince me of that, if they keep standards up and it's not easy to cheat.
On the learner's side, it's a really interesting space of possibilities for mixing-and-matching your own education. Since these certificates seem to be much finer granularity than degree programs, if they proliferate and maintain quality, you could more realistically do interdisciplinary programs of study while still being able to prove that you mastered specific things.
Re:Somewhat misquoted (Score:5, Insightful)
I'd guess it's going to be somewhat in between "MIT Certification" and "Internet-U Certification". They're trying to walk a line of ensuring that the regular MIT degree programs are differentiated, while still leveraging the MIT name to distinguish the online course from just any random online course.
Re:interesting times (Score:5, Insightful)
In a general sense of an educated society I agree, but I do think these kinds of courses will pose a significant challenge for CS programs outside the top 10. I don't think MIT and Stanford's online course offerings will be purely supplemental education taken in addition to a 4-year CS degree or by people who wouldn't have gotten one anyway, though there will also be some of that. I think they'll to some extent also displace some proportion of traditional CS education. Probably not a lot at first, but to the extent anything makes it easier to get a good tech job via a route other than a traditional 4-year CS degree, which is already possible but not super-easy, I think it may reduce enrollments in 4-year CS programs, especially outside the very top schools.
Put differently, just in supply-and-demand terms, MIT and Stanford professors can now each fulfill a much larger portion of the demand for CS lectures, since they can lecture to students outside their classrooms. Unless the new audiences are 100% new audiences (i.e. they bring new demand for CS lectures in a 1:1 ratio to the demand-for-lectures that they fulfill), it'll reduce demand for lectures from non-MIT/Stanford professors.
Re:Somewhat misquoted (Score:4, Insightful)
Do the MIT courses have any testing or homework? I just completed the Standford ML class, and it was about as much work as a standard college course. I would imagine that a tested class would carry more weight than a certificate stating that you pressed play on n videos.
Of course, I'd like to believe that the class I completed (and others) will mean something on my resume, but the application process is so streamlined these days that without a degree to make it through the initial filters, I'm skeptical that human eyes will ever see it.
Re:This will get lecture book publishers crying (Score:5, Insightful)
Education comes out of the dark ages (Score:5, Insightful)
It's interesting to see how the state of learning has changed in the last 10 years, and the pace of change is accelerating.
Does anyone know why we study the subjects we do in high school? Mostly it's because the subjects are classical - things are studied because it's been that way since ancient times.
Take geometry, for example. It's an important subject, but not nearly as useful to the average person as probability, yet we study one and not the other.
Then there's the mode of teaching, several hundred years old, where the student sits quietly in a seat watching the lecturer write things on a board and explain them.
Newer models have emerged. The Kahn academy still uses the lecturer/blackboard model, but improves it in many ways. The video can be viewed at a time of the student's choosing, parts can be rewound and replayed, and most importantly: the lectures can be improved by redoing them.
The Stanford and MIT online courses are just another example of the changing landscape. The Stanford AI course had lots of technical problems that they were unprepared for - ambiguous English phrasing, uneven level of practice versus test, missing technical explanations, and so on.
Despite the problems, they will get better. Indeed, they will get a lot better even the 2nd time they give the course.
We're apparently watching a competition for "esteem" between the top end universities. The colleges are competing for clarity of presentation, comprehension, and usefulness of the data.
In 10 years or so the traditional university model will be gone. There will be no need to go to college when all the standard subjects can be learned very well online, using methods which have evolved to present the material in the best possible way.
It'll be fun to watch as this evolves over time.
Online education - really? (Score:3, Insightful)
While I agree that expanding the access to education is a great idea, there is no substitute for attending a brick-and-mortar university. Online courses and online lectures are a supplement to learning -- in the same sense that a text book and a lecturer is a supplement to learning. There are certain skills that you will only learn by living on your own; such as learning to balance your social life, classes, managing a schedule, and other activities. However, if you sit in your parent's house (or basement, like most
Re:This will get lecture book publishers crying (Score:3, Insightful)
MIT has a long history (decade) of offering their entire courseware online for free:
http://ocw.mit.edu/about/next-decade/ [mit.edu]