They will be free (or very very cheap) for ever - it's a great business model
There is no threat to MIT/Stanford/etc from running free courses. Even IF somebody decided = I could afford to go there, I could get accepted, I'm not planning to do a startup - but I have decided to only do the online courses instead of attending - their place would immediately be taken by somebody else.
The real impact of these courses is that they destroy the market for 3rd rate universities. Why did you waste your time going to NoWheresVille U when you could have done the MIT courses and something relevant to me as an employer?
A very similar thing happened 10years in the UK. All the community college equivalents were allowed to call themselves universities. All the 'IVY league' were promoting this as 'access for all'.
Of course what it meant was that employers weren't sure if somewhere they hadn't heard of was a decent 2nd tier university or some degree mill - and so to cover themselves began demanding IVY league degrees only.
Since Udacity is not exactly forthcoming on exactly what their business model will be, could you add some insight to exactly what the great business model is? Right now, I see the MITx and Stanford programs being a small loss per student compared to the normal amount universities spend on instruction, but a small loss a hundred thousand times over doesn't exactly turn a profit.
If you are going the freemium route (free content, pay for testing/credentialing) you have the classic freemium dilemma, what is your conversion percentage from free to paid and what is your margin on the paid version. If you anchor your product pricing at zero, you can't exactly charge hundreds or thousands for testing. Plus, the vast majority of students in the Stanford experiment were overseas which makes the paid option at a recognized testing center much more difficult.
It's all a percentage game. Revenue = Total Students * percent who complete the class * percent who can pay for testing * percent who do pay for testing * margin on testing. We ran the numbers in a number of different number of scenarios when we looked at going that exact same route and found time and time again that even if you can break even at a small number of very popular classes, the chances of being able to build out a broad curriculum of classes with less mass appeal was not good.
To use some realistic numbers, let's take the Intro to Databases class and make up some inputs:
Number of students: 90,000
Percent who completed the class: 7.2%
Percent who could take tests at approved centers: est 40%
Percent conversion from free to paid: Classic freemium less than 5%. Hugely optimistic: 20%
So, we have a probable range of $1,300 - $26,000 based on some real numbers and some educated guesses. $26,000 may sound like a lot, but that amount can easily be absorbed just in the professors time prepping and managing the class. Now add on the rest of the overhead of high paid SV engineers, offices, support, etc...
The end result, an enormous amount of things have to go just right in order to cover your base costs. Even if those things go just right, you still have a huge amount of pressure to create classes with the broadest possible appeal with the highest possible number of students completing the course (in order to have a chance of testing revenue). Those market pressures do not coexist well with academic integrity and high standards. They also don't support classes that might "only" have a few thousand students.
There other ways to get to free education for all (we're on that path ourselves), but starting out as free without the backing of a large university is a tough proposition.
The business model of Ivy league free courses is protect the market, destroy competition, preserve monopoly.
At the moment Uni education is a pretty open market, there is Ivy league, there are top state colleges, there are 3rd tier and then the Phoenix type places - you pay your money and take your choice.
The real tough competition, at least in technical subjects, is between ivy league and top state schools. The students can save a lot of money by going to UCB over Stanford, learn exactly the same stuff and get exactly the same job - why pay the extra $100k for a Stanford t-shirt?
What these free courses are saying to the students headed to state Uni outside the top 10, and in fields that don't require a paper qualification (like software) is: don't bother paying to go to UCLA, do a startup or do opensource AND do our courses for free and you will get the same job.
Gradually (at least in CS) UCLA etc will disappear from the market and the employer radar who then see MIT/Stanford vs self taught.
It's really no different from why Nike pay Tiger Woods a zillion $. The idea is to put in the minds of the consumer (or CS student and employer) that there is only Nike or nothing.
There is no threat to MIT/Stanford/etc from running free courses. Even IF somebody decided = I could afford to go there, I could get accepted, I'm not planning to do a startup - but I have decided to only do the online courses instead of attending - their place would immediately be taken by somebody else.
The real impact of these courses is that they destroy the market for 3rd rate universities. Why did you waste your time going to NoWheresVille U when you could have done the MIT courses and something relevant to me as an employer?
A very similar thing happened 10years in the UK. All the community college equivalents were allowed to call themselves universities. All the 'IVY league' were promoting this as 'access for all'.
Of course what it meant was that employers weren't sure if somewhere they hadn't heard of was a decent 2nd tier university or some degree mill - and so to cover themselves began demanding IVY league degrees only.