Lots of new Coursera classes throughout 2012 posted
coursera.org
coursera.org
There is beginning to be too much fragmentation in subject matter.
There is no discernible ultimate goal (degree or overall subject matter certificate) or any kind of solid path to coalesce all of the work a student puts into the classes into a coherent structure -- a well defined learning path.
I am hoping this is part of a strategy to gauge demand for different subject areas and, from there, create a cohesive program because if it is going to be more or less random offerings...I don't know how that is going to remain relevant or viable.
It just seems confusing as to what their intentions are.
Companies can set curricula based on what their requirements are. And I can see possible demand for influential people to create and share curricula related to their areas of expertise.
For me the answer would be so I know where I stand.
e.g. I come back after a year or two and I know what material I have covered and what steps are next in the learning journey. I don't want to get three lectures or more in and realize I'm rehashing material, or, alternately, that I am hopelessly lost because I've skipped a pre-requisite and have to go digging for the missing piece.
I tend to think about learning in two different categories:
1) Learning for the sake of the subject itself: the goal is simply to learn - no need to specifically prove competence to a third party.
2) Learning for credential. We need to pay the bills - or want bragging rights. Credentials are required to signal that we have obtained competency in a subject area.
For myself (not-quite-youngish with a BA, MA, other graduate work/certifications) , I sign up for online classes with 1) primarily in view. For my kids (high-school age and younger) 2) is also a concern.
It is likely that Coursera (or Udacity or MitX) will eventually create certifications based upon groups of programs. A third party might well group classes in logical sets across institutions. So I hope you are not too fearful - change may bring about some remarkable new opportunities :).
I'd much rather than go crazy with the experiments in this early stage and figure out what works, then try to stick to some super organized and rigid model that probably isn't the best.
I was happy with ml-class from last semester, and I also pretty happy with pgm-class this semester.