I speak on open access (OA) and open educational resources (OER) regularly, and I'm often asked what I think about the Open Education Movement (OEM). That question generally comes in the form of "Do you think OEM is going to destroy universities/higher education?"
Generally, my answer is "No." Among other things, it's hard to destroy anything so deeply rooted in a culture as higher ed is in ours. But that doesn't seem to be a sufficient answer, so I started thinking more specifically about it. I'm working my way though those thoughts here.
As I see it, higher education has traditionally performed several functions that have all gotten mushed together: content, teaching, certification, and research. Each is affected differently by OEM.
1) Content. Sorry folks, content has flown the coop. With some self discipline, learning skills, and a decent Internet connection, you can pretty much learn the basics of anything you want without setting foot on a university campus. You may not be able to advance too far, and some fields are harder than others, but you can learn an awful lot about an awful lot of things for the cost of Internet access. Actually, you always could learn on your own, but the self educated scholar of the past needed access to a physical library, so geographical concerns were more important. (Geographical concerns are still an issue; just ask anyone who has had to deal with trying to connect to the Internet when 1) electricity is an issue; 2) Internet connections are few and far between; or 3) politics dictates what you can and cannot view online. But it is easier now.)
2) Teaching. I mentioned self-discipline and learning skills above. Teaching and learning are what transform information into knowledge. I keep thinking about that every time I go on a weather site. These days, you can get a ton of data about the weather: temperature, pressure, humidity, wind speed, radar pictures and videos, etc., etc. However, that won't tell me what the weather is going to be this afternoon, because I don't have the knowledge to interpret that data. I probably could sit down and read meteorology textbooks, watch weather videos, and eventually figure out what I need to know. But it would be much easier to take a class in meteorology, where someone will distill the most important points, demonstrate them in several ways, ask students to demonstrate the analysis and interpretation needed and correct them when things go wrong.
That's the hardest part about learning on your own. You may know something's wrong, but you can't always figure out what or how to correct it quickly. And the more you do something wrong, the harder it is to get it right--just like any bad habit. A good teacher catches those bad habits before they get reinforced. A good teacher can head you off before you go down an unprofitable avenue of learning (unless that detour will be good for you!)
There is the question of whether a poor teacher is worse than no teacher. It probably depends on the particular type of poor teaching. We also seem to remember poor teaching better than good teaching, and even more we remember poor learning, which is affected by much, much more than the quality of the teaching. It's possible for a good learner to learn from a poor teacher and a poor learner will struggle with the best teacher. I often feel that opportunities for learning are expanding while the skills for learning are dwindling.
3) Certification. Also know as grades and degrees. As a society, we trust our institutes of education (higher and otherwise) to tell us when someone has reached a level of proficiency in a subject. This is why diploma mills upset us, and why the talk of a "higher education bubble" seems so threatening. What if educational certification doesn't mean what we think it means? This is probably the greatest threat OEM poses currently: what if there is another way to certify proficiency, and what if it doesn't match up with traditional grades and degrees?
This is probably related to the scholarly publishing crisis in a way. We count "good research" as research published in "good journals", but what if there is another way to judge? For instance, the aggregator model would turn journal publishing on its head: publishing is the easy part, in open access repositories, but then comes the peer review via some sort of commenting system, and finally journal "aggregators" would select particularly well received research to be republished. Journals would become a stamp of approval, not a gateway.
4) Research. One of the purposes of universities is the production of new knowledge. Even small teaching colleges encourage research (though they can't always support it.) One of the great draws for a good college or university is being able to learn from people doing neat stuff, and possibly having the opportunity to do neat stuff with them. That would be difficult to replicate outside of a higher education institution, though it's probably possible in some fields and for some projects. Citizen science is a lot of fun but it's not the same experience as being in the lab with someone. I've done both, and learned a lot from both; they just aren't the same.
So content is lost, certification is chancy, but teaching (at least good teaching) and research could be what higher education needs to concentrate on.