A colleague of mine, Sebastian Thrun, has decided that education should be free and open to all. Just over a week ago, he created a website:
http://www.ai-class.com/
and has already had several thousand people sign up for an introductory artificial intelligence (AI) course he is teaching this fall. Of course, by only offering “instructor certificates”, there is really nothing at stake for the online participants, so that cheating should not be a problem. But the reality is that participants will learn a “real” AI course, taught from some of the best AI researchers/teachers in the world.
Will such an approach “work” at the collegiate level?
What does it mean for this class to be successful with respect to the thousands of online participants?
Should/could real credit be given, and if so, how might one assess student work?
What are possible implications for K-12 CS education?
Can we teach students to program in a non-apprentice-based approach?
Steve Cooper
Chair, CSTA Board of Directors
MIT Open Course Ware has been doing just this for years. I have used some of their courses in conjunction with my classroom materials in teaching Computer Science in a Maryland public high school for the last five years or so (since the courses have been available on-line).
I cannot answer most of your questions, but I can speak a little to the question of the putative impacts of these programs at the K-12 band (in public schools). Before doing so, however, I am ambivalent about your last question: As someone who was a “real” programmer for over twenty years, what we do (or did) is different than what’s expected in an academic setting, that I don’t see the apprenticeship model as being relevant … but that’s a whole different discussion, I suppose.
I find that most of the material (again, ‘material’ here refers to the learning outcomes, etc., in the MIT packets) is problematic for high-school students who:
1) Lack any exposure to authentic mathematics—i.e., they suffer from an over-exposure to operational mathematics;
2) Have no understanding of logic;
3) Have no experience with formal languages. Actually, on this last point, a vast majority of students in my classes (and this is a high-performing school community) have never diagrammed a sentence and have NO understanding of grammar, or the recursive application of reduction, etc.;
4) Lack any motivation to acquire anything more than a cursory understanding of the “technology” in isolation from the “science,” (in other words: they are interested if they think that CS is about playing games on the computer) and
5) Lack the time in their schedules to devote to address points 1-4, above.
I don’t think that the free availability of these kinds of courses are problematic to K-12 CS. Actually, they are helpful. What’s problematic for K-12 CS is the culmination of irreversible events:
1) Around 1996, when the political discussions began about “standards,” and the like, CS had no lobby; in effect, CS in K-12 education allowed others (their competition) to define them;
2) The events of 2001-2004, specifically the revelation of widespread (or what appeared to be widespread) out-sourcing turning many parents opinion negative;
3) The lack of a simple and appealing (this is the United States, after all) articulation of what comprised CS and why it was important during this time; and,
4) The contemporary political landscape which could only be described (kindly) as focused on other issues.