Thursday, May 12, 2005

The Perfect Teacher...

I wanted to follow up on my thoughts about “the perfect teaching device”. It seems historically when a new technology comes along, we get this perception of how it's going to change our lives. There is the old anecdote that a former president of IBM predicted that the world market for computers would be a grand total of five. They were big bulky machines, and very expensive. And what would a common house dweller need a house sized adding machine for, anyway?

Similarly, predictions concerning robots were severely inaccurate. The predictions were that robots would clean our house, do our shopping, drive our cars. Robots would take on human form and become our friends (read Issac Asimov's first short story in I, Robot). In fact, robots have proven to be very useful, but for the most part they work on assembly lines and look nothing like us. They are just giant arms that perform a repetitive task.

In fact, as science has advanced, we haven't created a more intimate relationship with technology, rather we've used it to strengthen our relationships with other humans. The internet, e-mail, the cell phone, P2P file sharing. All of it helps us communicate and share with other humans, not technology. Technology is the means by which we share experiences with humans.

So why in education do we insist on creating a computer that is the epitome of the perfect teacher?

Dave Wiley made an interesting comment recently on this subject. He said, “As I interact with an intelligent tutoring system, what will be the source of my inspiration? Who will be the teacher I remember forever, with whom I form a transformative bond of trust, who I know cares and worries about me?”

So I pose a question; instead of trying to make the computer a perfect instructor, why not make it a perfect instructional designer?

Think about it. A good teacher does a lot of things at once. He delivers content. Simultaneously he is evaluating whether or not the students are understanding or are interested in that content. He is watching the eyes, the eyebrows, subtle movements of the mouth and forehead. He listens to the whispers in the room and determines if they are questioning, “I am lost” whispers, or “OK, this makes sense” whispers. Based on that evaluation he adjusts his lesson on the fly to better suit the students' needs. At the same time making mental adjustments for future lessons. Spend more time on topic X, and glide through topic Y.

We can't even get a computer to recognize speech, let alone determine if a brow furrow is a questioning furrow, or a 'wow, this makes sense, let me think about this for a moment' furrow.

But what if we had a computer that just observed. It watched how we solved the puzzles in MYST, it observed what web sites we went to, and which ones we returned to. Maybe it asks us a question here and there; “did you find this web page informative? Do you find this discussion board helpful?” No, I didn't enjoy the full 17 pages of Plato's Apology, but I found the wikipedia summation very helpful. Yes, I understood Coase's Penquin, and found it very insightful, what else is out there on the subject?

Based on the system's observations, it begins to build a learner profile. It begins to steer us toward web pages, discussion boards, and learning objects, not because they are the most popular, but because others people with similar learning styles and interests have found them helpful. It may even begin to tie pieces of instruction together. Before you read this article, perhaps you might enjoy this background found here.

And what I like most about this idea, is that the human element isn't taken out of it. Rather than attempt to teach me the subject, it may take me to a discussion board set up where the subject is being discussed. The computer would serve as a guide, directing us to places where learning and experiences are taking place. Together humans would learn, not from some unfeeling computer programmed with rudimentary commands, but by other living, breathing, intelligent beings.

Plus, the programming would be so much easier.

1 comment:

David said...

Check out a couple of things:

http://www2.parc.com/istl/projects/uir/publications/abstracts.html
http://www.adaptivewebsites.com/
http://jamesthornton.com/cf/

More when I have a chance. This is not as hard as it would seem. The eduCommons could use something like this.