New BuddyPress Plugin: Better Membership Directories

Telegistic today reported on the new BuddyPress plugin being developed for the  CUNY Academic Commons.  This is exciting news for educators who are building and developing self-sustainable LMS apart from what their universities provide through what used to be called Global Learning Systems (GLS) such as OnCourse, BlackBoard and Angel.   It’s also a boon to  those who believe in the democratized education movement and see  BuddyPress along with ScholarPress as coming a long way in filling the customization gap that instructors need.

The new BuddyPress plugin is only in Beta right now so its not recommended for general public use just yet.

I’ll be discussing more of the newer ScholarPress/BuddyPress goodies  and what you can do with them to deliver a customized course experience in an upcoming post.


Personal SuperComputers, AI, and What Does That Have To Do With the Future of Education In A Web 2.0 World?

Online Education, AI, Web 2.0 and Web 3.0

We’ve already discussed Web 3.0, the Semantic Web, in a couple of previous NanoWeek blog entries and we have briefly mentioned the subject of Intelligent Agents. And while this week’s entry is not about Intelligent Agents, per se, it is about another kind of machine intelligence…Artificial Intelligence or AI for short. So what’s the connection between Web 3.0, online learning in the Web 2.0 and 3.0 age, and AI? The “glue” or connective tissue between these three concepts (and of interest to educators now living the reality of these concepts) is discussed in a book written a decade ago By Roger C. Schank.

Tell Me A Story
It’s called: Tell Me A Story – A New Look At Real and Artificial Memory, a very fascinating work considering how we index and construct stories, how we use stories as metaphors for understanding the context of things and relate to others and how this relates to computers and artificial intelligence. It’s a concept that noted writers and authors, G. K. Chesterton and C. S. Lewis, hit upon in their day when realizing that we tend to place our ideas not in the construct of the abstract but in clear mental images. This realization has a great impact on the way we design computers to “think” and this innovation in thinking was picked up on by Professor Schank in connection with his research in the field of AI. I don’t know what he is up to now, but Schank was at one time (and may still be) a John Evans Professor of Electrical Engineering and Computer Science, Psychology, and Education at Northwestern University. He used to be the Director of The Artificial Intelligence Laboratory at Yale University. But back to our story…..

You Gotta Let Them Create Pictures in Their Mind

In her article Read Aloud, Virginia, Patricia Muller states one of the corrolarry discoveries being made as technology was making greater advances in “streamlining” visual communications: “Recent brain research confirms that optimum brain function and development is still dependent upon storytelling. In fact, it is the most effective stimulation for young developing brains. This is because brain development in children is dependent upon induced imagery. That is, in order for the brain to develop properly, it must be stimulated; instead of “use it or lose it,” it’s “use it or never get it.” Induced imagery means creating pictures in the mind.” (emphasis mine)(1)

Recall, that the writers G. K. Chesterton and C. S. Lewis (who were well versed in the field of apologetics) had come to the realization that people tend to think in mental images. Chesterton consistantly used the imagery of language when forming logic arguments from the 2nd level which is Drama (Level 1 is Theory and Level 3 is “Kitchen Table Talk”). For example, he illustrates that when we rely on reason alone, we tend to overlook the problem of the individual becoming trapped in the “clean and well-lit prison of one idea”. And in order to break a person thinking that way out of that “prison”, the writing cannot successfully argue from an abstract position but must “snap the spell” because we, as listeners and readers, think naturally in metaphoric images, in stories. (2) Thus, Schank’s research into story indexing has a great impact on how we design a computer to “think” as artificial intelligence. As Shank writes: “Stories give life to past experiences; stories make events in memory memorable to others and to ourselves.”1 In other words, memories are really stories which can be recalled at a later time. Children who are exposed to information in the context of a story can better recall it later.” (3)

Indexing the Story

Now this explanation of recalling is very important to the idea of indexing information for later recall when it comes to complexities in thought and deriving connections. A good illustration of what I am talking about is one in which I used Schank’s observations on memory and indexing while analyzing one of Shakespere’s most complex plays The Winter’s Tale:

“The Winter’s Tale is an invocation of the oral tradition, the narrative act to expose the power of what we are told when we are young and without the predisposition to doubt. (Lamb) Some of the stories we are told are, without a doubt, true and sustain us throughout a lifetime and some of the stories we are told (and we tell ourselves) do not last much further than the damage they cause. Stories seem to be an inconsistent form to convey truth, don’t they? But we still need stories. Why? Schank asserts people need a context to help them relate what they have heard to what they already know (15). He goes onto to relate: “When a decision-making heuristic (a rule of thumb), is presented to us without a context, we cannot decide the validity of the rule we have heard, nor do we know where to store this rule in our memories.” So how do we learn to distinguish what is true and what is false? How do we train our perceptive powers to distinguish between right and wrong? (NWT)” (4)

Why is this seemingly disparate piece of information important and pertinent to our discussion? Because it illustrates the point Schank is making about indexing. Shakespeare’s play The Winter’s Tale is to complexity of idea illustrated in words as Rachmaninoff’s 3rd Piano Concerto is to complexity of idea illustrated in music. The Winter’s Tale is considered the most difficult of Shakespeare’s plays to understand. Likewise, The Rach 3 Concerto is considered the most difficult of musical pieces to play. Both the play and the musical piece are as one professor of mine put it: “a big rambling barn of a context” used to distill a very deep and complex concept/answer within the context of a verbal story or musical story. This is the crucial point on which designing artificial intelligence hangs: searching for facts is not searching for stories. It’s an entirely different matrix we need to address and connect with. The semantics are vast. Schank further outlines the matrix of storing stories in memory by referring to their framework as “story skeletons”:

“The key point here is that once we find a belief and connected story, no further processing, no search for other beliefs need be done. We rarely look to understand a story in more than one way.” (p.73) “The skeletons we use indicate our point of view. Storytelling causes us to adapt a point of view. With this adaptation comes a kind of self-definition, however. We are the stories we tell. …As we come to rely upon certain skeletons to express what has happened to us, we become incapable of seeing the world in any other way. The skeletons we use cause specific episodes to conform to one another. The more a given skeleton is used, the more stories it helps form begin to cohere in memory. Consequently, we develop consistent, and rather inflexible points of view.” (P.170)

This is why some have espoused the old Julius Caesar quote: “Experience is the best teacher” as a maxim, if rather flawed. That perception is based on the filter with which we may see a certain situation, event or person. And that filter may (and often does) lead us to many erroneous conclusions and labels

In regards to the memory concept of labeling, Schank states: “An incident is remembered in terms of how it is seen in the first place. That is, labeling is in many respects an arbitrary process. He goes on to apply this to the concept of categorization which we would naturally do in a relational database of singular facts for which a particular query would define the results as facts combined in the form of information. But with stories the categorization is much more complex as in the following statement which recognizes the effect of perception on information gathering:”…And, of course, even that last categorization is arbitrary since one person might characterize the victim as being blond, while the other might characterize him as being fat.” (P.222)

Intelligence Depends On Clever Indexing

The process of learning for an artficial intelligence, then, is much more complex than we might imagine. “A good memory, then means an attentive labeling facility during processing or you aren’t going to remember what you don’t find interesting, so the more that interests you the better memory you are likely to have.” (emphasis mine) (P.223-224) “Yet what we learn is still entirely up to us. No one teaches us how to index after all. We make up our own way of seeing the world,…” (P.113) “Knowing a great deal about a subject means being able to detect differences that will reflect themselves in differences in indexing. In other words, intelligence depends on clever indexing.” (Italics mine) (Schank uses an example of a man who had a well-read background that included military history for one observation:” Our expert is intelligent about military history. He sees nuances where others would not. He analyzes new stories well enough to be able to relate them to old stories that might not obviously be the same.” (P.113)

Creating New Stories By Making New Connections

The intelligence we are looking for then in Artificial Intelligence (and the path that Schank alludes to in his book) is one that works off the capacity of random access/case-based connections which allow for both lateral and vertical nuanced thinking. Years ago, there was a memory game called Trivial Pursuit that was quite popular. It worked off the premise of a knowledge-based thought pattern. Call and response-question/answer regurgitation. Access the right memory bank in your brain and you won. But later on another game came out that took the Trivial Pursuit idea one step better and combined 2 ways of accessing information and was a more precise illustration of what Schank is talking about and a step in the direction of what AI needs to do: 7 trivia questions were asked as if in a sequence, all seemingly disparate in nature, but had a commonality which pointed to a final answer. Use sequential, knowledge base memory to answer all the separate questions correctly. Discover the commonality and you won. I loved this game because I was good at it. It required the kind of random access/case-based reasoning that comes from being able to see connections over vast territories of seemingly unrelated factoids, disciplines, experience, and histories. The game led your mind through the process of creating new stories by making new connections. That’s the territory AI has to conquer.

The Need To Mentally Index Our Own Stories

Even with this grandiose ambition, wisdom should caution us as was illustrated in the famous children’s story Danny Dunn and the Homework Machine, that we will only be able to get the machine to appear to do what we can do…faster. Yet more important questions remain for us: what really was going on in the human mind in those moments when we were doing those same mental computations more slowly? Have we discovered all there is to the process? Those are the questions that are going to get answered in the coming years of Web 2.0 and Web 3.0. And if the speed of information delivery is a worry to many of us now when it comes to getting students to reflect, to be pedestrian for a purpose, on what they are doing (which is the process of learning as we have understood it up till now), what might be the price paid to society when we don’t have to mentally index our own stories at all?

That’s it for NanoWeek. Keep Clicking….

©NanoWeek, M.S.Reed 2008

(1) Read Aloud, Virginia; Patricia Muller, 2000

(2) Orthodoxy, The Maniac, G. K. Chesterton, Editor, Craig M. Kibler, 2002, p41

(3) Read Aloud, Virginia; Patricia Muller, 2000

(4) Knowledge and Belief Aided By Time: The Winter’s Tale and Shakespeare’s New World of Faith, M. S. Reed, 2004, Epsilen ShareIT object and uploaded under class wiki