.... a first year college-level course suitable for all disciplines from humanities to technology. The subject matter of machine psychology is a broadening of the subject matter of AI, ie the substitution of machinery for the phenomena which have traditionally been included in the study of psychology.To take up discussion of the machine-taught lessons in this course with other students and teachers, send a blank email to <[email protected]>.
<http://ai.about.com/cs/onlinebooks> presents itself as "a comprehensive list of free online AI-related books" including the book "Machine Psychology" at <http://iio.ens.uabc.mx/~mbarbosa/atoma>. IMP will broaden and deepen the material at that web site Though this text is being entered by a human instructor, when it is completed as "courseware" the lessons will serve on a "stand-alone" basis, taught only by machine.
This is a course on machine psychology taught by machine.
Our opening lesson is on the Geocities Teacher at <http://www.geocities.com>. Try it out.
How well can you follow the lessons on building a free web site? If you have no more aptitude than myself, you will find them to be time consuming, frustrating and unpleasant.
For a machine teacher to get a high grade, the lessons must be as easy to read as Canadian Geographic magazine. Good machine teaching adheres to the slogan, "The student is always right". In a conflict between human student and machine teacher, the judgement always favours the student. That slogan and related issues are dealt with in <[email protected]>. However, the particular lesson of the Geocities Teacher (GT) will be discussed on the IMP list above. Please take your comments there. GT is a machine. You can be ruthlessly honest about GT's teaching ability! Score a plus right away for the advantages of machine teachers over human teachers.
This exercise works best if you first try to build a geocities web site by yourself and are completely honest with yourself. Write down every difficulty, confusion, frustration you experience. The chances are almost zero that you are alone among students experiencing any given difficulty. Thus if you want to develop courseware for students or understand the concept of automated courseware, doing this will be instructive for you.
Some students will of GT will have a background in computing science and /or high aptitude. They may find GT to be a satisfactory teacher. But the prerequisties for IMP are only K-12 equivalency and the ability to send/receive emails and look up a web site.
My background experience is that last year I called in a computing consultant and asked him to install a scanner plus give me instruction on using the scanner to set up a web site.
After a half day sessional I had a scanner and seven pages of notes which I scribbled while he was teaching me scanner and geocities web site building basics. Releaved of several hundred dollars I then put all that behind me until I pulled out the file for present purposes.
But I am not complaining about the particular consultant...far from it. I have received similar short courses from two dozen such people over the past decade. This was the best of the lot.
GEOCITIES TEACHING MACHINE TODAY:
What I will complain about is the INADEQUACY OF PRESENT MACHINE TEACHING.
There is not one among those two dozen consultations referred to above which could not have been done by machine alone and done better by machine teacher than by human teacher. Thus here we have the first example of machine psychology, ie a machine substituting for the human computing consultant-teacher who taught me to use the geocities web site. How far I go with this exercise presently depends upon a number of considerations, one being the extent to which I can benefit from those seven pages of notes from last year and the use of the GT instructions available presenly online. I was not
able to keep up with the instructor as I made the notes. That is, I was given at least twice as many instructions as I set out in my notes. Score another plus for a machine teacher. Once GT is perfected it will have all notes stored in the machine's memory. The human student then does not have to rely on either human notes or human memory.
Human memory is another concept from psychology. Now we are talking about a machine which substitutes for the broadly based human behaviour of the teacher and also for the memory of the student.
Try setting up your own teaching module for GT on your own geocities web site. You can discuss the experience with others by subscribing to the IMP list. You may find it advantageous, at least at first, to use a web mail account (eg a yahoo account) under an alias. There are psychological inhibitions which sometimes prevent us from disclosing to others when we do not understand something. We may feel "stupid" for not understanding something and we are inhibited from disclosing that "stupidity" to others. Thus I WOULD RECOMMEND USING AN ALIAS. You can reveal your name later if you want. Remember that in automated teaching, "the student is always right".
Other than avoidance of taking some aversive emotional hits, another benefit of anonymity is to the entire courseware-building process. The more you can reveal your difficulties as a student, the faster the GT courseware can be perfected. You might think of each difficulty as associated with a frequency distribution if, let us say one thousand students were to encounter GT as it is is. The chance that a given difficulty is experienced by only one student out of one thousand is extremely small. Your complete and uninhibited expression of learning difficulties is to be encouraged. And one way to lessen inhibitions is with anonymity.
TBC
GT ON THE COUCH
This is a test of another format for laying out text.
The problem encountered by the column on the right has been presented on the IMP list. Rather than wait for a solution to the problem of eliminating the top left line on the web page which prevents me from widening the column of text, I have been able to start a new column on the left by starting the PageBuilder session with clicking on the T (Text) icon on the top horizontal tool bar upon which this left-side type box drops onto the screen.
But these are the details which one could spend much time trying to figure out. A well-functioning GT would avoid time consuming and wasteful trial-and-error learning. Given that what one learns from GT is modelled by maze learning, there are only so many moves which one can make in a maze. If GT gives all of them in a language we (the students) can all understand, GT is a complete success. Until then it is not a first rate machine teacher. The problem is that of how to teach the teaching machine.. If its human managers do not want to do so, these words are in vain. This problem has been addressed by various eminent persons in computing science like Donald Norman ("The Psychology of Everyday Things"), David Gelernter ("Machine Beauty") and Richard Stoll ("Silicon Snake Oil").
The affliction of GT has been presented to the Geocities management for remedy (see IMP postings)
so a remedy may be forthcoming....
In "The Psychology of Everyday Things", Norman confesses, "Over the years I have fumbled my way through life, walking into doors, failing to figure out water faucets, incompetent at working the simple things of everyday life...My difficulties were mirrored by the problems of others. And WE SEEMED TO BLAME OURSELVES" (p. vi). Blocks are mine.
Does a technology education solve the problem for users of machinery? It will for particular applications in which technology students are trained, some of which will generalize to a degree. But note this Norman statement:
"You would need an engineering degree from MIT to work this, someone told me, shaking his head in puzzlement over his brand new digital watch. Well, I have an engineering degree from MIT. (Kenneth Olsen has two of them and he can't figure out a microwave oven). Give me a few hours and I can figure out the watch. BUT WHY SHOULD IT TAKE HOURS?" (p.1). Again, blocks are mine. And why should it take hours to figure out GT?
Computing science professor, David Gelernter applies what the layman calls the KISS principle (Keep It Simple, Stupid) to "machine beauty" , title of his text. See page 46. Directly relevant to the cyber-therapy of GT we have his comment on a related machine teacher, equally troubled, called :Windows. Gelernter says, "...I have acquired ten useless new characters permanently stuck to my screen-desktop litter, the electronic equivalent of trash on the sidewalk. And some software designer somewhere, formerly the kind of sanctimonious twelve-year-old who always kept his three-ring binder neat, is grinning triumphantly." (p.48)
And before we go any further, do a conventional search via google or other search engine for online computing dictionaries like <http://www.techweb.com/encyclopedia>.
In most social interactions, expectations are placed on all parties interacting. Exceptions are made for the very young and very infirm. But for the social psychology of human-machine interactions we need new norms. We do not apologize to the vacuum when it fails to pick up dust from the floor and similarly we do not apologize to GT when IT FAILS TO TEACH US. Pardon the fumbling about with attendant shortcomings in present lesson layout but that trial-and-error cost is part of what this exercise is about. Lesson plan web sites for subsequent lessons can be found on:
<http://www.geocities.com/machine_psychology/Table_of_Lessons>
| Add your text here |