LESSON TWO


MOTIVATION

Every marketed high tech machine that I know of comes with an owner's manual. In that sense it is a teaching machine. If you define a teaching machine otherwise, then scan the manual into a computer and read it as instruction in a machine and by a machine.  If a survery were done on how well the manuals for VCR's, telephone answering machines, fax machines, photocopiers, printers, digital watches and so on do as teachers. the results would be dismal. But good pedagogy is not the primary motivation in this marketing strategy. It creates a secondary tier of technical support personnel and the profits which accompany that tier.

There are a number of motives behind the current technolord-technopeasant social stratification as there are different motives in the opposition to this social order which may some day turn into a kind of revolution. Later lessons will explore some of these motives. Beyond that, more senior courses in psychology and related disciplines like philosophy and political science are required to advance the study of human motivation. Students going on in the behavioural sciences will pursue this path. Those going on in technology will hopefully have a more user-friendly attitude instilled in them by IMP and will take up the cause of developing better teaching machines. Start by asking yourself the question as to how far technology could go in the development of the teaching machine. Could you develop a MACHINE WHICH TEACHES COMPREHENSIVELY ABOUT MACHINES? If it teaches comprehensively to humans, could it also teach comprehensively to other machines?
This is futuristic no doubt, but worth considering at present because its actualization is only a matter of time.


COURSE PLAN

You should be reading through the material in the online "Machine Psychology" text, up to and including Chapter 1 which is titled "Total Automation-Now!" Some of the web sites may no longer be current as this text was written from 1997-99. Please use the standard search engines to find related information. All technologies as well as their social analyses are advancing at an astonishing rate. As for the "GT on the couch" exercise, you can wait until I get around to setting out some easy-to-follow procedures for the use of Geocities "PageBuilder" program if building a web site is your primary concern. However, if that is the case, you will miss the major reasons for  this exercise. It is better that you fumble around with PageBuilder for a while as I have done. Use some introspective psychology and MAKE NOTES ON HOW THESE FRUSTRATIONS MAKE YOU FEEL as well as how much time the trial-and-error learning costs you.


TEACHING MACHINES: THE BIGGER PICTURE

At <http://www.worldwidelearn.com/index.html> we find what is purportedly the world's largest directory of online courses. Thousands of courses from thousands of universities can be accessed. What if they were presented as much as possible, taught by machine-only?

Take it to the next step. Common to all of these fields of knowledge we find that QUESTIONS are asked and when the correct ANSWERS are given, competency is considered as having been established. Likewise for IMP: if you are asked questions on the subject matter presented and you give the correct answers, you are considered to have passed the course. Can we develop new methods for quickly entering this maze of Q-A sets and finding our way around? That will be the challenge ahead and presently the answer is not known in much detail. It may even be that powerful new search engines will be invented. There is a search engine at <http://www.alltheweb.com> which claimed in 1999 to be the first search engine to index all web sites, a total of 200,000,000 then. All of these web sites, whether used to present courses or not, are in effect teaching machines. Presumably you could ask the human developers of those web sites the answers to all of your questions and they would be able to answer correctly. What if the machine's search engine were designed to do that job as well as any human...or better?

Given that we don't yet have such a search engine, we are left with more conventional methods for dealing with Q-A sets pertaining to teaching machine web sites. However, the teaching machine slogan scrupulously employed for IMP that "THE STUDENT IS ALWAYS RIGHT" should lead us toward a solution. The first year level of courses like IMP would be written in a version of Standard English which introduces a certain number of new technical terms but all lessons and Q-A sets are expressed as we go along in English vernacular. The lexicon and syntax are BUILT UP FROM THE VERNACULAR, step by step. Take the previous sentence as an example. "Lexicon" and "syntax" are not everyday words, but they can be found in a standard dictionary.
Words which are more novel and technical must be defined as the lessons proceed. Their usage (synatx)  is further explained as the lessons proceed. Standard pedagogy...yes; with the exception that THE STUDENT IS ALWAYS RIGHT. A teaching machine which does not teach is as much a failure as a vacuuming machine which does not pick up dust.

Thus the trail through this TEACHING MACHINE MAZE will be marked at every choice point with clear signs. Confusion is the biggest enemy of the teaching machine. The result will be a knowledge road map. With a road map of any world city we can decide what our destination is and by applying a simple rule like going to the index and then the x and y co-ordinates on the map we find our route. We may discover new rules for teaching machines as we proceed in this way.

One immediate implication of web-based teaching machines might have become apparent already. Every IQ or other intelligence test item I know of is explictly or implicitly a Q-A set. A question is asked. If the subject correctly answers, that subject is credited with a point. A machine which can quickly and correctly answer our questions pertaining to thousands of subjects must be considered as an accomplished AI. Indeed it would be an AI functioning at a super-human level, beyond that of any single human on earth. That too is part of the "bigger picture" of web-based teaching machines.


PSYCHOLOGY

J.P. Chaplin's Dictionary of Psychology defines psychology as the science of human and animal behavior. That definition would include simulations of such behaviour or substitutes for human and animal behaviour, in other words it allows for the subject matter of machine psychology.

As explained in the introductory sections of your text, "Machine Psychology", the short definition of robot will be used in this course, though it will encounter controversy if you want to discuss it online with experts in robotics and computing science. According to this definition, ie "reprogrammable, multifunctional manipulator", a computer is also a robot. That is a useful reminder for IMP instructional purposes because your present teaching machine can morph into many forms, some of which will look very much like humans and will have greater degrees of autonomy.

To illustrate how the desktop computer is morphing into more human-like machines, consider Hewlett Packard's innovation in person-to-person conferencing. Yes, the standard computer uses various software packages for tele-conferencing. Online courses all use tele-conferencing software and that includes IMP. A May 26/03 article in Wired by Elisa Batista titled "Working Remotely, Robots in Place" describes the HP "Surrogate" as HP prefers to call it rather than robot. The Surrogate is a 5 ft. 8 in. tall mobile plastic machine with a computer monitor "head". The remote operator can project his or her facial image onto the monitor, and move the Surrogate about with a joy stick. It is a mobile computer with audio-video in and audio-video out as well as a human face via monitor. HP's Surrogate could be teaching you IMP. See <http://www.wired.com/news/business/0,1367,5891,00.htm>

For that matter, "GRACE" could also be your teacher and GRACE is more of an autonomous robot. Like HP's Surrogate, GRACE has an expressive face on a monitor. It was built for the Mobile Robot Competition in Edmonton, 2002. See <http://palantir.swarthmore.edu> and <http://www.cs.uml.edu/aaairobot>.

Hosted by www.Geocities.ws

1