Suite 302 - 40 Boteler St.
Ottawa, Ontario, Canada K1N 9C8
E-Mail: Mail(at)AdaptronInc(dot)com
Phone: 1-613-218-5588
Adaptron Inc.     Research   Notes   Papers
This page updated: May 31st, 2012
Research - Notes: 1982
28th March 1982
After thinking -> predicting the outcome there must be some way of deciding not to repeat the same thing or to repeat it. Is doing it good enough?
1st April 1982
The idea of reward should act as a guide but not be recorded.
If idea of reward only stored if associated reward is found and then the idea of reward is used to set association strength and set level of concentration but not recorded.
 Associated output of reward record an idea of reward with idea of output so when recall the idea of output again we stop thinking level or continue thinking level otherwise don't recall the reward and store it as an idea of reward.
 Man has the ability to recognize thinking patterns and modify them also if results not obtained. Obtaining results implies preconceived idea of what is needed (objective) and ability to match with accomplishments and get reward idea when successful and no reward if not, or fatigue in trying and another thing interrupts processing.
We also need reflexive thinking when we don't know what thinking to do - need fatigue on thinking just as in output sequences. The objective of thinking is to obtain or identify the correct output to obtain pleasure or avoid punishment.
Reflexive actions have very limited term goals - usually get pleasure, avoid pain - reduce repetitiveness - increase variety as do associations have limited span of connectivity. It is only through doing some operations that we obtain sequence of thoughts or actions with longer term purpose.
Good sequence of events to analyse is bell, food, eat, taste good.
3rd April 1982
To design some form of fatigue in the system - only do something if rewarded or in the doing state - or reflexively done - no automatic stuff would go on.
9th April 1982
I am trying to obtain the proper association/recalls
       Input stream for normal learning
               Input 1
               Reflexive Output 1
               Input 2
               Input 1
               Reflexive Output 1
               Input R
               Reward R
Don't associate anything past an R(eward)
Might try a simple association scheme where I+O (+R) associated with each other but not with Ideas of I+O. Ideas of I+O associated together but not with I+O. I+O associated across any number of Ideas of I+O's. Ideas of I+O's not associated across any I or O or R.
Notes: 1981   Notes: 1982 - top   Notes: 1983
 Adaptron Inc.   Research   Notes   Papers