21st March 1981
I have just read chapter 6 of "The Brains of Men and Machines" and come to the conclusion that the mechanism I am designing must have in it the ability to continue the pattern recognition hierarchy on and on forever - grouping symbols and giving it a single symbol identity - using this symbol in a group etc. So the function of short-term memory to do this is most important just as the ability to ungroup a symbol into its components.
20th Apr 1981
There is the need for the retention of a goal state. I have [in my model] an increase in the level of attention-interruption-concentration as a result of a good thought but no remembering of the goal state trying to be achieved - unless that is implicitly stored in the address of execution.
21st May 1981
Need to add inter-modal associations - i.e. between different input sources - touch - sight -sound etc.
- Generalization and over discrimination requires that objects have a number of properties that are common or different. This implies that a single concept is insufficient for generalization and over discrimination. This means my current model can not generalize.
- I need an association which is the next sight, next sound, next idea, next output etc. which will attract attention just like the associations already generated. This gives locality of associations - remain in the context of the original cue and help in the sequentialization of the order of doing things.
- Associations should be made based on the assumption of cause and effect relationship between events.
26th July 1981
Interest is the drive. Interest in what to do (if anything) when situation recognized. Interest in what to do (if anything) when new situation. Therefore a new situation provokes attention - observation - by default second time observed - its familiar probably has an associated input. But that is all (implies it is expecting something) Can do a number of possible things.
1/ reflexive motor NO!
YES! 2/ execute associated input -> listen for it. Expect it.
If input different - new situation
If input same as expected - then reflexive motor.
Possible associations (cause and effect)
- Input associates next Input across any number of thoughts but not intervening Output.
- Input associates next Output across any number of thoughts but not intervening Input.
- Output associates next Input across any number of thoughts but not intervening Output.
- Output associates next Output across any number of thoughts but not intervening Input.
i.e. I associates with next I or O which ever comes 1st.
O associates with next I or O which ever comes 1st.
- Rewards / punishments should also halt any further associations.
Thoughts are symbolic manipulation which allows one to model ones environment without suffering the consequences of real life. One can think through a situation sequentially searching for cause and effect and evaluating desirability of actions and consequences before acting.
2nd Aug 1981
Ideas cannot be associated across inputs.
- The context problem we must resolve is that a resulting input is only associated with an output if the output is done in a specific context i.e. an original input cue.
16th Aug 1981
Thinking is a process to model the external world by going through the conceptual representations (internalized inputs/outputs) and predicting the outcome. Thinking is made up of expectations - predictions based on experience.
- Presumably when one is thinking and comes across a concept that "has negative connotations" one does not do the actions which would result in it. If one comes across "positive connotations" one would do the actions - starting from the original input that instigated the line of thought.
- Two modes of thought are executing mode where the thought process exists and one is just following the steps and the other is free thought where associated ideas interrupt the process. One could look upon it as procedures that are continually being perfected.
We currently execute the past experiences even if it was not rewarded. This implies past actions taken were acceptable (not punished) but we are really not furthering our goal by doing this since we are not pursuing interesting things.
Let us reflect on 26th July 1981 some more.
1 New Input - interest
2 Same input - familiar expecting previous associated input - store idea of input
3 Next input - not familiar - not same as expected before(idea of input) -> interest -> any associated output, input etc. go to 1 or 2
- same as expected - reflexive motor
4 Same input - familiar - associated output - store idea of output - do it.
- expecting some input so go to this process #3 - not familiar
- In this scenario - no punish or reward - always trying new things on environment so never have a repetitive input stream.
- The same process should apply to thought process.
New concept - interested
Same concept - familiar - expecting previous associated concept
Or should it? Or does it? The purpose of actions in environment is to be interested in it and adapt to it. In the absence of rewards & punishment the above scenario works - trial and error. The purpose of thought is to model the actions and environment. We must learn to do correct actions and we must learn to think - therefore one must be trying to think - and not repeat unsuccessful thought sequences. Does one learn to think?
No such thing as unsuccessful thought sequence! Is it the content of thinking that is learned or do we learn new + different thinking processes?
- Thought models environment - Inputs and action - outputs as well as reward/punishment, therefore it models trying to achieve concept of reward and avoid concept of punishment.
- Go back to process applied to thought!
1 New input idea - interested
2 Same input idea - familiar - possible associated same input
Or possible associated different input.
If input idea is same as associated input then look for associated output.
If input idea different from associated input then store new idea of input, go to 1 or 2.