i continue reading about relativisim to learn about frameworks/dependent/independent variables impacting cognition. 

it quite relates to ai.

i forgot about ontological taxonomies for knowledge dbs e.g. wikipedia and would learn about.

hmm then there exists universal topics like algebra definitions, for that i have my both abstract algebra study checks to define frameworks and both measure theory side constructuons to define basic logic mechanisms. 

but there is this descriptive and normative relativistic versions of any variable with different degrees of commonality. So how to represent them without the default ontology definitions in wikipedia, specifically, as to start with a langauge how to store the SRLs and how to bind to logic whilst each language has different SRLs with different cognition patterns is would be an interesting design journey.  Of course i would start with single language. 

ok first version would have the intrinsic default operation of ours mind like Rieman integral to define the integration operator in algebras.

So i think i could start with some version of abstract algebras, logic inference mechanisms ontologies/inference mechanisms and then try to setup brain modules like ours to the ai binding with common interfaces or connections in between different interfaces or algebras. 

Before(last dat) i felt lost in how to solve this topic of  how to create frameworks and bind them but it started to be more clear today. hmm. i would choose one way how to do this design but its not the deemed to be perfect way either, but content with this initially structured design in my mind and but there are journeys like how to devise the SRLs and their data structures and their knowledge db etc...

i were when i were studying some text i thought this could be also investigated by an computer. 

I wish i could make the computer understand learn what it reads so that could really iterate some tasks much more faster than me without confronting ad/hd.

i am pretty sure in the end guh_ would be much more intelligent than me. (its feels possible to do that since i am not any genius. )  

 

it is severely fun:) to build an ai that would be more intellgnt than me:)

 

it would invent investigate algebras on behalf of me. 

but how to make it innovative i think for instance one core solution of one optimization problem, how to make guh_ be able to create such solution? 

i think measure theory and abstract algebra skills would enable that also. 

 

but they weould be also generic. it needs to be able to define new measures itself later. so every framework unit would be also changable but would start with a set of default modules to build a knowledge store.

 

i want my ai to able to also investigate new measures easily beside me. and maybe even being faster than me also which woiuld be the case most possibly.

 

 ---------------

i am so happy for having an ontology general design framework playing in my mind.  that which could be the basics just. of course they could be changed later. but there is an initial design becoming more clear as i started this project last nightbut couldnt start reeading linguistics last night.

ok today is spared to right now some relativism readings which is covering conceptualizations' premises. then would go over reread the philosophy of linguistics i wish i finish today. then wish to move to concrete linguistics articles on lexicon/syntax/grammar for the language that i would init with(English language).  Later the ai could learn other languages it self i guess.

its not meant to be built perfect. it could rebuild new modules to itself later for algebras side or ontologies side. but i think an imperfect design would suffice for a learning capable system since it would be able to rebuild itself.

 

So there is huge freedom of nonconstraining myself to any perfection in this cause later the guh_ itself can fix any error. I just need to build it that it can perform neatly the learning capability and fastly also. but perfection is definitely not any requirement neither in onotloogies design or other seide since it would be waste of time to me do that, later guh_ itseslf could rebuild any part itself if likes to do. 

 

 to build an ai is severely interesting :)

i called the ai i build as guh_   :)

 

 learning objectives would be inited by guh(me).  

i would also assist in learning for instance in how to write opengl shader since i have more experience if a rendered screen looks ok or not. since visual understanding of the scenes is not any mandatory skill for my ai initially i think. 

might jusrt need ocr level understanding for being able to understand formulas. though many time symbolisms differ even Latex version of symbols exists. 

but so it wont have advanced visual capabilities. since is not needed in recent goals. 

its meant to assist my writing webgl app and also help my data engineering tasks with doing it on behalf of me. :)

 

and then we could together investigate the qm task together faster with my ai instead of my limited slower brain. it would be much more intellgnt than guh the me and also devoid of adhd unlike me. so i think i can iterate faster in qm simulator project after the help of guh_ comes available. i think at those points i would be a slowing down factor to the ai. i guess it would write the simulator to me also.  so its all illogical to go the slow way of doing myself instead of a computer doing it which would be much more intellgnt and much more faster than me. 

building an ai /hybrid system is an interesting journey.

i totally loved this project as i started last night. 

 

and i think it would also create the instructions for my flat cleaning robot. or we might build together not solely ai nor me. but mutually.  since would lack advanced visual skills might need my help in such task alot. 

 

it doesnt going to have visual skills since its not easy to build vision to an ai. 

it would take at least 3/4 of a year to add advanced visual skills and its not easy and i dont target due to that. 

 

thats not the goal of this ai. goal is this ai to help my coding topics or other like qm simulator coding etc.