turns out i need to study to topology a little bit more. not the entire sections of algebraic topology topic but up to some level.

hmm i need to start preparing my luggage on tomorrow i think. hmm.

i need to print some topology lecture notes. ok. 

i think beside analytic/algebraic topology study for this layer,  in addition to this, i would need to build agebraic geometry layer to this single maths layer ok. 

hmm. i think might hav e 2 maths layers. one for analytical geometry other for this layer i build. 

hmm might need to set up some ontology RDF mappings maths versions also as would train with wikipedia later if possible. hmm.

 

hmm. in most coding tasks we dont do geometry based tasks alot, but for a fullstack coder, geometry understanding must be present. i mean in thecoder ai project.  hmm.

more challenging part is semantics layer and its intentionality layer connections.

hmm. i need to train bert including context handling, those bert architecture enables that as i remember. hmm 

i think this weekend i need to materialize this BNF latest. then asap move to semantics/intentionality based tasks. 

just thinking how an ai could invent mutexes/semaphores right now. i mean mathematical representation of that problem and how ai would have had solved/designed.

 

i mean when thinkign a coder ai, one should not just generate a code generator, but should have capability set of solving easy or hard  problems with designs it would design, hmm. ok. this be another thought exercise to think of. of how this ai is built.

or make itunderstand index structures of relational dbs so that it creates relational tables with correct index settings and make it understand index concept. 

or for instance make it understand map reduce functionalities in elastic search's js layer. or that crystallize information related to a specific db system  and create code along with that knowledge or understand  or understand need of some factory or creational patterns or understand object oriented or alternative patterns to generate code in ways other people would understand.

how could ai had invented object oriented approach? hmm this be another thought exercise or another design requirement as constraint to design of this system. 

it should be able to invent such stuff itself. if it were going to be a coder ai right? ok. it seems possible. ok.

ok tonight's rest of time is planned as:

applying to jobs. and finishing this topology related study with printing out a topology lecture notes. i have one book at another city and one hjere but here its algebraic topology and more advanced discussions in topology area. so but i need basics.  that i would print some lecture notes of and intending to finish today.

 ok lots of tasks remains but i think would be finished at least before new year. mid october might be kind of not enough time frame :S anyway. lets iterate fastest as much as possible in recent month's end and also new month.

new mont would be of healing stomach illness, running jogging by seaside, swimming in pool, attending excursions, attending events and job applications. 

hmm both holiday and both studying period it would be and i need to iterate very much in nlp related challenges in following weeks.  semantics related challenges are difficult but its ok, manageable level i think. hmm so it would be a lot studying whilsts its also holiday, next month and the other month i guess, would be alot include alot studying beside being also holiday sionce i would attend excursions/events/do swim/do jog by sea / attending sailing events to see dolphins/watching stars/sky events/off road traveling events/learning local food cooking styles/yoga by sea events etc etc. 

 

i guess by end of october i would have some mvp at least but not the finshed product.  but some features present version. than would become to have more crytallized information like general ai as would be studied to coding and i think by end of noivember it might become a coder in some areas. hmm. ok. 

as told i wont disrupt coding industry any. its for to get singular not much jobs and code them. i have no intentionality any of disrupting coding sector with such product. but then i could study together on qm simulator product with it's help in coding that alot kloc projects. ok. cool.