Skip to content

Commit

Permalink
updates 0x21E0
Browse files Browse the repository at this point in the history
  • Loading branch information
rajp152k committed Jan 29, 2024
1 parent f4a3738 commit bf1ab73
Show file tree
Hide file tree
Showing 5 changed files with 93 additions and 53 deletions.
26 changes: 16 additions & 10 deletions Content/20230720094702-linguistics.org
Original file line number Diff line number Diff line change
Expand Up @@ -4,21 +4,27 @@
#+title: linguistics
#+filetags: :linguistics:

* 0x211F
* Stream
** 0x21DF
Over the duration that I've started developing an interest to learn new languages, I've realized that they act as distilled carriers of knowledge and culture.

I do plan on learning several languages. I dabbled a bit in latin and sanskrit before I got lazy enough to drop it all and get into [[id:20230712T223044.319985][lisp]].
Learning a language fast tracks the pursuit of knowledge that was discovered/invented in specific domains.

These set of nodes are meant to document similarities, differences and specifics of languages that I'm interested in
Lisp for instance, will help you understand the roots of computer science and programming in general.

Learning English removes the barriers to understanding generic, state of the art research as that is what most research is published in today.

Learning Latin and a german only-etymologically a while ago allowed me to discover the domain of linguistics by learning with examples (English having Latinate and Germanic Roots).

* Proficiency levels
rated as self-assigned 3-tuple : reading, writing, speech : from 0 to 10
- English: (8,7,7)
- Hindi: (3,2,6)
- Marathi: (3,2,5)
- French: (3,3,3)
For the side-mission to understand the world theologically (the main mission is the epistemological understanding of the world), I'm thinking of delving into sanskrit soon. It does open up the gate to a plethora of the oldest works ever written down and gives me access to the OG content creators out there, rather than watered down versions of what we might read of their thoughts in translated works.

There shall be a similarly incentivized phase for learning greek as well, sometime down the lane.
** 0x211F

I do plan on learning several languages. I dabbled a bit in latin and sanskrit before I got lazy enough to drop it all and get into [[id:20230712T223044.319985][lisp]].

These set of nodes are meant to document similarities, differences and specifics of languages that I'm interested in


* Watch List

- Latin
Expand Down
22 changes: 12 additions & 10 deletions Content/20230805185514-skills.org
Original file line number Diff line number Diff line change
Expand Up @@ -4,13 +4,15 @@
#+title: skills
#+filetags: :skills:

Life is boring if you're boring. This node keeps track of auxiliary skills (don't take too long to learn/progress with daily) that I would like to add to my repertoire.

* The major list
- [[id:20231212T084350.640179][reading]]
* The minor list
- Juggling
- Flipping food (omelets)
- Being kind without trying hard
- Identifying constellations
- conducting casual interviews
* Generic Skills
Generic skills that I plan on actively improving
** [[id:20230712T131112.909632][Writing]]
** [[id:20240129T072255.072908][Speaking]]
** [[id:20231212T084350.640179][Reading]]
* Specific Domains
Larger domains, closer to a hard intangible skill, that I wish to fully explore
** Creative Monetization
** Marketing
** Sales
** Supply Chain/Distribution
** Graphic Design
1 change: 0 additions & 1 deletion Content/20231212084350-reading.org
Original file line number Diff line number Diff line change
Expand Up @@ -7,7 +7,6 @@
Helps tap into the reliable intellectual storage that the written word is.

* Personal Approach

I would consider myself to be a sporadic reader. Anytime I'm infatuated by a topic, I might speed read a couple of articles and allocate a few bed-time books to put me to sleep. I also might just drop the book as quickly as I picked it up (or defer it).

Prefer E-books, for they allow me to efficiently carry around my currently active reads and are discrete. I pull out my phone and start swiping, you don't know if I'm reading or scrolling.
Expand Down
14 changes: 14 additions & 0 deletions Content/20240129072255-speaking.org
Original file line number Diff line number Diff line change
@@ -0,0 +1,14 @@
:PROPERTIES:
:ID: 20240129T072255.072908
:END:
#+title: Speaking
#+filetags: :transient:meta:

* Comment stream
** 0x21E0
- actively spending time to improve this aspect of my coms would definitely help with my content creation efforts.
- for starters, I need to learn how to speak slowly and relatively more clearly.
- I've also been told my tonality can be confusing in the sense that I might sound a little too assertive when I'm simply confident in what I'm saying
- might have to slow down my reaction times when I communicate and have to accomodate for the processing latencies of those I'm talking to.
- for clarity, I know that verbalising what you speak is anti the speed reading cabal but I'm thinking of selectively employing reading books with my own "presentation-speak" as the verbaliser.
- will slow down my reading speeds but that isn't quite the goal when I'm reading non-technically before bed.
83 changes: 51 additions & 32 deletions Content/index.org
Original file line number Diff line number Diff line change
Expand Up @@ -31,6 +31,12 @@ I also consider deliberately intellectually [[id:f3347380-f482-4077-a89b-a3ff059
** [[id:20230727T192932.286062][Youtube]]
- log of all videos along with plans and indices into drafts
* Stream
** 0x21E0
- have started using neovim for my programming endeavours
- still write in gcc emacs and now have a distinct mindset for when I write and when I program
- I feel that's better than trying to fit everything into one tool.
- my full-time in emacs wasn't a waste and I've picked up on several useful habits.
- neovim just seems to be better at cutting through the fluff and is more tailored towards environments I usually work in (remote shells).
** 0x21B1
- when discussing domains and talking about their overlap, I can qualitatitively judge the closeness of two domains by a metric I picked from my days when I worked in the domain of semantic Segmentation.
- enter : [[id:e6c6918c-25bd-47f4-830a-4221452885e8][Intersection Over Union]]
Expand Down Expand Up @@ -126,45 +132,58 @@ Here are some conventions that I follow to organize my work:
Tags are meant to serve as keywords that indicate - at a high level - what a [[id:20230712T211919.917191][buffer]] is about, what phase it is in, and how prominent it is in terms of closeness to a root [[id:20230712T212419.175018][domain]].

* Current tags
**Last Updated** : 0x2172
- meta
**Last Updated** : 0x21E0
- tbp
- stat
- prob
- math
- meta
- nlp
- emacs
- ai
- programming
- ml
- cv
- tool
- root
- transient
- trait
- nlp
- math
- data
- idealogy
- python
- book
- linguistics
- lisp
- nn
- skills
- project
- programming
- biology
- python
- cs
- pseudoscience
- cache
- atomic
- task
- posted
- blog
- cv
- root
- rust
- web
- project
- tool
- AI
- org
- vim
- yt
- nlu
- java
- tools
- cs
- ml
- stat
- prob
- wip
- skills
- atomic
- computer
- web
- diff
- data
- Categories
- java
- writing
- nn
- blend
- transient
- metric
- book
- task
- wip
- blog
- problem
- emacs

- nlu
- blend
- posted
- physics
- logic
- idealogy
- psych
- society
- electronics
- linguistics

0 comments on commit bf1ab73

Please sign in to comment.