AI Researcher

26/09

• Praise Meds: Finally able to take my ADHD meds again (after 2 days rest, 1 day break, 2 days rest again). They make me feel "alive" whenever i take them. Im starting to think that they keep me sane during this traumatic period since my dopamine is shot from immense stress -- the meds feel akin to when id play video games to escape my stressful reality, but i can actually make progress to try fix things instead of temporarily escape and return to the original position

• Learned when listing things w/ "etc" the correct punctuation is "[..], etc., [..]" to continue after the list, or "[..], etc. [..]" to end the list. Since ive been writing so much ive discovered an interest to learn proper grammer bc i try to use new words as often as possible (to expand vocab -- seems like ill never "master" English) and formulate what im trying to say better w/ experimentation after reading research papers and books :) which leads me to wondering *"hmm, what is the correct way to do this?" *

• Everyone has their own shit going on and the hardship is relative to their subjective experience. I apologised to some friends for leaving them on delivered by saying ~"sorry for late reply. been emotionally exhausted recently and didn't want to be dismissive".

• 13:22: Reading Becoming a Research Engineer at a Big LLM Lab -- 18 Months of Strategic Job Hunting where a master's student of German Aerospace Center ("research wasn't ML related") became a self-taught MLE, Max Mynter, landing a research eng role w/ Mistral (ML foundation model labs w/ billion $ funding)
  
  

  He mentions he focused on "most notably distributed systems, data structures, and algos" which reminded me i bought the pdf of "Think Distributed Systems" to support the author. I dont usually do that but bc it was a book on distributed systems i had to -- ive been itching to learn it low level for some distributed AI experiments so i thought id pay my respects.

• 14:02 Just finished reading it. It was mostly an article towards optimising for successful MLE job application. I thought it woud of been about the experience of creating cool custom AI projects. Bit disappointed, but extracted some useful notes:

  • "[..] strongly considering applying to role x. Is there anything you can share to help me make the best possible application?" and "concrete offers gave a lot of signal"
  • imagine a hiring manager reading your CV on their phone, glancing at it from time to time. therefore, one column pdf with bold, italic, underline, hyperlinks, + minimal language., easy to skim. I'd say projects >> work experience > skills > education

• Random notes from today:

  • people who don't truly care about your problems/trauma/feelings aren't going to read this career blog, so i can speak my mind and only people who give a shit will read. thats the beauty of public online autobios. the wrong people aren't willing to read it all to use it against you, unless you've really done horrible things to them. -* "People can only find you interesting if they know what makes you interesting. So post about your projects"* this was reinforced to me a few days ago when i was reflecting on my extremely lonely upskilling phase -- all of my friends were gained during the period of marketing my programic progress on x. but now im not working on anything there isnt as much "action". im aware it will be revived once i begin my experimental phase :)