• 0 Posts
  • 631 Comments
Joined 3 years ago
cake
Cake day: July 17th, 2023

help-circle
  • There’s always something that will stick. Unfortunately, the things that seem to always stick are anti-LGBTQIA+ and anti-foreigner. The 80s were very anti-gay with the AIDS scare and (lack of) response, crackdowns, anti-sodomy laws, etc. The 90s continued that, and added on some extra anti-black since they were becoming quite prominent in sports and popular culture. The 00’s added on a nice thick layer of anti-muslim because I guess all brown people were responsible for 9/11 or something? Then the 10’s got some more anti-black with Obama and George Floyd, and the 20’s with queer/trans people.

    Going back we had more cycles like this:

    • 1900-1915: Irish people
    • 1915-1920: + German people
    • 1920-1930: + Jewish people
    • 1930-1945: + German and Asian people
    • 1945-1950: + Russian people
    • 1950-1970: + Asian (Korea/Vietnam) people
    • 1970-1980: + Black people
    • 1980-1990: + Gay people
    • 1990-2000: + Black people (again)
    • 2000-2010: + Arab people
    • 2010-2020: + Black people (again), queer people (again)
    • 2020-present: All of the above, and somehow also white people too?




  • It’s amusing how each generation seems to independently come across the “empty chip bag” meme and the actual explanation - the bag is intentionally oversized and slightly inflated to give the chips a buffer against being crushed. I remember seeing it come up about once a decade, including pre-(consumer)internet.








  • An LLM can only be derivative. It can come up with “unique” patterns, but only based off it’s training data. For something like your use case, it would have very little data to work on given that you claim the documentation is poor. Your best case would be to use a high reasoning model and feed the documentation into it’s context before asking anything - but it will only give you answers based on that input. If the documentation is wrong or outdated, you’ll get answers based on wrong and outdated data.

    As far as “who could get left behind?”, I feel that is more people who’s job can be impacted by LLMs - programming is a big one, and the reason is largely in my first paragraph. The amount of “training data” on the internet for coding is absolutely massive and incredibly well organized because programmers are nothing if not incredibly pedantic. The result is that the task of writing code can be largely offloaded and with recent models, the quality of the code being produced is extremely good, especially in tasks that are well solved/documented.