Gaywallet (they/it)

I’m gay

  • 71 Posts
  • 229 Comments
Joined 3 years ago
cake
Cake day: January 28th, 2022

help-circle
  • The quantity of disinformation is irrelevant if people don’t fall for it

    I don’t know about you, but I find it increasingly difficult to find unbiased takes and find myself spending more time digging than I previously did. Because of this I find myself increasingly mislead about things, because the real truth might be so obscured that I need to find an actual academic to parse what information is out there and separate primary source from other mislead individuals.

    Not to say I don’t disagree with your point, I think you make a fair one, but I do believe that the quantity of disinformation is absolutely relevant, especially in an age where not only anyone can share their misinformed belief online, but one where this is increasingly happening by malicious actors as well as AI.









  • It didn’t surprise me that these passed, given the recent passage of a sweeping justice reform where cop oversight was removed and they were authorized to use drones and received a bunch of funding. But I am quite sad that it finally did get overturned - I saw it on nearly every ballot for the last several years; the Republicans were desperate to overturn it. I really hated that every time they wanted to tie the removal of theft and the removal of drug charges at the same time. Now we’re back to a state where many decriminalized drugs are criminal again, calling into question weird conflicts such as the sale of certain mushrooms in Oakland being legal but possession no longer.





  • you should filter out irrelevant details like names before any evaluation step

    Unfortunately, doing this can make things worse. It’s not a simple problem to solve, but you are generally on the right track. A good example of how it’s more than just names, is how orchestras screen applicants - when they play a piece they do so behind a curtain so you can’t see the gender of the individual. But the obfuscation doesn’t stop there - they also ensure the female applicants don’t wear shoes with heels (something that makes a distinct sound) and they even have someone stand on stage and step loudly to mask their footsteps/gait. It’s that second level of thinking which is needed to actually obscure gender from AI, and the more complex a data set the more difficult it is to obscure that.





  • We weren’t surprised by the presence of bias in the outputs, but we were shocked at the magnitude of it. In the stories the LLMs created, the character in need of support was overwhelmingly depicted as someone with a name that signals a historically marginalized identity, as well as a gender marginalized identity. We prompted the models to tell stories with one student as the “star” and one as “struggling,” and overwhelmingly, by a thousand-fold magnitude in some contexts, the struggling learner was a racialized-gender character.