9 Matching Annotations
  1. Last 7 days
    1. No articles retrievedbetween 1990-2000 met inclusion/exclusionsearch criteria for the final review.

      They may have had more luck in findings pieces that suited their criteria if they allowed non-academic works. The 90s have been referred to as the "Gold Age" of print (e.g., news papers, magazines, etc.) as a cultural influence. They were daily staples in peoples lives and could have included incredible potential for this review.

    2. However, LGBTQ+ topics rank low onthe list among other diversity topics

      I wonder if this remains an issue of performativity on behalf of the university (similar to what was described during Thursdays readings), or if the policies in place are uninformed/ ineffective. I have seen people describe Universities as "all lip service" when they assume that hanging pride flags during June makes them "inclusive". It's incredibly disingenuous on the part of the school.

  2. Jul 2026
    1. actually they are a societal “superstructure” that enforces the stability of a particular system.

      If these AI driven algorithms are pushing these particular beliefs, I can imagine people having biases being reinforced and keeping this oppressive system in power. If we are continuously made to believe that this is "just the way things are" by those in control, how can we as a society feel capable of creating change?

    2. Valuing this data

      I don't think it is just as simple as valuing data; it is who values it. If biases are so ingrained in LLMs to this degree (and potentially worse in the future [not very a very optimistic take, but thinking on least favourable outcomes]), who knows how it may radicalize this generation of children and youth growing up with this digital landscape? It is so easy to fall down rabbit holes of hateful extremism via algorithmic funnelling, that truly anyone could be victimized by, and if those who are creating these platforms harbour those beliefs (or make it easy enough for these dangerous ideologies to infiltrate them), I have to ask perhaps a conspiratorial question: what is the true motive behind all of this?

    3. “low resource”

      Reading texts such as this is a very real point of privilege for me. As someone who has grown up so far removed from the issues raised in this article, its genuinely startling. This brings the question of who get's to determine these rankings of language, who is placing the value of use? This reinforces Sam Altman's quote regarding the selling of knowledge. In this case, I feel as though the primary customer base would begin to become more and more exclusive as time went on. Again, we can grapple with the idea of the right to knowledge. Thinking even about the UN sustainability goals (4: quality education to be specific) set for 2030, is this not directly going against humanitarian rights?

      (link for more info on SDGs if anyone is interested: https://globalgoals.org/goals/4-quality-education/)

    4. LLMs will happily disguise those useful absences with opinions of how Americans imagine Indonesians see the world

      It feels like these particular LLMs are a form of digital colonization. We are unknowingly fed information about different cultures and ways of being through a westernized lens, potentially developing a bias similar to that of the LLM, and possessing an unknowingly false representation of another country/ culture. The danger of inaccuracy is how hate spreads as a whole.

    5. Don’t encourage users to commit suicide

      One of my first "culture shocks" of the potential harms of AI was learning about several stories where young teens had taken their own lives in connection with their AI usage. This ranged between researching methods without the content of the conversation being flagged, acting as a "suicide coach", and incidences of AI psychosis. Much of the backlash for the families and public sparked companies to redesign the platforms, but the concern still remains as to if there is enough being done to protect teenagers/ children that are using these tools. Should there be parental controls on AI platforms? An age restriction?

    6. significant biases

      This is yet another reminder that nothing is truly neutral. It raises questions of who is actually in control of what we see? Is what we are seeing accurate/ factual, or is it manufactured by those who are the figure heads of these larger organizations? The idea that these biases are so ingrained within something so widely used is incredibly dangerous. Who is incharge of the narrative of what we consume online?

    7. squeezing it down

      The phrasing of "squeezing it down" implies a condensing of culture - when I first read this, it felt as if LLMs are essentially breaking down the vast mosaics of what makes different cultures so valuable and simplifying it. Is this not a dehumanization of cultures? While I can understand and appreciate the complex processes that go into the creation of these algorithms, is it worth potentially losing a piece of humanity?