Pangram social justice misinformation wars

(statmodeling.stat.columbia.edu)

3 points | by stratos123 4 hours ago ago

2 comments

  • stratos123 3 hours ago ago

    Originial title: "The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hit", had to crop it for submission.

      What I really dislike is the way that people (including academics like Bender) are adding confusion by ignoring (and even actively arguing against) how a positive classification from a classifier like Pangram can have evidential value for some of the decisions people care about right now.
    The author is being very charitable in her interpretation there. I'd go further and suggest that it seems quite a lot of people would be better off ignoring all their instincts about what media is AI-generated and instead outsourcing their judgement to Pangram, since empirically their instincts are not reliable even when it's very blatant.
    • Rochus 3 hours ago ago

      Assuming that Pangram is reliable by concept is yet another category error. I just had a case where it classified all quotations from 1968 in a study as "AI generated".