Artificial Intelligence

Out of context: Reply #1732

  • Started
  • Last post
  • 2,577 Responses
  • ********
    -1

    ********
    -1Permalink
    UpvoteDownvote
    Flag
    • Going forward, I guess we will be using left or right models, depending on biases and mood

      < although I suspect this is from Grok 3's "unhinged" mode
      ********
    • But a model trained on Right's news sources, 4chan, 8chans, all-the-chans, even reddit will give you this kind of reply

      Biased data in > biased replies out
      ********
    • SURE, JANutopian
    • Jan, lol, good onemonospaced
    • Ah yes the two stooges come to the AI table again with nothing to add, just to downvote lol
      ********
    • The idea of left and right is flawed and not fit for use anymore. Both left and right biases have crossovers making it hard to use as group identitiesIanbolton
    • it seems lazy to group people as left and right now. It's unhelpfulIanbolton
    • Models get there from training data
      ********
    Show [[ numHiddenNotes ]] more notesAdd Note
    SaveCancel

View thread