• lepinkainen@lemmy.world
    link
    fedilink
    English
    arrow-up
    1
    ·
    2 days ago

    Yes, it doesn’t generate funny pictures but they do generate player “stories”, basically synthesising the way players play the game.

    And a multimodal LLM can look at the game and “see” if something is wrong. Like “go here in the map, look here and there should be a monster of type X there”.

    • atopi@piefed.blahaj.zone
      link
      fedilink
      English
      arrow-up
      2
      ·
      2 days ago

      Do you have access to a resource with more detailed information about how this works? It seems really interesting