• catloaf@lemm.ee
    link
    fedilink
    English
    arrow-up
    26
    ·
    2 months ago

    Maybe we shouldn’t be treating text generators as sources of truth.

    Isn’t there some liability for someone who provides inaccurate voting information? Perhaps that could be used to influence Google et al. to stop providing AI summaries on their results pages.

    • roofuskit@lemmy.world
      link
      fedilink
      English
      arrow-up
      6
      ·
      2 months ago

      Yes, there is. Even Xitter quickly changed their LLM to point to a government website whenever voting questions were asked so they have no liability.