22 pointsby abbe989 hours ago6 comments
  • internet_points7 hours ago
    > Let's follow one example: Nigeria is the most populous country in Africa. In Abstract Wikipedia, this might be stored as: Z27243(Q1033, Q138758272, Q6256, Q15, Z27243K5)

    Haha that's like John Wilkins' "Real Character, and a Philosophical Language"

    https://en.wikipedia.org/wiki/La_Ricerca_della_Lingua_Perfet... is a great intro to the weird and wonderful world of abstract/universal/ideal/a priori languages.

    • zozbot2345 hours ago
      It's not that different from how LLM tokens work, only in a tree structure as opposed to a plain sequence. Having a tree structure makes it easier to formally define rewrite rules (which is key for interpretability), as opposed to learning them from data as LLM do.
  • rustyhancock6 hours ago
    One issue with projects like this just show me what it is on the front page.

    Even the featured article section is empty!

    • brettermeier6 hours ago
      Correct, what is this and why is this?
  • orbital-decay5 hours ago
    Are they trying to reinvent Cyc?

    https://en.wikipedia.org/wiki/Cyc

  • zinekeller6 hours ago
    For context, this was proposed way back in 2013 (https://meta.wikimedia.org/wiki/Abstract_Wikipedia), when machine translation is just plain bad (and LLMs are only known in academic circles). Surprised that AWiki is now active though.
  • 5 hours ago
    undefined
  • casey27 hours ago
    So rather than machine translation... really primitive machine translation with extra steps?
    • Hasslequest7 hours ago
      Would you rather use a compiler, or have an LLM generate assembly code based on source code?

      The purpose is to establish a new high-level lanugage