• unn@lemmy.ca
      link
      fedilink
      arrow-up
      1
      ·
      edit-2
      4 months ago

      I think there might be some existing algorithms like evolutionary algos that if scaled could led to beating the ARC-AGI challenge for example