I enjoyed the piece, but wonder if the author could have found some benefit if they used the corpus to train an (ideally locally-run (for privacy)) ai - so that questions could be asked of it and some value extracted..