I guess that, these ML approaches are trained on lots of existing text, and therefor derivative works are to be expected?
I guess that, these ML approaches are trained on lots of existing text, and therefor derivative works are to be expected?