I listen to a lot of stories off of Utube while I work. It keeps my mind calm and focused on what I do when it mostly is do it again Sam for the 5000 time. It becomes quite obvious that it is A.I. reading it when words are mis-said like wound and wound. One is an open gash, the other is wrapped around something. The A.I. doesn't know the difference from context. Live and live are another as in a-live or live-ing space. Again context says what it should be but A.I. is stupid and can't tell the difference. The text following along is worse than the CC you are all talking about. For the most part though I only notice that once in a while if I look down at the screen to see how much longer the story is going to run.