I wish modern "closed captions" were accurate. It is one thing to have a computer screw up voice recognition on a live broadcast. It is another thing to allow the same on a prerendered video in which somebody has a chance to check the input and the output.
No I mean English subs on original English audio in new videos. I have my own rants about "localization" which aren't relevant here but I will offer you some jelly donuts in exchange.
