I wonder how this works. It sounds like it's transcript driven, but then the next question is - were the transcripts automatically created or user-defined? If

I wonder how this works. It sounds like it's transcript driven, but then the next question is - were the transcripts automatically created or user-defined? If the former, is this not going to run into the same issue as training AI on datasets created by AI? I experience so many mistranslated words when using automatic transcripts that I can't imagine the quality of data is excellent without supporting the transcripts with video inference.