Source Reference: https://www.youtube.com/watch?v=7leZaJuzn6g
Comment Reference: https://www.youtube.com/watch?v=7leZaJuzn6g&lc=UgzUcBNzGb-kRXU9eKF4AaABAg
56:12 In the call itself, you're busy with other things, keeping track of what has been said and preparing responses, and your brain encountering ideas/expressions for the first time, processing the language and no way to jump back, but watching things a second time: your brain state has changed since then, you notice other things not recognized in the call, it's also a different emotional state, not "in it" and the other people present with you, but more of an outside observer of yourself and the others.
Comment Reference: https://www.youtube.com/watch?v=7leZaJuzn6g&lc=UgxkNu2_QDPnRuTG4m14AaABAg
1:06:02 Yes, that's the one the crowdsourcing would work from, to make it perfect, tag speakers, tag topics, links to websites/places/people that were mentioned, organizing them and attaching additional metadata, and/or other things, to be later commented on. Other speech-to-text engines could be used as well, if they're producing a result of similar or better quality.