It’ll work great I’m sure with time, it’ll be better at transcribing than any other model it’ll be better at speech synthesis, understanding etc