Autogenerate captioning on video or audio files

Autogenerate captioning on video or audio files. This speech to text service potentially saves significant human effort but will work less well in languages that have not yet had the full research effort of - say - English and Chinese, where error rates remain a challenge.