In times when users can already hold full-fledged conversations with an AI, and when deep fakes are virtually indistinguishable from authentic content, the next step for translating human interactions seems logical: real-time speech translation in your own natural voice.