Microsoft Corporation MSFT unveiled a text-to-speech artificial intelligence, or AI, model that can generate realistic voice imitations using a three-second audio sample. In contrast to how ...