Beginning today in the US, Alexa will now read out the news in a much more natural, human-like way by selectively focusing particular words in a sentence in the same way a real newscaster would. The new voice, which Amazon first declared last November, can be heard when you query Alexa for your daily news briefing using the question, “Alexa, what’s the latest?”
Amazon developed the new voice by using machine learning to evaluate audio clips from actual news channels to determine patterns in the way newscasters converse. This “neural text-to-speech” technique means that Alexa understands which syllables to highlight and which to read over past, rather than pronouncing them all similarly as it recently did.
Apart from the newscaster tone of voice, Alexa also adopts a different style when reading out facts from a Wikipedia webpage.
Amazon has recently experimented with distinct Alexa voices, such as a whisper mode, but this machine learning method has the potential to fairly quickly train Alexa to speak in a number of different styles. It required just a few hours of data to train Alexa to speak like a newscaster, Amazon says, recommending that more voices are a very real possibility.