Starting today in the US, Alexa will now read out the news in a much more natural, human-like way by selectively emphasizing certain words in a sentence in the same way a real newscaster would. The new voice, which Amazon first announced last November, can be heard when you ask Alexa for your daily news briefing using the query, “Alexa, what’s the latest?”
Amazon created the new voice by using machine learning to analyze audio clips from actual news channels to identify patterns in the way newscasters speak. This “neural text-to-speech” method means that Alexa knows which syllables to emphasize and which to skim past, rather than pronouncing them all equally as it previously did. You can hear the difference in the sample audio clips below.
In addition to the newscaster voice, Alexa also adopts a different style when reading out information from a Wikipedia page. You can hear how this sounds in the clips below.
Amazon has previously experimented with different Alexa voices, such as a whisper mode, but this machine learning method has the potential to quickly train Alexa to speak in a variety of different styles. It took just a few hours of data to train Alexa to speak like a newscaster, Amazon says, suggesting that more voices are a very real possibility.