The Google Assistant is getting more voices thanks to artificial intelligence.
Google (GOOG) CEO Sundar Pichai said Tuesday that the search giant’s voice-activated digital assistant would now include six different voices that were created with the help of artificial intelligence technologies. Sundar made his comments at the company’s annual Google I/O conference for developers in Mountain View, Calif.
Pichai said that the company worked with its AI subsidiary DeepMind to create a technology called WaveNet. This technology, Pichai said, helped Google create new computerized voices that sound “closer to how humans speak.” It’s likely that Google didn’t get rid of human voice actors to help be the “voice” of Google Assistant, but is using the WaveNet technology to augment their recordings so that those recorded voices can respond to more inquiries without the actors having to record each individual answer.
The goal is to make Google Assistant sound more natural during conversations, he explained.
Google showed the audience of coders some demonstrations of the new male and female sounding voices. They sounded a bit more lifelike than the previous voice, and when finishing a sentence, they didn’t abruptly stop, thus ruining the natural cadence of a human conversation.
Get Data Sheet, Fortune’s technology newsletter.
Pichai also said that musician John Legend also recorded his voice to be used for the Google Assistant. The company will presumably use the WaveNet technology to create more possibilities out of Legend’s responses than what he has recorded.
“John’s voice is coming later this year,” Pichai said.
6 new voices coming to the #GoogleAssistant. And, later this year, @johnlegend's, too. #io18 pic.twitter.com/qXSTe6FB0r
— Google India (@GoogleIndia) May 8, 2018
