This week, Amazon is organizing the conference re: MARCH (Machine learning, Automation, Robots, Space) in Las Vegas, where they will show some news about the company’s upcoming products and services. Among the news that was presented was a new feature in Amazon’s smart assistant Alexa that lets this one speak in the same voice as a deceased person.
For it all to work, the user must first feed Alexa with an audio recording from the dead person’s voice. Amazon then fixes AI algorithms for a synthetic voice that Alexa can use to talk like the dead person writes TechCrunch. In the fictional example that was shown on re: MARCH, a boy could have a goodnight story read by Alexa. Then the smart assistant used a synthetic voice generated from a recording of the boy’s dead grandmother’s voice. During the presentation, an Amazon spokesperson said:
“This required inventions where we had to learn to produce a high-quality voice with less than a minute of recording versus hours of recording in the studio. The way we made it happen is by framing the problem as a voice conversion task and not a speech generation path. We are unquestionably living in the golden era of AI, where our dreams and science fictions are becoming a reality.”
If or when Amazon plans to release the functionality with synthetic voices from dead people is not something the company has stated so far.
–
