Amazon has unveiled plans to develop a system that would allow its voice assistant Alexa to mimic anyone’s voice after hearing a short audio recording. The new feature was announced at the company’s re:MARS conference in Las Vegas on Wednesday.
Amazon Senior Vice President and Chief Science Officer of Alexa Rohit Prasad introduced the new feature to conference attendees and explained that the goal of this system is to “Making memories last.”
While demonstrating the feature, a child asked the voice assistant if “Can Grandma finish reading the Wizard of Oz to me?” » After a few moments, Alexa confirmed the command and apparently started speaking in the child’s grandmother’s voice.
Prasad said this new feature, which is still in development and has no announced release date, required inventions where “We had to learn how to produce high quality vocals with less than a minute of recording versus hours of recording in the studio.”
He explained that the way they achieved this was to define the problem as “a voice conversion task” and not a “speech generation path”, noting that “We are unquestionably living in the golden age of artificial intelligence, where our dreams and science fictions are becoming reality.”
Amazon’s VP suggested that while the feature could apparently be used to reproduce any voice, it could also be used to help commemorate a deceased family member. “Many of us have lost someone we love” during the Covid-19 pandemic, Prasad said, noting that making artificial intelligence conversational and companion-like has become a key goal for the team.
“While AI can’t take away that pain of loss, it can certainly make memories last,” he added.
Amazon said it wants to make its Alexa voice assistant more natural overall and has been constantly improving the program’s functionality. The e-commerce giant has rolled out features that allow it to have more human conversations with consumers, even going so far as to ask questions.
You can share this story on social media: