Amazon demonstrates Alexa reading a bedtime story in the voice of a boy’s late grandmother


Amazon’s intelligent, voice-controlled assistant Alexa has become an integral part of everyday life. Alexa receives more than 1 billion requests a week, Amazon said Wednesday, while customers have access to more than 100,000 Alexa skills.

Now the tech giant is developing a new skill for Alexa so it can help you remember loved ones who have passed: the ability to communicate with the voices of others. On Wednesday, Amazon’s Rohit Prasad briefly described the capability at re:MARS (Amazon’s Machine Learning, Automation, Robotics, and Space Event).

He showed a short video of a boy talking to an Amazon Echo speaker. “Alexa,” the boy asks, “Can grandma finish reading ‘The Wizard of Oz’ to me?” A woman’s voice begins to speak and Prasad confirmed that the voice should be his late grandmother’s.

“One thing that surprised me the most about Alexa is the camaraderie we have with it,” said Prasad, Alexa AI SVP and Chief Scientist. “Human qualities like empathy and affect are key to building trust. They have become even more important during these times of the ongoing pandemic when so many of us have lost someone we love. While AI can’t erase that pain of loss, it certainly can keep their memories.”

Prasad didn’t say when the capability will be available – he said Amazon is working on it. An Amazon representative told ZDNet that it has yet to announce when it will be available.

Many questions have already arisen the ethics of replicating the voice of a real personbut Amazon’s Nate Michel told ZDNet that it’s “early days” and that the technology is “still exploratory at this point.”

Generating such a voice presents a technical challenge, Prasad explained in his remarks, as it requires recording a high-quality voice in less than a minute, as opposed to hours of recording a voice in a studio. Prasad’s team took up the challenge thereby as a language conversion task and not as a language generation task.

“We are undoubtedly living in the golden era of AI, where our dreams and sci-fi become reality,” Prasad said.

To make Alexa even more human, Prasad shared how Amazon is building generalizable intelligence into the tool. Generalizable Intelligence comprises three key attributes: learning across many different tasks, continuous adaptation to user environments, and learning new concepts through self-monitoring.

Amazon is working on approaches like think before you speakin which Alexa effectively uses “implicit commonsense knowledge” (built with a large language model and a commonsense knowledge graph) to generate answers for a user.

For example, on Valentine’s Day, if a customer says, “Alexa, I want to buy flowers for my wife,” Alexa could use world knowledge and time context to respond, “Maybe you should get her red roses.”

Source link

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.