Amazon is working on giving Alexa the ability to mimic anyone’s voice, dead or alive, from just a minute of audio — and Twitter isn’t doing that…


  • An Amazon executive said his team taught Alexa to mimic voices with short audio clips.
  • The ability can help people remember loved ones who have died from COVID-19, the manager said.
  • Many Twitter users have expressed concerns about possible misuse of the technology.

Amazon teaches Alexa to mimic anyone’s voice, dead or alive, with just a minute-long recording of that voice.

Rohit Prasad, Amazon’s chief scientist for Alexa, said at a live event Wednesday that his team instructed Alexa to take a voice from a short audio clip and convert it into a longer audio output. Prasad presented at Amazon’s re:Mars conference in Las Vegas.

He showed one short video how people might use Alexa’s voice-changing ability in real life. In the clip, a boy asks, “Alexa, can grandma finish reading The Wizard of Oz for me?”

The smart speaker acknowledged the request with its default chirping voice, then switched to a less robotic voice narrating an excerpt from the children’s novel.

“This required inventions that required us to learn how to produce a quality voice in less than a minute of recording instead of hours in the studio. We achieved this by framing the problem as a voice conversion task rather than a language production pathway,” Prasad said.

Prasad said Alexa’s ability to pose as familiar voices is especially important now that many people have lost loved ones to COVID-19.

“While the AI ​​can’t take away that pain of loss, it can definitely make her memories last,” he said.

Prasad didn’t say when Amazon would unveil Alexa’s voice impersonation capability to the public. An Amazon spokesman declined Insider’s request for comment.

Alexa’s ability to mimic voices is a form of artificial intelligence (AI) that Prasad called “generalizable intelligence.” The ability helps Alexa adapt to different situations and acquire new knowledge from the experiences with little supervision, he said.

It differs from the “all-knowing, all-capable” artificial general intelligence, or AGI, which aims to understand human tasks and intellect to solve problems, Prasad said. Organizations including Google’s deep mind and Elon Musk’s OpenAI both focus on perfecting AGI.

Amazon isn’t the only company working on developing technology that can mimic human voices. Last month, Japanese toymaker Takara Tomy debuted an egg-shaped device called the coemo that copies adult voices and uses them to read stories to children.

Many are startled by AI’s ability to mimic human functions

On Twitter, people were divided over Amazon’s plans to teach Alexa to mimic human voices.

One person who uses the Twitter handle “Maltese Mama” said Alexa can keep her parents who have dementia and who live far away mentally active. “We have caregivers who are on the road daily, but it’s amazing to be able to drop by with a video call, or even better, drop by,” they say tweeted in response to Prasad’s presentation.

But many others raised concerns about the technology.

“Um, how soon will criminals be able to call your family members and ask for Venmo cash with this? Or ask them for social security numbers? Or bank info?” tweeted one user with the handle bitty_in_pink.

Others, including a Twitter user who calls himself “Luke,” said the thought scared them.

“It’s cute but incredibly creepy at the same time…I lost my mom last August and would die to have one last real conversation with her but I wouldn’t do it for a goddamn circular device,” he wrote.

Experts have long been concerned about AI’s ability to mimic human functions. In 2015, Musk funded several AI projects, including OpenAI, to ensure researchers only use the technology for charity. Earlier this month, an engineer claimed so a Google chatbot had become sentientbut AI experts said it was far from confident.

Source link

Leave a Reply

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