How many times have we wanted to hear, even if just one more time, the voice of someone we loved who passed away? Alexa, Amazon's virtual assistant, will be able to give its customers that opportunity in the near future through audio messages that will bring back to our ears the person we miss so much.
Using Artificial Intelligence (AI), application engineers will be able to imitate the voices of deceased people whom we wish to hear again in audio messages lasting up to one minute.
“Many of us have lost a loved one, and while AI can’t take away the pain of loss, it can make memories of them last,” said Rohit Prasad, Alexa’s senior vice president and chief AI scientist, during his keynote address at the re:MARS conference in Las Vegas, Nevada.
The development of what will become one of Alexa's many features is based on replicating the speech of a dead relative, based on less than a minute of recorded audio from the original person.
During Amazon's global AI event for machine learning, automation, robotics and space ‒re:MARS‒, the executive gave a demonstration of what the assistant will be able to do through an Echo Dot.
“Alexa, can Grandma finish reading me The Wizard of Oz?” Prasad asked out loud. The assistant, in her typical voice, replied, “Okay!” Then, when the device began narrating a scene from the Cowardly Lion calling for courage, Alexa’s robotic tone was replaced by a more human-sounding narrator.
“Instead of Alexa’s voice reading the book, it’s the voice of the child’s grandmother,” Rohit Prasad enthusiastically explained.
The goal of the Alexa feature, Amazon's still-in-development virtual assistant, is to build greater trust with users by infusing artificial intelligence with the "human attributes of empathy and caring."
The engineer mentioned that Amazon developers learned to “produce a high-quality voice with less than a minute of recording” in addition to focusing on the ability to reproduce the voice of a deceased loved one as a conversion task rather than a speech generation task.
Rohit Prasad stressed that the device's advances in functionality have come from the same needs that users have pointed out over time.
Although he did not indicate when this function would be available, he did emphasize that they are working on it.
“We are certainly living in the golden age of AI, where our dreams and science fiction are becoming reality,” Prasad said.
Not everyone was enthusiastic about the announcement, with experts saying that while the prospect of hearing a dead relative's voice can be heart-warming, it also raises a host of ethical and safety concerns.
"I don't feel like our world is ready for easy-to-use voice cloning technology," Rachel Tobac, chief executive of San Francisco-based SocialProof Security, told The Washington Post.
He further explained that if a cybercriminal can easily and credibly replicate another person's voice with a short voice sample, they can use that sample to impersonate other people, which can lead to identity theft, fraud, data loss, account takeover, and more.
With information from The Washington Post.
You may be interested in: Burlingame to be home to Meta's first physical store
