Most of us are used to asking Alexa today’s weather forecast or to add coffee to our grocery list, but we may soon be able to get a response back in the form of a dead relative’s voice. That’s right, folks. Amazon unveiled new functionality for its virtual assistant, one that will sound eerily similar to a deceased loved one.
Amazon’s senior vice president and head scientist for Alexa, Rohit Prasad, announced the unsettling new feature during the e:Mars conference in Las Vegas. Apparently, technologists can synthesize short audio bites of a person’s voice and turn them into longer clips.
“While AI can’t eliminate that pain of loss, it can definitely make their memories last,” Prasad said. In a video played at the event, an Amazon Echo Dot is asked: “Alexa, can Grandma finish reading me ‘The Wizard of Oz’?”
“OK,” Alexa’s voice responded. But “instead of Alexa’s voice reading the book, it’s the kid’s grandma’s voice,” Prasad said, which was developed using, “less than a minute of recording.” Arguably, this could be extended from reading a bedtime story to your deceased grandpa telling you a bad dad joke or your dead mother-in-law continuing to tell you what a disappointment you are even after she’s gone.
People were understandably uneasy about the idea that their deceased loved ones could come back in Alexa form:
Trap your dead grandma in our orb so that she can offer oracular advice in response to your questions and demands!
— Daily Grail (@DailyGrail) June 28, 2022
Occult A.F.https://t.co/rE6IexviRx
"Alexa, order pre-made shortcrust pastry"
— Sam Whyte (@SamWhyte) June 23, 2022
*Dead grandma's voice* "Even in death you disappoint me" https://t.co/1HG7hG3trn
https://twitter.com/EmilyGorcenski/status/1540719486797434880?s=20&t=4495R0bLVhiZajFaoxVsSA
What
— Geoffrey A. Fowler (@geoffreyfowler) June 22, 2022
Were
They
Thinking
"Alexa will soon be able to read stories as your dead grandma"https://t.co/oHcNnigPEc
Amazon isn’t the first product to use deepfake audio to recreate a family member’s voice, but it may be the first to do so without their consent. The Takara Tomy smart speaker uses AI to read children bedtime stories using a parent’s voice (like, say, if their parent is away for extended periods of time). But the parents upload their own voices, meaning they are very much a part of the process.
“In this companionship role, human attributes of empathy and affect are key for building trust,” the exec continued. “These attributes have become even more important in these times of the ongoing pandemic, when so many of us have lost someone we love.” He added that the feature “enables lasting personal relationships.”
The feature, he noted, is still in development, and no word has been given (from here or beyond) as to when it may publicly launch.
