Amazon is working on new technology to synthesize a person's voice. The technology aims to make memories last, so one can save the voice of the people dear to them and listen to it whenever they are apart by distance or death.
Amazon Working on a New Technology to Synthesize a Person's Voice
Amazon announced at the re:MARS 2022 conference that it's working on a feature that could synthesize a person's voice in a short audio clip and then reprogram it as a long speech. The company's senior vice president and head scientist for Alexa, Rohit Prasad, demonstrated features at the event.
In the demonstration, viewers watched a child requesting Alexa to have his grandmother finish reading him "The Wizard of Oz" in a video clip. Following the boy's request, the digital assistant reads him a piece about the Cowardly Lion in what is probably the grandmother's voice.
Prasad pointed out that the company could produce this audio output with just one minute of talking. He said they made it happen by framing the problem as a voice conversion task rather than a speech generation path.
"We are unquestionably living in the golden era of AI, where our dreams and science fictions are becoming a reality," Prasad said.
He also highlighted the technology's potential to evoke memories of deceased loved ones before the clip. According to him, the technology was intended to make memories last because many people have lost someone they love.
Currently, information is limited. There is no schedule yet or detailed information.
Biased Opinions of Users on Artificial Intelligence Products
The technology divides people's opinions because of how the interaction with the disease works with someone actively working through grief. It is the question asked by Florence Ion of Gizmodo.com.
Pcgamer writer Ted Litchfield wondered about the unfortunate implications of a loved one's voice being given to a digital assistant.
"Oh yeah my nona lives in a plastic box on the shelf and reads the New York Times daily brief to me," he said.
Litchfield also added that he couldn't wait for his voice to one day be sold as a product to his loved ones after his passing.
On the other hand, Takara Tomy was launched for the benefit of the children and parents who are sick of reading books repeatedly.
Another smart speaker, Coemo, is launched to entertain children with 60 different stories and songs worldwide. Children could listen to two nursery rhymes, songs, Grimm's fairy tales, and Japanese folktales.
Coemo allows the users to create copies of their voices. However, it doesn't produce a perfect copy but at least closely matches how the parents speak with an indication of their unique infections and intonations.
Artificial Intelligence: Deepfake Technology Threat
Although Takara Tomy's new AI-powered smart speaker may calm a child who wants to hear his or her parents' voice at bedtime, some users would not buy the idea, especially when it threatens their financial privacy.
In 2019, a deepfake vocal copy of a CEO of an energy company's voice imitation was used to convince an employee to transfer $243,000 to the scammers. In 2021, another news story was published after an AI-cloned voice was used as part of a $35 million heist.
Check out more news and information on Technology in Science Times.
