Amazon has announced work on a feature that can synthesize short audio recordings from the human voice and reprogram them for longer conversations. This was announced at the conference re:MARS. Amazon’s Senior Vice President and Head Scientist for Alexa, Rohit Prasad held a demonstration where, according to the description of TechCrunch, “the voice of a deceased loved one (a grandmother, in this case), is used to read a grandson a bedtime story”.
Prasad noted that the company can make such audio from about a minute of conversation. According to him, they see this as a task to transform the voice, not a way to reproduce speech.
Apart from this demo, the company does not disclose many details about the future feature. According to Prasad’s words, the goal of technology will be to “prolong the life of memories” because “many of us have lost those we loved.”
It’s like smart speaker Takara Tomy, who could imitate his parents’ voice before bed to they could read to their children even from a distance.