During Amazon’s annual re:Mars event, the company announced a new feature that can synthesize short audio clips into longer clips.
According to TechCrunch, a scenario featured at the conference was the voice of a deceased loved one; specifically, a grandmother is used to read a bedtime story to a grandson. Senior Vice President and Head Scientist for Alexa at Amazon, Rohit Prasad, stated that using the new technology, Amazon can just turn one minute of speech into endless possibilities.
“This required inventions where we had to learn to produce a high-quality voice with less than a minute of recording versus hours of recording in the studio, the way we made it happen is by framing the problem as a voice conversion task and not 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 notes.
Prasad further shared that they are working on Alexa to become more human-like. The company is going for a more “think-before-you-speak” approach, where Alexa would use “implicit commonsense knowledge” to generate responses. For example, if a customer on Valentine’s Day was to ask “Alexa, Alexa, I want to buy flowers for my wife,” Alexa would respond with, “Perhaps you should get her red roses.”
Amazon has not revealed a timeline on when we can expect this feature to roll out; the executive claimed that they are “working on it.”
Elsewhere in tech, FedEx will start taking pictures of your delivered packages.