The realism of artificial intelligence has reached a new level – The Times Hub

Illustrative photo/rawpixel.com

Sonantic startup engineers taught artificial intelligence incredible speech realism thanks to the introduction of very important changes in the program. The technology is already being used in games.

The technology that imitates the human voice was shown on February 14, teaching how to confess one's love. It is simply impossible to distinguish it from a real person.

Demonstration of the work of artificial intelligence: video

The artificial voice in the video sounds quite natural and very realistic, especially the sigh and laughter that fit perfectly into her speech. That's why it's so surprising when she suddenly confesses: I'm not real. I was never born. And I'll never die. Because I don't exist.

Remind. Startup Sonantic announced the creation of a neural network to generate speech with imitation of human emotions in 2020. For two years, the developers have made great progress. In the first commercials, IS's voice did not sound very natural, and the speech was accompanied by distortion. The company is already cooperating with game studios. Her work was used by Obsidian Entertainment to generate the voices of some minor characters in The Outer Worlds. In August 2021, a startup used technology to replicate the voice of actor Val Kilmer, who lost the ability to speak as a result of laryngeal cancer.

Excerpt from:

The realism of artificial intelligence has reached a new level - The Times Hub