OpenAI presented a neural network for cloning a person’s voice
[ad_1]
Artificial intelligence startup OpenAI releases preview of digital voice generator
Artificial intelligence startup OpenAI has released a preview of a digital voice generator that can reproduce natural-sounding human speech from a single 15-second audio sample.
How reports NBC, the software is called Voice Engine. In their blog, the creators of ChatGPT say they have tested the generator for a variety of purposes, including helping children read, language translation, and voice restoration for cancer patients.
Discussing the new product, some social media users highlight potential abuses, including potential fraud using unauthorized voice impersonations or deepfakes.
In response, OpenAI said that there are no plans to widely release the software due to the possibility of illegitimate use, including during a US presidential election year.
[ad_2]
Source link