The American company Meta has announced that it has developed a new model of generative artificial intelligence, which can reproduce the voices of dead people as well as living people based on sound files.
It is noted that the AI model called Voicyebox can speak six languages. The Meta company hailed this development as “a breakthrough in generative artificial intelligence for speech.”
According to the description of the model, artificial intelligence can also imitate the human voice in foreign languages. According to the company’s representatives, this will help dubbing movies and allow content creators on the Internet to translate videos into other languages.
He added that the developers decided to close public access to the model, considering the potential risks of the technology.
“There are many interesting uses of the generative speech model, but due to the risk of misuse, we are not publishing the Voicyebox model or code at this time,” a company representative said.
Meta concluded that a balance must be found between openness and accountability in the development of artificial intelligence to prevent the technology from being used for nefarious purposes.