OpenAI, U.S based artificial intelligence research firm has introduced technology that can re-make somebody's voice. The high-profile artificial intelligence startup added that a team of organizations was trying another OpenAI framework, Voice Motor, that can re-make an individual's voice from a 15-second recording. Assuming that you transfer a recording of yourself and a section of text, it can peruse the text utilizing an engineered voice that sounds like yours.

 

The text doesn't need to be in your local language. Assuming that you are an English speaker, for instance, it can re-make your voice in Spanish, French, Chinese or numerous different dialects.

 

OpenAI isn't sharing the innovation all the more generally on the grounds that understanding its potential dangers is as yet trying. Like picture and video generators, a voice generator could assist with spreading disinformation across web-based entertainment. It could likewise permit lawbreakers to mimic individuals on the web or during calls. The organization said it was especially stressed that this sort of innovation could be utilized to break voice authenticators that control admittance to internet banking accounts and other individual applications.

 

"This is a sensitive thing, and it is important to get it right," an OpenAI product manager, Jeff Harris, said in an interview.

 

The organization is investigating approaches to watermarking manufactured voices or adding controls that keep individuals from utilizing the innovation with the voices of legislators or other noticeable figures.

 

In February, OpenAI adopted a comparative strategy when it divulged its video generator, Sora. It flaunted the innovation yet didn't openly deliver it. OpenAI is among the many organizations that have fostered another type of AI technology that can rapidly and effectively create engineered voices.