OpenAI, the corporate that makes ChatGPT, is stepping into the voice assistant enterprise and has unveiled new know-how that may clone an individual's voice. However, the know-how won’t but be launched publicly on account of safety considerations.
The new know-how known as Voice Engine, and it was unveiled simply over per week after OpenAI itself filed a trademark software for the title. With this know-how, OpenAI states, it will likely be potential to recreate an individual's voice just by ranging from simply 15 seconds of recording of that particular person talking.
Reserved for “beta testers”
OpenAI says that it plans to preview Voice Engine to some “beta testers”, “but that it does not intend to release this technology widely, at least at this time”, as a result of risks of improper use.
“We know that generating speech in the voices that resemble those of real people carries serious risks, which are especially important in an election year,” the San Francisco firm stated in an announcement.
In New Hampshire, authorities are investigating robocalls despatched to hundreds of voters shortly earlier than the presidential main that featured an AI-generated voice imitating President Joe Biden.
https://www.ilsole24ore.com/art/openai-ha-realizzato-tecnologia-grado-clonare-voci-umane-reali-AFvFGYFD