After the text, the photo, the video… here is the voice. Open AI presented Voice Engine at the end of last week, an audio cloning tool capable of fully reproducing a person's voice from an initial 15-second sample and making them say whatever you want. The results, from what the company has been willing to suggest, are impressive. Artificial intelligence models can now reproduce the nuances, intonations and specific characteristics of a person's voice. Cloned voices are becoming increasingly indistinguishable from the real voice, which is beneficial for many of the positive use cases that OpenAI wants to encourage.
To discover
PODCAST - Listen to the latest episode of our Tech Questions series
But, well aware of the risks that these technologies present in malicious hands, the company is currently restricting its use to a few hand-picked partners to experiment with it. “
We take a careful and informed approach before a broadcast…
This article is reserved for subscribers. You have 80% left to discover.
Do you want to read more?
Unlock all items immediately.
TEST FOR €0.99
Already subscribed? Log in