The Fugatto language model creates a sound you’ve never heard before

Nvidia recently developed a great language model that has the ability to change people’s voices or accents.

Nvidia has unveiled its big language model; This language model was developed for music and sound production applications and can modify sounds and create new sounds. Nvidia’s language model is called Fugatto, but Nvidia has no plans to publicly release this model.

Fugatto’s capabilities include converting text description into sound and music, this language model can even create unheard sounds; Including new sounds like a trumpet instrument barking like a dog.

But the main difference between Fugatto and other artificial intelligence models is its ability to change or modify existing sounds; For example, it can transform a piece played on the piano into a human-like voice or modify a person’s voice and change its accent and expression.

Fugatto is trained on open-source data, and its release date is not yet known. However, it should be noted that any productive technology always carries some risks, because people may use it to produce content that is not appropriate. That’s why Nvidia is cautious about its public release and has no immediate plans to make Fugatto available.

It should be said that the creators of generative artificial intelligence models have not yet been able to find a way to prevent the misuse of this technology, such as deepfaking or copyright infringement.