In the 2013 movie Herthe character played by Joaquin Phoenix falls in love with a female voice generated by artificial intelligence (AI). Now, an application that supposes a revolution in this field has come to real life and has left speechless to experts.
It’s about Skew mean application developed by the co -founder of Oculus, which replicates the human voice with a reliability that makes it very difficult to differentiate from the real.
You can choose the voice of a woman (Maya) or a man (thousands), and although for now They speak in American EnglishThey understand Spanish and even dare to chap.
Sesame’s key is not the absolute fidelity with which human voices replicates, but also the way of speaking: Doubts, imperfections, expressions, pauses … It is even able to interrupt ourselves when we talk.
“I’ve never found anything like Sesame”says Eric Hal Schwartz in his review for Techradar.com. “It is fluid, expressive and unpredictably human. He laughs briefly when he says something slightly funny, he doubts before answering a question and even seems to change his mind in the middle of the phrase, pause and starting a new one,” says the expert.
The key is in the conversational speech model (CSM) of Sesame, which combines text and audio in a single process. This means that it is not limited to generating a prayer and then ‘reading it’, but that it creates a speech that reflects how humans really speak, with pauses, changes in tone and other traits of reality.
“Pauses, hesitations, filling words, how he changed his mind in the middle of the phrase … everything is disturbingly real,” says Schwartz, who adds that “I forgot instantly that I wasn’t talking to a real person. “
Schwartz wonders “If AI speaks in a realistic way, would we even know that we are talking with an AI?” The expert considers that “could Be really difficult Know when we are talking with an AI, at least in short conversations. “
Another analyst, Sean Hollister, explains to The Verge that his proof was “The closest to a real conversation than any other that has so far “with similar tools.
Mark Hachman, in an article to PCWORLDhe says: “Fifteen minutes after ‘the phone with the new realistic of Sesame, I’m still scared. ”
Sesame’s future has no limit. His drivers are working on the project of realistic glasses that allow you to See what you do. The company itself says that talking to Sesame is “crossing the disturbing valley of the conversational voice. “
#artificial #voice #intelligence #put #hair #tip #difficult #differentiate #real #person
