The exposure that documentary filmmakers used voice cloning software to force the late Anthony Bourdain to say words he never spoke was criticized in ethical concerns about the use of powerful technology. ..
The movie “Road Runner: A Movie About Anthony Bourdain” will be screened at the cinema on Friday, featuring real-life footage of a beloved famous chef and a television host traveling around the world before his death in 2018. However, director Morgan Neville told New Yorker: The dialogue fragments were created using artificial intelligence technology.
This renewed the debate about the future of voice cloning technology not only in the entertainment world, but also in politics and the fast-growing commercial sector dedicated to translating text into real human speech.
“Unapproved voice cloning is a slippery slope,” said Andrew Mason, founder and CEO of voice generator Descript, in a blog post on Friday. It can be ethical and it won’t be long before something happens. “
Prior to this week, most of the general controversy over such technology was Deepfake difficult to detect Use simulated audio and / or video, which can foster false information and political conflicts.
However, Mason, who previously founded and led Groupon, said in an interview that Descript repeatedly rejected requests to regain voice, including voices from “people who have lost and are sad.”
“We don’t want to make that much judgment. We’re just saying that you have to have some bright lines about what’s okay and what’s not,” he said. Said.
The anger and unpleasant reaction to voice cloning in the Bodin case reflects the expectations and problems of disclosure and consent, and is the program director of Witness, a non-profit organization working to use video technology for human rights. Said Sam Gregory. With consent, he said it was appropriate to disclose the technology in the workplace. Instead, viewers expressed their dissatisfaction online, first surprised by the fact that the audio was fake, and then the director seemed to dismiss the ethical question.
“It also touches on our fear of death and our thoughts on how people can control our digital portraits and force us to speak and act without a way to stop them.” Gregory said.
Neville did not specify which tool was used to reproduce Bodin’s voice, but said he used it in a few sentences that Bodin wrote but never said aloud. ..
“In the grace of his property and literary agent, we used AI technology. It’s a modern story used in several places where it was important to bring Tony’s words to life. It was a storytelling technique. “
Neville also told GQ Magazine that he had the approval of Bodin’s widow and executor. The chef’s wife, Ottavia Busia, replied in a tweet: “It was certainly not me that Tony said it was cool.”
While tech giants like Microsoft, Google and Amazon have dominated text-to-speech research, there are now many startups like Descript that offer voice cloning software. Applications range from customer service chatbot conversations to video games and podcasting.
Zohaib Ahmed, founder and CEO of Resemble AI, a Toronto company that sells custom AI voice generator services, said: “Everyone’s voice needs consent when creating a voice clone.”
Ahmed said the rare opportunity to be allowed to duplicate his posthumous voice was for academic research, such as a project using the voice of Winston Churchill, who died in 1965.
According to Ahmed, a more common commercial use is to edit TV ads recorded by real voice actors and add local references to customize them for the region. He also said that it is also used to dubb anime movies and other videos by speaking in one language and speaking another.
He compared it to past innovations in the entertainment industry, from stunt actors to green screen technology.
Professor Rupal Patel of Northeastern University can teach AI systems how to generate their own synthetic speech by creating recorded human speech in seconds or minutes, but Anthony Bourdain’s speech clarity. Perhaps more training was needed to capture the rhythm. We run another voice generator, VocaliD, which focuses on customer service chatbots.
“To be able to speak like him requires a lot of, perhaps 90 minutes of good, clean data,” she said. “You are building an algorithm that learns to speak as Bodin spoke.”
Neville is a highly regarded documentary that directed Fred Rogers’ portrait “Would you like to be my neighbor?” The Oscar-winning “20 feet from stardom”. He started making his latest film in 2019, more than a year after Bodin committed suicide in June 2018.