Anthony Bourdain documentary deepfaking his voice draws mixed reactions
Anthony Bourdain arrives at the Creative Arts Emmy Awards in Los Angeles, held on Sept. 11, 2016. Image: Invision/Richard Shotwell via AP
The use of artificial intelligence to “deepfake” Anthony Bourdain’s voice in Morgan Neville’s recently released documentary, “Roadrunner,” has received mixed reactions.
Some critics find artificially re-creating the late chef’s voice unethical, while others defend the documentary and the artistic freedom in telling a narrative.
Helen Rosner, author of The New Yorker’s July 15 piece on the documentary, for one, took to Twitter to express her shock after finding out that some of Bourdain’s lines did not actually use his voice.
“[Because] of Bourdain’s massive [voiceover] archive, he’s sort of ‘Roadrunner”s ghostly narrator. But there was one line I was amazed they had audio for, and Neville *blew my mind* when he told me that for that line (and 2 others) they made an A.I. of Bourdain’s voice,” she posted on the same day.
Bc of Bourdain's massive v/o archive, he's sort of "Roadrunner"'s ghostly narrator. But there was one line I was amazed they had audio for, and Neville *blew my mind* when he told me that for that line (and 2 others) they made an A.I. of Bourdain's voice https://t.co/HkW7VAtoh8
— Helen Rosner (@hels) July 15, 2021
“This IS very ethically…fascinating. The transcript of this part of our convo is a lot of me going ‘wait, WHAT? Really? Are you serious?'” she said in another tweet.
This IS very ethically…fascinating. The transcript of this part of our convo is a lot of me going "wait, WHAT? Really? Are you serious?" (FWIW, Neville told me they made the A.I. with full green-light from Bourdain's estate.) https://t.co/B7xepmU0i3
— Helen Rosner (@hels) July 15, 2021
Twitter user @angremlin, meanwhile, replied to Rosner, “So people who receive money for its use said it was okay to make the dead guy’s voice say things he didn’t say and absolutely hope people believe he really said it?”
Rosner pointed out, however, that the filmmakers did not make Bourdain “say” things he did not actually say, but rather used audio deepfake to say things he had already written.
“To be clear, he didn’t say them *out loud*. But they’re not fabricated — they’re lines he wrote,” she said.
Rosner then admitted that she wass still not sure how she felt about the use of AI in this case, saying, “People use voice mimicry for v/o all the time, but there’s something about the tech that does feel eerie.”
To be clear, he didn't say them *out loud*. But they're not fabricated — they're lines he wrote.
(I'm not sure where I land on the A.I. tbh. People use voice mimicry for v/o all the time, but there's something about the tech that does feel eerie.) https://t.co/skxsE3SCYP
— Helen Rosner (@hels) July 15, 2021
Sam Gregory, a program director at nonprofit organization Witness, meanwhile, argued, “Faking Bourdain’s voice? Is it appropriate to deepfake a dead person’s voice? Under what circumstances and with what disclosure?”
“We’ve been working on ethics principles around deepfakes @witnessorg. Consent, deceptiveness, disclosure, audience expectations in genre all key,” he tweeted yesterday, July 16.
Faking Bourdain's voice? Is it appropriate to deepfake a dead person's voice? Under what circumstances and with what disclosure? We've been working on ethics principles around deepfakes @witnessorg. Consent, deceptiveness, disclosure, audience expectations in genre all key.. https://t.co/yWMd616Mk5
— Sam Gregory (@SamGregory) July 15, 2021
Podcaster Lauren Walker, for one, pointed out via Twitter yesterday, “People obviously have very strong feelings about this because it’s Bourdain but [I] think this is fine in the abstract.”
“[Documentaries] don’t and shouldn’t have to adhere to strict reality to tell the narrative the creator wants to express. cf. [Werner Herog], truth versus ecstatic truth,” she added.
doubly okay with this given it's just putting voice to words he wrote. get over yourself black mirror sometimes technology gives us cool new tools to push the boundaries of form
— Lauren L Walker (@ActNormalOrElse) July 15, 2021
Walker also claimed that it is not unethical as Bourdain’s estate had allowed the team behind “Roadrunner” a deepfake of his voice. Moreover, she argued that “sometimes documentarians should be a little unethical.”
"it feels unethical"
1. it's not, his estate signed off on it
2. sometimes documentarians should be a little unethical— Lauren L Walker (@ActNormalOrElse) July 15, 2021
A spokesperson for the documentary, meanwhile, told Entertainment Weekly also on July 15: “”There were a few sentences that Tony wrote that he never spoke aloud. With the blessing of his estate and literary agent we used A.I. technology.”
“It was a modern storytelling technique that I used in a few places where I thought it was important to make Tony’s words come alive,” the spokesperson was quoted as saying.
Neville also revealed in an interview with GQ, published on July 13, that he also had gotten permission from Bourdain’s widow.
“I checked, you know, with his widow and his literary executor, just to make sure people were cool with that. And they were like, Tony would have been cool with that. I wasn’t putting words into his mouth. I was just trying to make them come alive,” Neville was quoted as saying.
On how they pulled it off, Neville explained that they “fed more than ten hours of Tony’s voice into an A.I. model. The bigger the quantity, the better the result.”
“We worked with four companies before settling on the best. We also had to figure out the best tone of Tony’s voice: His speaking voice versus his ‘narrator’ voice, which itself changed dramatically over the years,” he told the magazine.
“Roadrunner” comes around three years after Bourdain committed suicide at the age of 61 in France, where he was shooting for his show, “Parts Unknown.” Ian Biong /ra
RELATED STORIES:
5 facts you probably didn’t know about Anthony Bourdain
A deepfake bot generated ‘nude’ pictures of over 100,000 women
