One of probably the most attention-grabbing elements of “Roadrunner,” the Morgan Neville-directed documentary about Anthony Bourdain, was Neville's use of generative AI to recreate Bourdain's voice.
When I look back now, Neville told Wired that he saw this as a “fun” approach to “keep (Bourdain's) voice alive within the film”. But his approach met with strong criticism – While the synthetic Bourdain only read words that the true Bourdain actually wrote, Neville said many viewers assumed, “Oh, they simply made up (expletives).”
“Quite a lot of people told me that there have been other documentary projects doing the identical thing and everybody responded; They either modified their approach or put huge disclaimers on all the pieces,” he said.
Since then, the director has “assiduously” avoided using AI. Also in his latest documentary “Piece by Piece”, by which he dramatizes the lifetime of the musician Pharrell with Legos (Yes, really), Neville made sure to avoid him.
“Carl Sagan says 'Pharrell' in 'Piece by Piece' and I made it clear to everyone that along with his widow's permission we might get him to say 'Pharrell' without using AI,” Neville said. “We actually experimented with making the word out of syllables (he actually said).”
