Fans notice Ye using Audimee in the making of “Bully”
On The Download podcast with Justin Laboy, Ye (formerly Kanye West) pulled back the curtain on his creative process for Bully, revealing that AI played a major role in the album's production. He likened the backlash against AI to the initial hate Auto-Tune received, suggesting that, just like T-Pain's signature sound, AI might eventually find its place in the culture.
“People are like, 'Stay away from AI.' It's a more negative reaction than Auto-Tune,” Ye said. “I remember I did Auto-Tune 'cause people thought, 'Man, this Auto-Tune is trash.' [AI] is in the same family, except that people have a more visceral reaction.”
When asked whether AI represents the future of music or simply a tool, Ye responded:
“It's time for me to explain to people the power of AI in music. […] Right now, you can take any song and separate it—just get the vocals, just get the bassline, the drums—and completely isolate it. So when I send a song or a sample to my engineers, I just say, 'JS, AI.' My engineer John Scott. It's like we're speaking AI, like THX 1138, the movie that George Lucas did before Star Wars.”
At one point in the interview, Ye's engineer pulled up his screen, prompting some curious viewers to zoom in and identify the AI software they were using—Audimee.
In the behind-the-scenes clip, Ye and his producer break down how they used an AI clone to morph an existing vocal into Ye's own, sharing a before-and-after of the transformation. To top it off, Ye played a snippet of an unreleased track—presumably intended for Bully—featuring the AI-powered vocal in the mix.
These days, recreating samples in a new voice is as simple as training a personalized AI voice model with 10-30 minutes of recordings. The more data—and the cleaner the vocals—the sharper the results, letting any voice take on a new identity.
For those looking to tap into Ye's workflow, Audimee's My Voices section lets users train their own voice models, using the same Custom Voice tech shown in the interview.
