To preserve the diversity of Latin American music, Nicolas Madoery is training AI audio tools with traditional voices from his home country. He fears many regional music styles could otherwise be lost.
Category
🗞
NewsTranscript
00:00 Nicolas Madueri and the FutureX team are researching how Latin American music is influenced by algorithms.
00:08 It's difficult for an AI model to understand the complexity and the specific characteristics of a region like Latin America.
00:16 That's the challenge. How can we create tools that represent our region?
00:25 A survey in 16 Latin American countries has shown that so far, creatives haven't really tapped into the potential of AI.
00:34 Here, we use AI mostly for project management, not creative work.
00:45 Musicians from all over Latin America work for the organization FutureX.
00:50 To test AI in music production, they composed a song together using various AI tools.
00:56 It took us about six to eight months. We tried different things and experimented with different tools.
01:15 At the start, you could hardly tell that we were using AI. The results were mostly instrumental.
01:26 Later, we trained an AI model using voices from famous Latin American singers.
01:37 The song has quite an experimental sound.
01:50 But it does give the sound researchers some valuable insights.
02:00 This song was supposed to tell us something about voices, about women and many voices becoming one.
02:09 It wasn't so much about making a nice song, but more to find out what happens when we bring all these voices together.
02:16 It's kind of like creating a mega voice.
02:24 Together with other musicians and scientists, he's training AI audio tools with the traditional sounds of his homeland.
02:31 He fears that many regional musical styles could otherwise be lost.
02:40 I think it's important to understand how AIs are trained.
02:46 Using these models shouldn't create a rift between us and our cultures.
02:54 In fact, the technology should deepen our cultural identity and enable more access to local traditional music.
03:08 This is what he has.