Grimes launches AI that regenerates her voice

Grimes, the musician and artist, has developed Elf.Tech, an AI-powered software that can generate her voice. The app is in beta mode and allows users to upload audio or record directly, receiving the output file with Grimes’ voice instead. Grimes can be listed as the main, secondary or featured artist in any music made using the software, with a 50% royalty split on the master recording. Distribution is available through Elf.Tech’s platform, or artists may choose to self-distribute or release through a label.

Grimes has made her voice open-source and available for download by providing a folder of stems that include her vocals from tracks such as 4ÆM, Oblivion, and IDORU. Developers can use the stems to train their own AI models of Grimes. The software’s Chaos Manual states that integrating AI into “God-tech” is a necessary step towards a society that can pass the great filter.

The musician also announced that she will be releasing two new songs, Music For Machines and I Wanna Be Software, in the coming weeks. Grimes said that she will split 50% of royalties on any successful AI-generated song that uses her voice and encouraged artists to use her voice without penalty, stating that she has no label or legal bindings.

See Grimes’ tweets below.