Canadian electronic artist Grimes has launched an open-source AI voice software called Elf.Tech, which allows artists to replicate her voice in their songs and receive 50% of the royalties. Grimes made the announcement after a song that used AI deepfakes of Drake and the Weeknd’s voices went viral but was taken down. She offered to split the revenue with anyone who wanted to use her voice in AI-generated songs. Elf.Tech is the prototype of CreateSafe’s Triniti platform, developed by music tech studio CreateSafe, co-founded by Grimes’ manager Daouda Leonard. The platform aims to empower creators to make music using AI tools without replacing humans with computers.
Triniti is an artistic intelligence platform that enables artists to create an AI voice clone, generate text-to-audio samples, ask music industry-related questions to a chatbot, monetize creations, and manage music intellectual property (IP). It leverages a customized version of a Realistic Voice Cloning (RVC) model and OpenAI’s ChatGPT for its virtual companion. Triniti also incorporates Stable Diffusion to generate images for album covers and a music model based on MusicGen for audio techniques. CreateSafe has raised $4.6 million in funding, led by Polychain Capital, with participation from notable figures such as Kendrick Lamar’s manager Anthony Saleh and Paris Hilton-founded 11:11 Media.
The most remarkable feature of Triniti is its voice transformation and cloning tool. Singers can record their voices and train the AI with different voice patterns and styles. Artists receive a digital voice clone and a voice likeness certificate, which allows them to set licensing terms for their voice. This aspect is particularly important as debate around voice AI in the music industry often revolves around copyright protections. CreateSafe aims to protect artists’ voices by giving them control over how their voice clone is used and enabling them to issue takedowns if necessary.
Triniti also offers an AI sample generator that creates tracks based on inputted genres or vibes, a virtual companion that answers music-related questions, and a management tool that automates workflow using LLMs. Grimes won’t be the only artist to use Triniti’s voice cloning technology, as CreateSafe has a cohort of 30 artists and record labels planning to release digital voice clones in 2024.
While some musicians feel threatened by AI, Grimes believes in the collaborative potential of AI music tools. She wants to empower artists to own and collaborate with AI to enhance their music creation experience.
CreateSafe plans to introduce editing tools, a MIDI processing visual, and launch iOS and Android apps in the future. Currently, Grimes’ AI voice tech, audio samples, and chat are available for free, with applications required for voice cloning, licensing, distribution, and management. The company intends to introduce a subscription model ranging from $99 to $150 per year.
The Triniti platform aims to address the issue of fraudulent streaming activity, estimated to have reached at least 10% in 2022. CreateSafe wants to contribute to a more ethical use of AI in music creation by ensuring consent and ownership over digital clones.
In conclusion, Grimes’ launch of Elf.Tech under CreateSafe’s Triniti platform enables artists to replicate her voice in AI-generated songs. The platform offers various AI tools for music creation while empowering artists to own and collaborate with AI. CreateSafe has secured significant funding and plans to release digital voice clones from a cohort of artists and record labels. The Triniti platform seeks to address copyright concerns and fraud in the music industry while paving the way for new and innovative music experiences.