Published on November 16, 2023, 4:54 pm
After creating a buzz with an AI-generated song featuring the voices of Drake and The Weeknd, Grimes surprised everyone by announcing that she would share 50% of the revenue with anyone who wanted to use her voice in their AI-generated songs. In a bold move, she went on to launch Elf.Tech, an open-source AI voice software that allows artists to replicate her voice and earn half of the royalties.
While some musicians may feel threatened by AI technology, Grimes, a Canadian electronic artist and partner of Elon Musk, is doubling down on her support for AI music tools. CreateSafe, the music tech studio co-founded by Grimes’ manager Daouda Leonard, has officially launched Triniti – a generative AI-powered platform available as a free open beta. Triniti empowers artists to create an AI voice clone, generate text-to-audio samples, seek advice from a chatbot about the music industry, monetize their creations, and manage their music intellectual property.
Elf.Tech serves as the prototype for CreateSafe’s Triniti platform. With over 1,000 songs already created using Grimes’ voice replicator, Grimes herself has joined CreateSafe’s advisory board. Additionally, CreateSafe announced that it had raised $4.6 million in funding for further development of the Triniti platform. Polychain Capital led the seed round investment, with participation from Anthony Saleh (Kendrick Lamar’s manager), 11:11 Media (founded by Paris Hilton), Unified Music Group, Crush Ventures, MoonPay, Chaac Ventures, and Dan Weisman (vice president of Bernstein Private Wealth Management).
Triniti is positioned as an “artistic intelligence” platform since its primary objective is to empower creators in making music. Daouda Leonard described it as not being fully reliant on AI but rather providing creators with interfaces to collaborate and make music on their own terms. Triniti combines a customized version of the Realistic Voice Cloning model (RVC) with OpenAI’s ChatGPT for the virtual companion feature, as well as Stable Diffusion to generate album cover images. CreateSafe is also building a music model based on MusicGen, which focuses on audio techniques such as traditional DSP (Digital Sound Processing).
CreateSafe emphasizes that its goal is not to replace humans with computers but to provide tools that collaborate with human creators and offer new experiences in music production. Leonard expressed their commitment to avoiding fraudulent practices in music streaming, which accounted for at least 10% of activity in 2022. By implementing licensing frameworks and providing distribution pathways on the Triniti platform, artists can protect their voice clones and establish ownership over them. They are also empowered to issue takedowns if other platforms misuse their voice likeness.
The standout feature of Triniti is its voice transformation and cloning tool, allowing singers to record their voices and train the AI using various patterns and styles. Artists receive a digital voice clone along with a “voice likeness” certificate, giving them control over how their voice is licensed. This provides a solution to copyright concerns surrounding AI-generated voices in the music industry.
Grimes may be the first artist to utilize Triniti’s voice cloning technology, but she won’t be alone for long. The platform has plans to release digital voice clones from 30 other well-known artists in 2024. Through these collaborations between artists and record labels who believe in this technology, CreateSafe aims to showcase ethical practices in model training and obtaining permission from IPO holders.
Triniti also offers an AI sample generator where creators can input desired genres or vibes for audio samples. For example, they can request a “catchy yet haunting love song.” The generator produces four tracks matched with album images. In addition, Triniti features a virtual companion capable of answering questions related to the music industry, such as how labels make money or how to achieve fame. The platform’s management tool employs LLMs (Language Models) to automate workflows, assisting managers in cataloging deals, creating invoices, adding new clients, and organizing projects.
Grimes’ AI voice tech, audio samples, and chat services are currently available for free to everyone. However, voice cloning, licensing, distribution, and management apps require approval upon request. CreateSafe plans to introduce a subscription model in the future with prices ranging from $99 to $150 per year. Triniti’s roadmap includes the addition of editing tools, MIDI processing visualizations, and the launch of iOS and Android apps.
Grimes expressed her pride in CreateSafe’s quick execution and vision for Triniti. She believes that art generates significant revenue while artists receive only a small portion of it. With tech advancements like AI and acceleration in mind, Grimes sees removing friction from the process of providing resources to artists as the first step towards abundance.
From DrakeGPT to Infinite Grimes: AI-generated music is making waves in the industry.