by Mat Dirjish

Global Copyright Exchange (GCX) by Rightsify, a self-described ethical, risk-free music dataset provider, is unleashing datasets it claims will enable large-scale music models (LMMs) that are allegedly emerging as the “gold standard” in artificial intelligence (AI) music creation. Reportedly, these large-scale music models answer every user prompt and can generate every musical style and instrument.

According to GCX, large-scale models need meticulous metadata that reflects deep knowledge of music and its structures and patterns. To meet this requirement, GCX provides more than 32 billion parameters and tracks labeled by trained musicologists. Its dataset offers 4.4-million hours of music, perhaps exceeding five lifetimes of total listening.

GCX’s datasets and large-scale music models target a diversity of use cases. These include do-it-yourself music production, music generation for ads, background music for videos, adaptive music for gaming, and music APIs, all of which promise to eliminate the analog human composer/musician from the revenue stream. But so what? They can do what most out-of-work musicians/composers have done in the past, i.e., become computer programmers.

Alex Bestall, founder of Rightsify and GCX, muses, “The era of large-scale music models demands an entirely new approach to datasets, especially in terms of copyright clarity and flexibility for refining groundbreaking models and apps. Legal and high quality datasets should not be a stumbling block on the road to better generative music models. We’re making sure everyone can find a firm foundation for innovative and exciting AI music use cases.” And that includes the 12-year-old kid who doesn’t play an instrument, never took a music lesson, and can’t sing a note, let alone a note that’s in tune.

Ready to ride on the coattails of previous and somewhat existing musical talent but, for whatever reason, you need deeper details and data? If so, visit the Rightsify website.

| info@matdirjish.com | 1-516-422-1431 |