The Atlantic creates a searchable database of music used for training AI models
The Atlantic has made publicly accessible four large datasets containing millions of tracks used to train AI, including sets with over 12 million and 9 million songs. These datasets have been widely downloaded and confirmed as being used by companies like Google and Stability in research.
Topics: Artificial Intelligence Copyright Data Privacy Music Industry