The Atlantic created a searchable database of the music used to train AI
Atlantic reporter Alex Reisner recently uncovered four datasets of music being used to train AI models and made them fully searchable for the public.
The Atlantic has launched a searchable database that exposes the vast amounts of music being used to train artificial intelligence models.
The short version
- Four datasets uncovered, including two massive sets of 12 million and 9 million tracks.
- Publicly searchable via The Atlantic’s "AI Watchdog" site.
- Contains copyrighted music from artists like Lady Gaga, Radiohead, and Wu-Tang Clan.
What happened
Reporter Alex Reisner uncovered four datasets used for AI training and made them searchable for the public. While two of the sets are relatively small at over 100,000 songs each, the other two are enormous, totaling 21 million tracks. Some of these sources, such as the Free Music Archive, require licensing for commercial applications despite being free for personal streaming.
Why it matters
The database reveals that millions of copyrighted tracks are freely available in AI training sets, even when they are not supposed to be. Google and Stability have both confirmed using these sets in research papers. This tool allows the public to see exactly which artists—ranging from Bruce Springsteen to experimental composer Hainbach—are being used to build the world's AI music models.
Summary by Nerd News Network. Read the full article at The Verge — AI via the links above and below.