Spotify makes social networks blush

AI threatens to collapse the music market. Last year, Spotify announced it had removed up to 75 million AI-generated songs, which competed with the 100 million tracks created (genuinely) by humans. The Swedish company has decided to put a stop to this and now requires that AI-generated music projects be identified as such. If they don't do so voluntarily, their review systems will flag them when detected. A service like Deezer explained, in April last year, that it received up to 18,000 AI-created songs daily. It's a tide of banal music, but one that aspires to be featured in some of the playlists that have massive reach and thus generate good income. In addition to labeling automatically generated tracks, Spotify commits to not integrating them into the lists it compiles. These are measures that should be applauded, provided they don't remain mere announcements and are applied effectively.

Spotify is grappling with several concerning problems: the meager returns to independent artists in exchange for their music, or algorithms suspected of recommending songs not based on user tastes but on the company's economic interests. However, in this AI journey, Spotify exposes social networks. Facebook, X, and the rest could, if they wished, detect the bots that pervert their service and attack the normal functioning of democracy. According to the company, users have expressed a demand to know if there are real people behind a brand or AI-created masks. Perhaps it's time for Musk, Zuckerberg et altri to feel the same pressure. (PS. Speaking of noise, it's time to hit pause and silence myself for a while. To read, walk, and observe a lot to be able to write. We'll meet again in five weeks).