That New Hit Song on Spotify? It Was Made by A.I (www.newyorker.com)

Nick Arter, a 35-year-old hobbyist-turned-semiautomated musician, has used AI tools to produce hundreds of tracks and score streaming hits that earned him hundreds of thousands of plays. Working under the name Nick Hustles, he writes lyrics, composes text prompts (he says a good prompt contains year, genre, instrumentation, mood and emotion), and iterates outputs from generative-audio apps like Suno and Udio, then generates cover art with Midjourney. His workflow—saving style shortcuts and polishing iterations—enabled rapid output (roughly 140 songs in a year) and viral traction via algorithmic recommendations; similar AI creations have reached the charts (an AI country song hit No. 1 on Billboard’s Country Digital Song Sales chart and passed 3 million Spotify streams) while platforms like Spotify currently do not label AI-generated music consistently. The story illustrates a broader shift: AI lowers barriers to music production and opens new creative personas, distribution channels and revenue streams, but it also exposes technical and ethical limits. Studies show listeners can only barely distinguish AI from human music (~53% accuracy); platforms have removed millions of “spammy” tracks (Spotify cited 75 million) but many unmarked AI songs remain. Musicians and producers report common AI flaws—sterility, flat structure and weak hooks—prompting a niche industry of “humanizers” who fine-tune AI vocals or train voice models on real recordings. The result is an ecosystem that amplifies novelty and scale over craft, reshaping gatekeeping, copyright and quality norms in the music industry.