AI Slop FM radio experiment puts machines on trial, one dodgy tune at a time

Listeners who enjoy music can now discover whether they prefer AI or human-created tracks without bias.

A Welsh music enthusiast named Mike Chaffin has launched AI Slop FM, an internet radio station designed to test listener perceptions of AI-generated music. The station broadcasts AI-created tracks alongside human-made songs without revealing the origin of any particular piece. Listeners are encouraged to rate the songs on a scale from S to E, tag their emotional response, and provide written feedback. Chaffin, operating the station solo from Ammanford, Carmarthenshire, aims to determine if listeners genuinely dislike AI music or if their aversion is based on prejudice. The station, currently in beta, plans to compile listener feedback into a weekly chart called the Slop 40. Chaffin has noted apprehension from some human musicians when approached about submitting their work, suggesting a deep-seated fear of AI in the industry. The experiment strips away artist names, cover art, and any AI labels, focusing solely on the music itself for evaluation.

The experiment at AI Slop FM highlights a critical juncture for the music industry, where the lines between human creativity and machine generation are blurring. By anonymizing song origins, Chaffin's project directly challenges industry biases and the market's reliance on artist identity over pure sonic merit. This approach could pressure platforms and labels to re-evaluate how music is discovered and promoted, potentially opening doors for overlooked human artists and AI-generated content alike. The results may inform future strategies for music production and distribution, impacting how artists are compensated and how audiences engage with new music.

Chaffin's initiative provides a unique, albeit informal, testing ground for the viability and reception of AI music in a real-world listening environment. The data collected from listener ratings and feedback could offer invaluable insights into audience preferences, potentially influencing AI music development and investment. If AI-generated tracks consistently outperform human ones, it could signal a significant shift in the creative economy, forcing a reconsideration of value and originality. Conversely, a strong preference for human artists would offer reassurance to the music industry about the enduring importance of human artistry.