Continuation of the analysis. In part one we looked at how platforms split into "resistance" and "let it be." What's happening with AI music from the listener's perspective? Let's look at the numbers:
1. AI is nearly indistinguishable by ear. 97% of listeners couldn't tell AI from human in a blind test by Deezer x Ipsos.
2. According to Morgan Stanley surveys, 50-60% of US listeners aged 18-44 already consume AI music about 2.5-3 hours per week (lo-fi for studying or background music in cafes).
3. Catalog growth outpaces moderation. Per Luminate 2025 Year-End, streaming platforms hold 253M tracks with 106,000 new ones daily. Human moderators physically can't listen to all of it.
4. A Goldmedia and GEMA study projects GenAI will take 27% of live authors' revenue by 2028. The AI content market in Europe alone will grow from 3 to 64 billion euros.
5. This will lead to classic elitization or segmentation. Listeners will split sharply. Some will be fans of "live music" subscribing to musicians as artists (Patreon, Bandcamp, vinyl). Others will endlessly consume mood-based disposable music generated on the fly. Streaming platforms will cater to the latter.
The problem is that music for many people is a form of therapy and mood regulation. AI tracks will make people even more hooked on music.
And separately: many musicians don't understand that survival requires having your own market, your own fanbase. There will be so much AI music that without building recognizability, it'll be extremely hard. The same applies to any other content. Now is the last chance to find your audience and your voice to stand out among AI slop.