Between 2000 and 2010, the music industry’s total grosses dropped from $14 billion to $6 billion. The change from selling albums, to selling single songs, to hoping fans stream songs on repeat has made it tougher to make a living (for mid-level acts, that is; the rich stars are richer than ever, mostly on the back of tours). Anyway, Silicon Valley thinks it can help by introducing artificial intelligence into the mix. But can AI save the music industry?
Forbes has a good overview of several startups hoping to make music profitable again. Most of them revolve around either helping musicians find fans, or figuring out the massive, messy world of music streaming. There are some interesting ideas, no doubt, like an AI that automatically mixes beats from streams, allowing playlists to be used in more places, while others will help musicians and labels figure out who owns the rights to a song being streamed and thus who gets paid.
While detailed accounting of which songs are played when (and where) is really to the benefit of both musicians and fans, the real problem still seems to be how we value music. Spotify alone streams almost a billion songs a day, but many bands see about $.004 a stream. Granted, sketchy accounting has plagued bands well into the digital age. Being able to better find potential fans and get your music in front of them will certainly help, but exactly how much? Spotify is a giant in the field, and it’s lost hundreds of millions of dollars, every year. Apple likely is losing even more money, as it has half of Spotify’s subscriber base.
The reality is that the vast majority of music industry revenue goes to labels, not musicians. No amount of AI is going to fix that, and the truth is, musicians need to get paid for their work. AI might be able to help them find a few more fans, but it’s going to have to be the industry itself that solves its econ problems.