The AI Revolution – Streaming Services Using AI

All streaming services already have or will in the future have more and more AI generated music in their playlists … 

… and of course they won’t tell us, the users, that this is the case. Because who wants to hear from the music service of his trust the words “So now we have something really fantastic for you, music generated by artificial intelligence, awesome!” 

The impetus for argument with this thought, gave me the publication of the Universal statement after the Drake/The Weeknd fake : Universal (and more will follow) prohibits the use of so-called APIs of the licensed music services, so that no more music can be used to train AI – fair enough, these are their rights.

Let’s think for a second: Music services, the DSPs, are 100% data-driven tech companies and they extract, analyze and aggregate everything that happens in their universe, of course including all content-based acoustic and physical properties of the music, because otherwise the necessary info to their features would be missing: personalization, recommendation, automatic playlisting, automatic radio stations … add to that the global data of consumers, across regions to the individual fan. 

With the help of these billions of daily information nodes, it is no problem at all a) to have special music produced for functional experiences as buy-outs and b) to create them automatically in the near future. And let’s not discuss “quality” and hits, the market for music that serves as a pure function, a soundscape for our activities, such as concentration while learning, relaxing, yoga, workout, showering, falling asleep, even dancing etc. is huge. The millions of plays and hundreds of thousands of monthly listeners of these “artists” of such content speak a clear language, on any music service, just search for “White Noise”, “Pink Noise”, “Rainforest”, “Focus” on Spotify & Co. …

Every single percentage point increase in margin is an extreme incentive for a music service to invest aggressively

and saving 70% compensation per play by using machines sounds like a dream to investors. To the address of the advocates of “real artists”: think of “The Monkees”, boy bands, the casting formats like Pop Idol and all the constructed models from the songwriting camps with the always same repetitive patterns in the respective taste of time for the charts. Pop was and is always a make-believe world to which AI fits like a glove. 

AI first takes over everything that can be imitated and automated well, so it is clear who and what will be replaced in the future. So let’s focus on creating something that is of value to the machine because it desperately wants to “learn” it as something previously unknown, creative and new – let’s create something worth imitating! 

As long as pop music and its protagonists continue to care more about their “brand awareness” or follower count and grind all the rough edges for possible sponsorship and no longer have any social or cultural relevance, so interchangeable and replaceable by AI are their creations and business models – with many greetings to all commercial pop radio stations with the approx. 200 titles in their monthly rotation, you are next too.

Scroll to Top