Music Ally has for some time been covering the emerging wave of startups exploring AI music: algorithms trained on catalogues of existing music, with the aim of then being able to create their own new melodies – and even entire tracks.
Thus far, the most prominent commercial model for all this has been that espoused by startups like Jukedeck and Amper Music: AI that spits out listenable backing tracks for video creators.
Something we’ve suggested a few times, though, is the potential for AI to play a role in the booming area of ‘mood music’ – epitomised by the playlists on streaming services designed to help people work out, relax, focus on work, sleep etc.
Those tracks are created by humans (even in 2017’s controversy about ‘fake artists’ on Spotify’s mood playlists, the creators were actual humans) but we’ve regularly wondered whether the real feather-ruffling would come when algorithms started getting involved.
Well, here they are. Startup Endel already has a mobile app available that generates “personalised sounds to help you focus and relax”. As we reported last year, it’s also been part of Amazon’s Alexa Fund program, which yesterday saw Endel launch as a skill for that company’s smart speakers.
But here’s the most interesting thing: Endel is also claiming (rightly, we think) to be “the first algorithm to ever sign a major music deal”. That deal is with Warner Music Group, with Endel planning to release 20 ‘mood and productivity-boosting’ albums through WMG’s Art Music division this year.
Five of them have already been released to streaming services: ‘Clear Night’, ‘Rainy Night’, ‘Cloudy Afternoon’, ‘Cloudy Night’ and ‘Foggy Morning’. They’re not popular yet: Endel’s artist profile on Spotify reveals it only has 109 monthly listeners on that service, although amusingly Spotify has already created one of its own-branded ‘This Is’ playlists for Endel, just like it has done for all manner of human artists.
There’s unsurprisingly no public details on the terms of the deal with WMG: we’re just hoping Endel’s algorithm doesn’t ever go sentient and end up complaining in the media about per-stream rates…
Enough flippancy though. The potential for technology like Endel’s to play a role in the rapidly-growing economy around mood-music is fascinating, especially if (well, when) we get to the point where this tech is generating music *for* each individual listener, on the fly, based on what it knows about them and their current context.
Equally, the question of how these AIs intersect with the established music industry – Endel may be the first with a label deal, but AIVA was the first to sign to a collecting society as an author – and musicians is an important topic for thoughtful, unsensationalist debate. Talking of which, check out our recent primer on the topic to equip yourself with some useful background.