X Does Not Mark the Spot-ify



Music streaming service, Spotify has recently released an “AI DJ” named ‘X’ which curates sets of songs for you to listen to with an AI-generated dialogue between. While it is a great concept, it hasn’t quite been executed in the best way, along with only being on mobile devices.

One of the main issues people have is how forced a lot of the song choices are. Picking from top Spotify playlists you might not even be interested in, and even if you press the “I’m not interested” button, it just throws more at you. When the feature works as intended it is cool, providing an accurate introduction to songs and their backstories while appealing to what you as an individual want to listen to. As is, however, it barely works.

A major issue I as a user have personally experienced is what “songs” it chooses to play. I used to have trouble sleeping and would play white noise on Spotify to help, this was a couple of years ago. Now, sometimes when the DJ introduces songs, all it plays is white noise, saying “I know you love this artist!” and bam, white noise. It is an extremely funny thing to encounter but there should be a filter for this in the Spotify system for it.

This DJ also seems to have an extreme bias for what it describes as “Indie Music”, which is essentially every kind of music it can suggest for some reason. This might have to do with Spotify not tagging songs correctly or could be the fault of the AI itself. The monologues themselves are also AI generated which could explain its “Indie” fascination if the bot wasn’t trained well and chooses its songs based on that. Spotify isn’t exactly forthcoming with information though so we don’t know for sure.

While the feature is good fun, it fails for finding new music. The AI either only plays random songs that are popular in general or songs you have already listened to repeatedly. Multiple people I know have had similar issues as well.

While this is a great idea with a lot of potential, it has a lot of work that needs to be done to live up to its premise. It is still a feature in beta so we can only hope it will improve over time.