Future Innovation: Customizing playlists based on real-time behavior of music app users

Once I had a dream that my music app worked for me...

Do you have that one song you're supposed to like because it’s your genre, vibe, energy… but you hate it instead? You always click on next when it starts.

For me, it’s The Lemonade Song by Pink Martini. I listen to jazz often, and yes - I can imagine why the algorithm always puts this song into my dynamically generated playlist. But I always skip it. I wonder if there is a chance for the algorithm to become so smart that it keeps the information about all the songs I skipped more than once, and never shows them again within my playlist??That would be nice.

And then again, there are songs that I like so much that I always volume up a bit when they start. Those should be a priority when creating a personalized playlist.?

That would be great.

Finally, I do listen to different music depending on my mood and current activity.?

Let’s imagine that a music app can detect my current activity and state of mind, and find the perfect songs for me according to that.

Is this only a dream, or is it possible?

Well, there could be a personalized invisible ranking algorithm that could detect all user activities during the listening - clicking on repeat, next, volume up. This one should be easy (if we imagine that we have endless resources).

A pair of headphones equipped with smart sensors that can detect activity should do the other part. Sensors can track users' heart rate and detect if a user is into running or riding a bike. Those sensors could be in a close relationship with users' music apps. The app could recommend some nice pampering songs.

But is it possible for a music app to detect our mood?

Apparently yes.

Researchers from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) led by MIT professor and project lead Dina Katabi have developed “EQ-Radio,” a device that can detect a person’s emotions using wireless signals.

By measuring subtle changes in breathing and heart rhythms, EQ-Radio is 87% accurate at detecting if a person is excited, happy, angry, or sad.

A team from the Korea University of Technology and Education and Nokia Bell labs presented GrooveMeter - Automatic Detection of Reactions to Music via EarableSensing. Their research is based on the natural responses of people listening to music - nodding their heads, tapping their feet, and singing along. These reactions are compelling to enable interesting music engagement-aware applications.

There is little room for completely new ideas, but there is plenty of room for combining all that great research into something extraordinary.

From the user that spent 41.376 minutes listening to music in 2023.

Our employees definitely dream big. Do you?