This series explores how brain-computer interfaces (BCIs) will change our daily life by the year 2022.
In the last decade, the way we consume music has been completely reimagined by streaming platforms like Spotify.
The biggest shift is the democratization of music - we don't have to pay $1 to iTunes for every song we want to listen to. Beyond that, Spotify has built an incredible recommendation system and created a library of playlists for every mood, activity, or genre you can think of.
How will music streaming industry continue to evolve by 2022?
Musical taste decoded
What if Spotify had direct access to your musical taste? It would instantly know whether or not you like a song as you listen to it.
Recent advances in brain-computer interfaces (BCIs) have made this a real possibility. EEG technology measures your brainwaves non-invasively, from outside the brain (link for more details).
Companies are starting to produce consumer EEG headsets that are becoming more and more subtle. For example, Emotiv has released a sleek EEG headset (left) and even discreet EEG earbuds (right).
Emotiv headsets are actively using in research initiatives like MyBrainTunes, which aims to build computational models that will help us explore and decode people's music-listening brain states.
The algorithm will know whether you like a particular song just by listening to your brainwaves. Here's a look under the hood:
In the image above, you can see the differences in how my brain responds to songs that I don’t like (left) and songs that I like (right). What changes isn’t necessarily the activity in specific parts of the brain, but rather the connectivity between brain regions (the degree to which they are in sync).
Think about what this would mean - if you were to be wearing an EEG headset while listening to music on Spotify, then the recommendation algorithm could tune itself without you ever pressing a button. No more likes or skips. Your brainwaves would tell Spotify everything it needs to know!
Let's explore some of the possibilities this technology opens up. In each scenario, I'll set the scene and compare the reality of today (2020) with the near future (2022).
Imagine you're dancing at a lively bar with your friends. You can't remember the last time the music's been this good! A song comes on that you've never heard of, and you really want to know the name so that you can add it to your playlist.
You ask your friends what the song is called, but you're met with shrugs. You don't want to pull out your phone and Shazam it while you're dancing - it would ruin the moment.
When you get home that night, you try to remember the song's lyrics but it's not coming to you. You spend what feels like an hour trying to Google it. When that doesn't work, you frantically try singling the melody into Shazam. At this point, it's 3am so you admit defeat and go to sleep with the melody stuck in your head.
You dance the night away without any worries, knowing your BCI is taking care of it. When you get home, there's a playlists with all the songs you liked earlier in the night. You quickly find that song you've been thinking about and save it to your dancing playlist. Just like that, easy.
Think through a typical day. You probably having playlists for different parts of your life. Personally, I like hip-hop for working out, electronic when I'm working, and light jazz when I'm with my family.
In other words, your preferences change depending on the context.
Over the years, you find a number of playlists for different activities. You choose a different playlist for each occasion, but sometimes it doesn't fit the moment.
Your preferences are sorted by factors like time of day, location, activity, and weather. Spotify generates playlists around different parts of your life, based on what you liked previously in similar moments.
Imagine you've been creating different playlists for a couple of years now. As per usual, your taste changes over time and you start to dislike certain songs. Some of your older playlists are littered with songs you don't listen to anymore.
You spend a couple of hours manually removing songs from your playlists. Nobody likes doing this.
Songs that you don't like anymore are automatically phased out of your playlists, right after hearing them. All your playlists are fresh with no work required.
Mind reading DJ
Imagine you're having people over at your place for a small party. As the host, you have to put together a playlist for the night. You hate being in charge of the music because people have different tastes - you can't please everybody.
You spend hours putting together the playlist. Your friends get to your place and seem to be enjoying themselves, but you're not sure whether they are just being polite.
You tap into your friend's Spotify profiles and generate a playlist based on everyone's musical tastes. When your friend get to your place, you get real-time feedback on whether people like the music you're playing.
Who's leading the charge
While most of the progress needs to be made in the research domain, there are a couple of startups that are laying the foundation for 2022.
Endel is an algorithmically-generated sound environment that adapts in real-time based on weather conditions, time of day, heart rate, and motion.
Weav Music is creating tools that take music from a static recording to an adaptive format that can be experienced at any tempo.