Spotify wants to suggest songs based on your speech and emotional state

The limitation of having a handful of songs at our disposal on vinyl, cassette, CD, or even the more flexible mp3 player is already just a distant memory. Digital natives who cannot remember a time before the iPhone or iPad expect to play a song within a few seconds of thinking about it. But no matter what age you are, how you consume music would have continuously evolved as you build the ultimate soundtrack to your life.

The traditional album can struggle to hold an audience's attention when the distraction of 50 million songs is just a swipe away. The labour-intensive effort of creating a mixtape that fuses different styles has also been replaced with themed music playlists to match our moods. But what hasn't changed is that music continues to bring people of different ages and cultures together. 

The opening bars of a song can stir our emotions and instantly transport us a few decades back in time. The link between music and memory is also creating new ways to communicate and engage with dementia patients. As technology and our understanding of music's power improves, we can expect the industry to evolve further.

Music suggestions based on your emotional state

patent filed by Spotify in 2018 was recently approved. It could enable the streaming platform to use speech interpretation to determine the emotional state, gender, age, social setting, or listeners' accent. The data would allow the Spotify app to make real-time recommendations based on their subscribers' mood and location.

Scheme of audio interpretation

The more you use the app, it could learn when you are happy, sad, or angry and seamlessly recommend music that would match your mood.

A recent study by Spotify also revealed links between our personality traits and musical preferences. However, most music fans will agree that their musical tastes are way too complex for a machine to understand.

In the patent, Spotify claim that listening out for background noises such as birds chirping, office noise, or traffic will help algorithms determine whether the user is at home, walking, or in the workplace. But many are understandably feeling uncomfortable at a music streaming app listening to every aspect of their life.

Welcome to Personalization 2.0

If we dare to look behind buzzwords such as personalization 2.0 and hyper-personalization, you will find a long list of always listening brands at every customer touchpoint. It is believed that the data created will help predict future behavior and intent so that brands can give customers more of what they want and when they want it in the same way that Netflix keeps their viewers watching one more episode.

In a world of digital assistants obeying our every command, even a few swipes on our smartphone can feel like a friction point. Many now find it easier to send audio messages rather than composing lengthy text messages.

For the next generation of personalization to be a success, brands must have a single unified data record for their customers and seamlessly drive real-time actionable insights they can act upon.

Many marketers believe that accurate personalization could eradicate spam and irrelevant marketing messages. But trading privacy in the name of convenience will not be perceived as a good deal by everyone. Monitoring our speech patterns, environmental noise, and even personality traits means that Spotify would always be listening, just to autogenerate a new playlist. Is this a trade you would be willing to make?

Empathetic AI

The recent Spotify announcement is just one example of the move towards a more empathetic AI. A few years ago, Amazon also patented voice analysis technology to enable Alexa to detect emotional changes. Emotional states such as happiness, joy, sadness, anger, boredom, fear, or even if you have been crying can all be detected in your voice commands.

Despite the headlines around Spotify's latest patent filing, right now, it's just an idea that has raised a few eyebrows.

Empathetic AI and technology that senses our emotional state and personality traits are still a long way off. But we also need to accept that our impatience and an insatiable thirst for speed and convenience drive this technology and progression of algorithmic recommendations.

As the pace of technological change continues to race ahead at breakneck speed, brands such as Spotify are challenged with providing the same service that the likes of Netflix and Amazon provide. To do that, they need to understand their customers across multiple touchpoints and devices. But by removing data silos to obtain greater understanding, they can quickly drift into creepy territory. 

Hyper-personalization in sterile echo chambers could unwittingly remove the sheer joy that creativity can bring to our lives. The surprise of stumbling across and falling in love with an artist's music that is outside of your music taste cannot be matched or understood by AI. When it comes to discovering new music, serendipity will always be one step ahead of any algorithm. AI cannot begin to comprehend the rules and art of creating a mixtape, and I think there is something quite cool about that.

Leave a Reply

Your email address will not be published. Required fields are markedmarked