Spotify AI's plan to determine your mood and suggest the perfect track for it.

• The streaming audio company, Spotify, has more than 100 million tracks accessible to its listeners and is increasingly focusing on AI to personalize song recommendations.
• The AI DJ recently launched aims to help listeners explore new music and step out of their comfort zone.
• Audio experts find it challenging to determine whether users prefer to explore new music or stick with their established preferences.

Spotify's mission is to provide personalized and meaningful recommendations to its 600 million subscribers, who have access to over 100 million tracks.

Spotify's recommendation tools have expanded over time, including the Home feed, Discover Weekly, Blend, Daylist, and Made for You Mixes. Recent data shows that the platform has been successful, with artist discoveries increasing to 22 billion per month from 10 billion in 2018. The company stated at its 2022 Investor Day that this is just the beginning, and they are not done yet.

Over the past decade or more, has been investing in AI and, specifically, in machine learning. Its recently launched AI DJ may be its biggest bet yet that technology will allow subscribers to better personalize listening sessions and discover new music. The AI DJ mimics the vibe of radio by announcing the names of songs and lead-in to tracks, something aimed in part to help ease listeners into extending out of their comfort zones. However, an existing pain point for AI algorithms - which can be excellent at giving listeners what they already like - is anticipating when you want to break out of that comfort zone.

The AI DJ employs personalization technology, generative AI, and a dynamic AI voice, allowing listeners to discover new and less-directly-derived music recommendations by tapping the DJ button. Behind the soothing tones of an AI DJ are individuals, including tech and music experts, striving to enhance the recommendation capabilities of Spotify's tools. With hundreds of music editors and experts worldwide, Spotify aims to leverage the generative AI tool to amplify the expertise of its human team.

The AI algorithm can analyze data on a song or artist, including musical features and typical pairings among millions of listening sessions, to generate new recommendations and make individual listener assumptions.

Spotify has discovered that listeners who hear commentary along with personal music recommendations are more likely to try new songs or listen to a track they may have otherwise skipped.

A great discovery tool can also provide relief to artists looking to connect with new fans.

Julie Knibbe, founder & CEO of Music Tomorrow, emphasizes the importance of balancing familiarity and novelty in a meaningful way, and how AI algorithms can help achieve this balance. She notes that this issue is central to all parties involved, including Spotify, listeners, and artists.

"Knibbe stated that any AI is only as good as the tasks it is programmed to perform. Despite the existence of recommender systems for over a decade, they excel in predicting what a person will enjoy. However, they lack the ability to understand a person's thoughts and preferences, particularly when it comes to exploring new musical genres or categories."

Spotify's Daylist utilizes generative AI to consider listeners' established tastes and the varying contexts that can influence their music preferences throughout the day, providing personalized recommendations that match different moods, activities, and vibes. Knibbe believes that the AI may continue to improve and better understand the formula for how much novelty a listener desires, but she emphasizes that people's desire to discover new music is not always constant.

Many individuals continue to return, with satisfaction, to familiar musical landscapes and routines.

Knibbe stated that there are different types of listeners, curators, and experts who put varying demands on AI. Experts are harder to surprise, but they are not the majority of listeners, who tend to be more casual. According to Knibbe, Spotify usage for these listeners often involves creating a "comfortable background" to their daily lives.

Music critic and author Ben Ratliff argues that algorithms are not the solution to the problem of listeners preferring the same songs repeatedly, but rather a further perpetuation of it.

"Ratliff stated that Spotify excels at identifying popular trends and creating playlists for them. For example, the Sadgirl Starter Pack playlist has a catchy name and over a million likes. However, under the guise of a gift, the SSP oversimplifies the nuanced complexity of young-adult depression into a limited collection of predictable music acts, resulting in clichéd music and sensibilities forming more quickly."

Ratliff prefers works of curation that are created by people with genuine preferences, even if a playlist is well-made, it may have been created without much intention or consciousness, but rather through a developed sense of pattern recognition, whether it's obscure or well-known patterns.

Spotify offers a vast music library, but it's important to remember that the app cannot truly understand users' preferences. Ratliff advises simplicity in streaming music journeys, suggesting that users should have a clear idea of what they're looking for or have some good prompts ready to find great music on the platform.