Its newly released AI DJ is an attempt to get people to listen to music they might not normally listen to. Spotify is spending more and more on artificial intelligence to provide users with more personalized music recommendations, and the firm already has more than 100 million tracks available. The challenge, say audio specialists, is determining when people want to play around in discovery mode and when they want to continue with their known musical preferences.
With over 600 million subscribers and 100 million tracks available, Spotify has a significant problem in helping listeners discover music they would love. A crucial part of Spotify’s aim is the promise of personalized suggestions and customization, which will give the extensive collection additional significance.
Over the years, Spotify’s set of tools for making suggestions has grown. These include Spotify Home Feed, Discover Weekly, Blend, Daylist, and Made for You Mixes. Additionally, evidence of its success has emerged in the last several years. At their 2022 Investor Day, Spotify revealed that the number of artist discoveries hit 22 billion, compared to 10 billion in 2018. The firm also indicated that they are far from done, adding to the data.
Spotify has been pouring resources into artificial intelligence (AI), particularly ML, for at least ten years. Its newly released AI DJ might be its greatest wager to far on the idea that users would be able to better tailor their listening experiences and find new music via the use of technology. With the goal of encouraging listeners to step outside of their comfort zones, the AI DJ creates an atmosphere similar to that of radio by stating the titles of songs and the lead-in to tracks. Although AI systems are great at playing back what their users appreciate, they have a hard time predicting when you might want to switch things up.
Listeners can hit the DJ button to hear something new and less-directly-derived from their existing preferences. The AI DJ combines personalized technologies, generative AI, along with a dynamic AI voice. The soft sounds of an AI DJ cover the efforts of experts in music and technologists who work to enhance Spotify’s suggestion capabilities. Hundreds of music editors and specialists from all around the world work for the organization. The generative AI technology enables human specialists to “scale their innate knowledge in ways never before possible,” according to a Spotify spokesperson.
From the millions of listens that the AI system has access to, it is possible to extract a few attributes from the data on a certain artist or song, including specific musical qualities and the typical pairing of the two. Easily obtainable details on the music include its genre, year of release, and emotional state (happy, danceable, gloomy, etc.). The speed, key, and instruments used in a piece of music are also detailed. By integrating this information from millions of listening sessions with other users’ preferences, new recommendations can be generated, and the gap between aggregated data and assumptions about individual listeners can be filled.
“People who enjoyed Y also enjoyed Z” is the simplest way to put it. Using the logic “We know you like Y, so you might like Z,” AI is able to discover compatible partners. Spotify even claims it’s effective. “Since launching DJ, we’ve found that when DJ listeners hear commentary alongside personal music recommendations, they’re more willing to try something new (or listen to a song they may have otherwise skipped),” added the spokesman. If it works, more than just the listeners will feel better. Artists looking to connect with new fans would also benefit greatly from a top-notch discovery tool.
Everyone is relying on AI algorithms to help them find a meaningful balance between familiarity and novelty, according to Julie Knibbe, founder and CEO of Music Tomorrow. The organization aims to assist artists in connecting with more listeners by understanding algorithms and how to work better with them. An important unresolved challenge for everyone involved, including Spotify users and artists, is finding a balance between remaining with known patterns and discovering new music.
Any artificial intelligence (AI) will only excel at tasks that are specifically given to it, according to Knibbe. Since the beginning more than ten years ago, these systems of recommendations have improved greatly in forecasting your tastes. What they can’t do, though, is read your mind, which is especially problematic when you’re looking to explore uncharted musical territory.
In an effort to utilize generative AI to create new suggestions that fit different moods, activities, and vibes, Spotify has introduced Daylist. This feature takes into consideration both the listener’s established interests and the different circumstances that can influence and change those tastes during the day. According to Knibbe, these kinds of advancements might keep happening, and AI could eventually figure out how much innovation a listener wants. However, she also said, “the assumption that people want to discover new music all the time is not true.”
The vast majority of listeners still gladly return to tried-and-true musical genres and styles. Knibbe explained that different types of listeners, curators, and experts place distinct demands on artificial intelligence systems. She claims that most Spotify users are more laid-back and whose listening consists of providing a “comfortable background” to everyday life and that experts are harder to surprise than the bulk of listeners.
One hundred million music are at your fingertips, yet if you listen to the same hundred songs over and over again, you can see why a new equilibrium is being pursued. A music critic and author of “Every Song Ever: Twenty Ways to Listen in an Age of Musical Plenty,” Ben Ratliff, argues that algorithms do more to prolong this problem than to solve it.
“Spotify is good at catching onto popular sensibilities and creating a soundtrack for them,” Ratliff remarked. “Its Sadgirl Starter Pack playlist, for instance, has a great name and about a million and a half likes. Unfortunately, under the banner of a gift, the SSP simplifies the oceanic complexity of young-adult depression into a small collection of dependably ‘yearny’ music acts and makes hard clichés of music and sensibility form more quickly.”
Ratliff continues to favor works of curation which highlight the preferences of real people. According to him, even a decent playlist may have been put up by someone with a developed sense of pattern recognition, “whether it’s patterns of obscurity or patterns of the broadly known,” rather than much intention or consciousness.
Within the 100 million-track universe, AI might become “either a utopian or dystopian solution,” depending on the person. According to Ratliff, the majority of people should simplify their streaming music experiences. “As long as you realize that the app will never know you in the way you want to be known, and as long as you know what you’re looking for or have some good prompts at the ready, you can find lots of great music on Spotify.”
