Since its official debut in 2016, the annual Wrapped has been a marketing staple of the music firm Spotify, receiving a remarkable success peaking with 120 million users in 2023.

 

Its success, anchored upon its in-depth analysis of users’ listening habits throughout the year, showcased unique data visualizations each year, like Audio Auras, Soundtowns, and Me In 2023 which all trended during their release.

This year’s data visualizations however, pioneered a new route with the expanding application of artificial intelligence, by introducing AI DJ, AI Playlist, and Your Spotify Wrapped AI Broadcast for the annual recap.

A sudden 360 from a meticulous user data analysis to mere generative AI marred the supposed commemoration of a decade of Spotify Wrapped, touted to take a bolder step towards refining the music data stories it is known for. 

The bolder step, turns out, is an expansive approach to the future of AI—receiving mixed-reactions from its global users.

A Year in AI?

Spotify CEO Daniel Ek  announced the music firm’s “all-in” focus on the power of artificial intelligence towards the music industry last 2023. This aligns with Ek’s show of support with Meta CEO Mark Zuckerberg against EU’s Open Source AI regulations in the same year.

The AI DJ, which first debuted for premium in selected countries in February, plays music for you according to what genre you want based on your listening history. Meanwhile, the AI Playlist allows you to select, or type in, a prompt which allows generative AI to create a playlist you can listen to. 

Your Spotify AI Broadcast, which sparked controversy during Wrapped’s first release, is an AI podcast built by Google’s NotebookLM to discuss your listening habits and music trajectory throughout the year. While innovative, it had its fair share of issues, including a reported obnoxious remark on One Direction’s trend following Liam Payne’s passing. 

Moreover, Spotify’s recent Music Evolution used key terms to “hyper-personalize” music genres for fans. According to their newsroom, this enables the firm to follow the music industry’s fast paced evolution. However, users worldwide slammed the weird mashup of words, which were accused to be generated artificially—like Pink Pilates Princess Strut Pop or Sweater Weather Softcore Indie era.

While Spotify newsroom denied the allegations in an online statement explaining how Music Evolution, the overall aesthetic of Wrapped, and other features are built upon collaborative work within the music firm, the welcomed use of AI flagged concerns with the subpar output they released compared to previous years.

Human (Dis)Connection

As netizens air out their frustrations with the allegations of generative AI on the underwhelming Wrapped, some pointed out that the downscale in quality of data stories may be related to the layoffs conducted by the music firm on December 2023.

Last year, approximately 17% of Spotify’s global workforce, or around 1,500 employees were culled of. Prior to this mass layoff, the music firm has also fired over 600 employees on January, followed by another 200 by June 2023 in what Ek calls “rightsizing.” 

“Today, we still have too many people dedicated to supporting work and even doing work around the work rather than contributing to opportunities with real impact. More people need to be focused on delivering for our key stakeholders—creators and consumers. In two words, we have to become relentlessly resourceful,” said Ek based from an internal memo obtained by Business Insider’s Jyoti Mann. 

This event is only one of the recent layoff trend among tech companies. According to data from Layoffs.fyi, nearly 1,200 tech businesses cut off workers in the same quarter—totaling 225,000 employees losing job on the global scale. 

According to Ek, Spotify took advantage of lower cost capital during the pandemic to invest significantly on music, but the slow economic growth and rising capital post-COVID led to job cuts. 

Among the laid off employees is their so-called data alchemist Glenn Mcdonald, whose team analyze the music firm’s vast database to program algorithms that recommend music according to listening history and other user information. 

His titular work, the fan-favorite Every Noise at Once (or EveryNoise), is a website for searching hyperspecific music genres which Mcdonald curated himself, allowing listeners to find a genre that truly fit their criteria worldwide.

“The project is to understand the communities of listening that exist in the world, figure out what they’re called, what artists are in them and what their audiences are,” McDonald said in an interview with TechCrunch. “The goal is to use math where you can to find real things that exist in listening patterns. So I think about it as trying to help global music self-organize.”

The site hosts over 6,000 music genres which is connected to Spotify after the music firm’s acquisition of music intelligence agency The Echo Nest, where Mcdonald first worked. 

EveryNoise powers Spotify’s genre system—leading to several trending features of the music player including the “Fans also like” feature, the personalized “Daily Mix,” and “Daylist” which immediately hit off with listeners since its first release.

In addition, the music genre map provides information that generate statistics making up the reports released on Spotify Wrapped. Among the key features that use these metadata are Soundtowns, top genres,  listening personalities, and audio auras.

With Mcdonald’s departure from Spotify, the database he created is also essentially dead, as it highly depended on the music firm’s huge array of user information. While Spotify has an application programming interface (API) for developers, a TechCrunch report cited that it is not as comprehensive as Mcdonald’s internal data.

While a welcome addition to our listening experience, usage of artificial intelligence takes away the human touch that Spotify has offered through the years. Mcdonalds’ efforts to name music genres across the globe and provide proper data visualization of how users listen yearly is an important aspect of why users subscribe to Spotify—and now, it’s only a shadow of its former success.

“It didn’t connect me to communities or the world, or put my listening in relationship to anything,” Mcdonald remarked. “So, for me, it misses the important potential of a streaming service where everybody is listening together and just treats it as if each individual listener is listening by themselves.”

While artificially-generated algorithms work at a lesser cost, they threaten tech workers’ job stability and take away human creativity that users pay for. We no longer have interactive data stories interpreted by humans—we have a sloppily arranged music revolution and AI discussing your listening habits in 2024.