Find Out How Bad Your Spotify Taste In Music Is With New AI Tool
Think you've got great taste in music? How Bad is Your Spotify? uses sophisticated AI to tell you how bad it thinks your listening history is.
In case users are feeling a little too proud about their Spotify habits, there’s a new AI tool to knock them down a few pegs. In a new project from the digital publication The Pudding, the How Bad is Your Spotify? tool uses a neural net to judge what they call “your awful taste in music.” If users are feeling especially secure in their playlists, this site might just change their minds.
A small digital publication located in New York, The Pudding consists of a team of eight self-described “journalist engineers” that craft their content with a combination of research, reporting, data analysis, coding, writing, and design. The publication prides itself on avoiding clickbait-type articles in favor of visual storytelling to make ideas more accessible. Their visual and data-driven stories include topics including politics, the economy, and, yes, music.
The How Bad is Your Spotify? test is among the publication’s most recent posts. The artificial intelligence (AI) will first ask users to log in to connect with their Spotify account to gather the data. Once they’ve opted in, the neural net will begin loading the library to run its analysis. Throughout the process, it will also ask a few clarifying questions, such as if users “really” listen to certain songs, and whether or not they do so “ironically.” Once the process is complete, the AI will give its scathing review, citing which tracks users listen to “too much,” which artists they “stan” an “uncomfortable amount,” and will even rate them on how "basic" they are. Unlike Spotify Wrapped, which celebrates trends and users’ music tastes, How Bad is Your Spotify? takes every opportunity to mock and cut down participants' choices. And while it might lack the flashy colors and pleasing design of Wrapped, the Spotify judgment tool makes its point.
After having their Spotify history ripped to shreds, participants might wonder just what makes the AI qualified to tell them their music taste is no good. Using a neural net, the tool claims it’s trained on nearly over two million indicators of “objectively good music.” To them, this means recommendations from record stores, obscure subreddits, and Pitchfork reviews. The trained neural network uses this background in combination with user interaction to let participants know just how it feels about the songs, artists, and genres they choose.
For those users who loved the data reporting of Spotify Wrapped, the How Bad is Your Spotify? AI may be a fun follow up. While it doesn’t have built-in sharing options like Wrapped, its simple design and biting judgments are definitely screenshot worthy. That is if users have thick enough skin.
Source: The Pudding