Is AI to Blame for the Spotify Song With 49 Different Names?

• Songs made with generative AI are making inroads into streaming services.
• One Spotify user was recommended the same song under 49 different names, and we suspect AI is behind it.
• The songs sound exactly the same, but each has a different title, artist, and art.

The latest earworm you can’t get out of your head? It may have been written by a robot.

After all, AI-generated songs are taking over the conversation in the music industry. Over the past week and a half, mainstream names like Drake, The Weeknd, Beyonce and Rihanna have been involved in one way or another.

Aside from technology’s ability to, in theory, recreate any song throughout recorded music history, another disturbing picture is playing out: Songs supposedly created by AI can be uploaded to streaming services. and recommended by listeners…but no one understands how they got there.

Spotify user Adam Phase gave to twitter try to find Faze said, “One song kept playing loudly on his Spotify radio this week. Every time I looked it was a completely different song title and artist.”

He decided to add each of these recommended tracks to his playlist. The playlist now contains 49 “different” songs. Each of these songs has a different title, artist, credits, and his cover art, but the audio is the same across the board, a 53-second recording of him that sounds like the intro to a famous TV show. , ripe with steady piano and strings. , it can be designated as classical music.

Song titles include “Fosticrians”, “Thorncutter” and “Vattio Bud”, while listed artists include Moditarians, Elval Woodridge and Trey Dawson. All of these artists have little to no presence, with just a handful of tracks that are often slight variations on the same classic number above.

Faze told Insider that he first discovered the puzzle while searching for 1950s orchestral music and began using Spotify playlists to satisfy that urge.

After listening to this classic collection, Spotify’s algorithm took over, making recommendations based on the playlist and starting playing the track in question.

Faze initially dismissed it as a strange coincidence, but after showing a friend what Spotify recommended, the two edited and compared all the songs.

Looking at the upload dates of the songs, we can see that they were all uploaded to Spotify after June 2022. The only label listed for some songs is Sky Tech Distribution, while others have no label at all.

In his Twitter thread, Faze points to a website that compiles all music related to Sky Tech Distribution. Associated with this label are his 1,944 albums, and Faze believes that, given the general nature of both, all the music and corresponding cover art for him is AI-generated. I’m here.

Sky Tech Distribution could not be reached for comment.

Spotify also didn’t respond to a request for comment.Given the nascent field of generative AI, there doesn’t appear to be an official company policy on uploading AI-generated songs to the platform.The streaming service turns AI into a DJ tool. Built-in to create personalized recommendations based on the user’s listening habits.

A Faze thread that went viral attracted several users with replies making similar findings.

“I found 10 more without even trying.” 1 Twitter user commentedsome link to their own playlists containing 24 versions of the same song, speeding it up or slowing it down.

Writing about AI and music in February, musician and cultural critic Jaime Brooks said the advent of AI-generated tracks would inherently devalue major-label music. As more songs are uploaded to music platforms (he had 100,000 songs a day as of October, according to Music Business Worldwide), the competition for user attention increases.

Universal Music Group last month asked Spotify and Apple Music to block AI services that train generative models on copyrighted music catalogs, according to the Financial Times.

One theory put forward by Faze is that Spotify is bringing AI-generated music to its own platform to increase discoverability through playlists focused on passive listening. If they were happy with the songs they could produce in bulk, the platform would rely less on licensing original material from major labels.

Of course, individuals or groups who collectively create cheap content and recognize the opportunity to make money from collective streams can also be behind AI-generated songs.

Regardless of who or what is behind the mystery of 49 songs, it remains to be seen if people will actually be tricked into listening to AI-generated music.

When asked how long it took to ask if the track was AI-generated, Faze was quick to respond. “