How AI is Impacting the Music Industry
(Total Runtime: 14 minutes, 38 seconds)
Transcript
INTRO VO: The following content may contain explicit language.
HOST INTRO: Have you heard this new song by Drake and The Weeknd?
*Audio clip of “Heart on my Sleeve” produced by @ghostwriter997 on TikTok*
HOST: What if we told you that both Canadian artists had no input in the music at all. Well, that’s because it isn’t them. This song was made with the help of artificial intelligence, also known as AI.
It exemplifies the new, elevated technology taking over the music industry. It was produced by a user on TikTok with the handle @ghostwriter997.
The song was taken down from various streaming platforms due to a copyright violation. However, it sparked a conversation about the impact of AI in the music industry and its legality.
*Transition music*
HOST: With the growth of social media and more intricate technology, AI composed songs that are being spread to masses.
Numerous viral TikToks and Instagram reels have been posted and circulated showing audiences their favorite artist’s singing songs from other artists, or singing brand new generated songs.
Popular AI softwares such as Amper MusicAPP, Amadeus Code, Open AI and more, are able to generate new musical compositions based on existing musical data. These systems are able to create original music that sounds similar to the music composed by human musicians. AI is also being used to automate aspects of music production like beat making, sound design and mixing to enhance existing music tracks or add new sounds to songs.
*Transition music*
HOST: Let’s hear an example. Have you heard “Munch” by Ice Spice?
*Clip of “Munch” by Ice Spice*
HOST: Here’s an AI cover of the song using Michael Jackson’s voice.
*Clip of “Munch” by Michael Jackson made by @officialai_covers on TikTok*
HOST: We spoke to Chris, the producer of this song and the owner of @officialai_covers on TikTok. Chris is a 16-year-old from Scotland whose account has over 85 thousand followers and over 4 million likes. His most popular video has over 7 million views.”
CHRIS (cover artist): “So basically I made my first AI cover just as a joke for my friend. I think it was like Michael Jackson rapping or something and my friend actually suggested that I post it on TikTok and then I did and I got around 20,000 views I think. People seemed to really enjoy it, so I just kept making and posting more and the account just kinda blew up from there.”
HOST: We then asked Chris about how he learned to make AI covers.
CHRIS: “I learned how to do it from a Discord community called ‘AI Help’ and they basically talked me through it. I made a bunch of friends.”
“So basically my process is that: I get a song that I want to do and I separate the vocals and the instrumental, using a software called ‘Ultimate Vocal Remover.’ Then once I’ve done that I’ll take the vocals and put them into an audio editing software called ‘Audacity.’”
“And then what I do is I go over to a website called musicfiy.lol and then I transfer these clean vocals from one voice to another.”
“On average it’ll probably take me 45 minutes to an hour and that’s factoring in the time it takes to download stuff and process the vocals and do all that.”
HOST: Chris went on to explain his creative process.
CHRIS: “I tend to just pick what I think will get the most views, what I think people will find the most entertaining, like Michael Jackson rapping or Spongebob doing some heavy rap song or something like that. People tend to find it more entertaining. Therefore, they comment, driving up engagement, pushing it to more people, getting more views and just increasing the entertainment level for everyone.”
HOST: What you heard in Chris’s covers is a software generated system using a technique called voice cloning. Voice cloning is digital replication of a person’s voice. It’s used to generate songs or other audio recordings that feature the voice of a famous artist who may no longer be alive, or who may not have recorded a particular piece of music. The machine learns algorithms of the target artist's voice to then study their unique speech patterns, accent, and other vocal characteristics.
HOST: Although, this content is entertaining and often lighthearted, making it can pose legal threats to cover artists like Chris. Just prior to this interview, Tiktok flagged Chris’ posts for copyright strikes, forcing him to take down some of his most popular videos.
HOST: There are also some obvious threats and problems that AI generated music poses for the music industry. AI music challenges traditional notions of authorship and creativity. When music is generated by an algorithm, it's difficult to say who the "composer" of the music really is. Some people see this as a threat to the integrity of music, while others see it as an opportunity for new forms of artistic expression.
We asked the public for their opinions. Ethan, a Hunter College student, said:
ETHAN (student): “I feel that it's kind of funny but it can be really manipulative in a negative way. Especially of people are able to monetize that, that can be really fucked up.”
HOST: Another student told us:
CHRIS (student): “I think it just gives a medium to those people who don’t have the voice of a rapper to be able to create musically and show their talents. I would definitely like to keep listening to some new things but at the same time it doesn't have to be monetized the same way that streaming services are.”
HOST: Universal Music Group, the music corporation that has artist’s signed to their labels like Taylor Swift, Drake, Bad Bunny, Billie Eilish and more, called AI music a “fraud” and wants it to be banned from all streaming services.
AI music threatens UMG’s deep library of music and artists that generate billions of dollars in revenue.
HOST: UMG sent urgent letters in April to streaming platforms, including Spotify and Apple Music, asking them to block artificial intelligence platforms from training on the melodies and lyrics of their copyrighted songs.
One of the main concerns from major music distributors is the potential effects that the technology will have on the money they are able to make. Chris shares his thoughts on whether or not AI producers should profit from their songs.
CHRIS: “Yeah I think 100% the money from AI music should be split with the original artist. That’s why I didn’t make any money from my TikToks or anything. I didn’t think it would be right that you should be able to make money from something that really isn't yours. It’s not your voice.”
HOST: Despite the growing concerns, Chris still believes that AI could have a positive impact on the music industry.
CHRIS: “I think if used in the right way it can definitely promote new creativity and new ways to make music.”
“For me, I think the best way that AI could be used in the future of the music industry would just be as a tool to enhance it. I don’t think it will ever surpass or take over traditional music. I think you can use it to spark creativity in terms of song writing or writing lyrics or music production. There’s a lot of ways you could use it.”
"Same thing, I don't think you could ever mimic or recreate the emotion of a real person.”
*Transition music*
HOST: We also spoke with an expert and professional, Professor Stephen Spencer, who teaches at City University of New York’s Hunter College. Spencer teaches music students about music theory and music technology. He believes that the software isn’t inherently bad and should be used in moderation.
PROFESSOR SPENCER: “I see it in both ways. I think that we can benefit from a certain controlled kind of approach to artificial intelligence. But I think that we should allow it to amplify what makes us human, rather than trying to make it the most human it could be. So I think that if the project of AI is to create the most creative person, then that's a project that I think is harmful.”
“It’s interesting the inversion that AI has presented. By inversion I mean historically we think of these tools as tools that help us and open up freedom for us to be human. Things like, I don't want to have to make my tea in the morning. I don't want to have to go and sweep up and I don't want to do all these menial tasks, I want the computer to do all that for me, and then I can be free to dream about the cosmos. But now what is AI doing, it's the one making the poetry for us. It's the one making the art. And we’re stuck doing other things. Right? It's like the technology is taking over the things that are most “human” and leaving us with the things that robots should do.”
“So AI I think is not stepping on the toes of a composer whose goal is to express themselves and they’re not trying to serve some sort of function or utility.”
HOST: AI machines can also generate the same product for instruments. A network could be trained on a dataset of classical piano music that could then be used to generate new pieces of music that sound like they were composed by a classical composer.
PROFESSOR SPENCER: “I do a little bit of coding music and I wanted it to just write a piece of code for me that would generate random tones that had a different color or “timbre” its called in music. So I had it do that but I gave it specific instructions and it wrote a little piece of code that made a bunch of really weird sounding tones that I then turned into a piece. So it was sort of a starting point that was kind of the springboard for my own creativity. It kind of came up with a strange way to do it and I was like, "Oh that's interesting, let's pursue that.” So it was useful that it was AI and not just some other computational method.”
*Transition music*
HOST: But AI music is not only being used to create new songs. AI systems have been making their appearance in our streaming services for quite some time now. AI can analyze users' listening habits and preferences to make personalized music recommendations. These systems can also be used to discover new music based on users' existing preferences.
CHRIS (STUDENT): “I definitely only pretty much mainly use Spotify and I know something that is in the news is that they’re making their own AI.”
*Audio clip of Spotify’s AI DJ speaking*
HOST: In February of this year, Spotify launched its own personal DJ “right in your pocket” for subscribers. The DJ serves as a curator by analyzing user activity to play songs the user already likes and recommend new releases that correlates to your taste. The DJ also plays songs user’s had on repeat previously to hit the nostalgic feels. Commentary and background context on the song and artist is also provided.
*Transition music*
HOST: As for the future of AI music, many artists have spoken out about generated music and their thoughts, good and bad.
Canadian artist, Grimes, controversially invites creators to use AI generated versions of her voice to make new music saying she could even provide the raw audio files to facilitate it. In April she took to twitter to say she’s willing to split 50% of royalties on any successful AI song which is the same deal she would give to any artist she collaborated with. “Without penalty,” she said. “I have no label and no legal bindings.”
HOST: Last August, Capitol Records signed FN Meka, a “robot rapper” powered by artificial intelligence, who had over 10 million followers on TikTok and over a billion views on the platform. Only a few weeks later, the label dropped Meka after facing backlash for their insensitivity towards the black community. The robot, portrayed to be a black male, had lyrics including the n-word slur.
HOST OUTRO: It is uncertain to know the future of artificial intelligence and the impact it will have on the music industry. But as AI progresses quickly each day, we know our future will be changing rapidly with endless limitations.
*Outro music - Kanye West AI - Jungle (by Drake)*
HOST SIGN OFF: This episode is brought to you by Alyssa Colon, Nefissa Kemech, Joash Nyarko and Kira Scott. Thanks for listening!