AI's Massive Role in Deezer Music Uploads

**SPEAKER_1** (0:00)
Welcome to the podcast. Today, we're talking about Deezer that says AI-generated tracks are now more than 50% of all their daily uploads. So half of all music going up is AI. And for those that don't know Deezer, it's essentially a Spotify or Apple Music competitor. It's the biggest one in France. When I, you know, I lived in France for a couple of years, that's basically whatever it listens to music on, and a lot of different places in Europe, it's pretty popular. And the reason why I think it's important is because I think it's indicative of what's happening in a lot of different geographies, not just France, because we were probably seeing something very similar on Apple and Spotify. So we're getting into the state of all of all of this with AI and music generation. But before we do, I want to tell you about the AI Hustle School community. Every single week, Jamie and I, we go and record a bonus piece of content and put on the AI Hustle School community. And we actually record a live stream right now for the end of the month. It is free to join the AI Hustle School community. It won't be free forever. And with the free membership, you get five of our bonus videos. You get live streams every single week for free. And you get access to a community of over 570 members. We have over 100 videos we've recorded of how we grow and scale our businesses using AI tools, the numbers, the metrics, the breakdown, all the stuff we're going to share publicly. That's $19 a month if you want to get access to the library of over 100 videos. But otherwise, you can join the school community for free and you can get access to tons of incredible members and an amazing community. Okay, let's talk about what's going on with Deezer. So this is a company that said that they're going to delist AI tracks that are unstreamed for six months because they said that they're getting 90,000 uploads per day. And so they're like, look, if you don't get any streams on your song for six months, we're just going to unlist you because we have so much music. Just the volume of it is insane. And I'm going to be honest, it's using a company like Suno. It's very easy to create a song that sounds good. And it's interesting because I know people have a lot of different opinions on this. I personally make a lot of money every single month from Spotify. A chunk of my catalog I made before AI was without using AI essentially, and a chunk of my catalog I've used with AI. Some songs I've remade with AI since, you know, like a lot of stuff I made when I was in college. My biggest part of my catalogs, I would just basically make relaxing piano music. I had a friend that was good at just playing relaxing piano music. I'd get them to record for a couple hours. I'd take that, chop it into two-minute chunks and just post it as songs online to study to, or relaxing music. Most famously, a lot of people laugh that I have a lot of music labeled relaxing music for your cat to help your cat relax, and I make $600 a month, and I have since I was in college, on like an album of relaxing piano music for cats. It's kind of funny. So what's interesting about all of this, though, is I do make a lot of money from AI, but it's also stuff that's getting streamed and listened to. And I think there's a couple different arguments. Like, I would say my preference is always gonna be a real human singing. You can just kind of tell that the voices sound better, the instruments sound better, the lyrics sound better. Like everything really sounds better about a human usually. And even if it's kind of unique or interesting, like that's the sound that you like to hear.
But that being said, I have listened to some AI music. I was recently scrolling through Instagram, tap it on some Instagram stories, saw someone's story. There was a song in the background. I was like, yeah, that's pretty catchy song. Looked it up. And sure enough, once I got it, I realized it was AI generated because by the way, this person was doing something pretty sketchy. I think they used the name of another famous artist as the artist's name. So that's how they got streams on their song. But their song was pretty good. I actually really liked it. It was very wholesome. But as AI generated. But the other day, I was actually on YouTube and I found a whole playlist of this like folk music and the style I like. And it was AI generated and I clicked it and I honestly sounded great. So I listened to it for probably like half an hour. My wife was laughing at me, but I was like, honestly sounds good. So I can see it both ways. And sometimes it's less obvious if it's instrumentals. It's not very obvious relaxing piano music that's AI generated sounds just like normal piano music to me. But if there's like singers and different instruments like guitar twang, I can usually hear tell AI generated. That's also because I'm generating a lot of stuff. So I get a good understanding of this. So there is money to be made and different streaming platforms are taking it different ways. We have a title that said if you use AI generated music, they're not going to pay you streaming royalties. So you straight up just won't get paid on their platform. But at the same time, a lot of people are making content that people do like. Now, where is the place for AI in music, in AI and AI in music, and what's the relationship with Spotify? Came up with a bunch of terms recently that basically, I mean, at the end of the day, if people like it, if it's high quality, then it's going to be streamed. My personal opinion is if you have content that sounds amazing, that is beautiful, using AI to help you with, I don't know, for example, my wife is an artist and she releases all her own music, and I'm not going to lie, one of the biggest struggles we have, and it's just because this isn't our ballpark, I'm a podcaster and she's a full-time mom for the most part, she's really good at writing songs, they sound amazing, she plays them live, but when it comes to recording, the recording process, the mixing, the mastering, all of that is basically the bane of our existence, we hate it, and I haven't gotten around to paying a producer to go and make it all sound good, or even to record it all, because sometimes I'm like, shoot, the AC unit was on when we were recording, and there was a hum in the background, and we got to put some compressors on it, and I don't know really what I'm doing, if I'm being honest. So the problem is, I wish there was a way to, if she could just record her music, how she does, put it in some sort of AI platform, and it goes and fixes everything for it, to make it sound good, but it's still her, it's still her playing, I would say, right on, and that's honestly what a lot of plugins, even on a lot of these music editing platforms do, they're going and kind of manipulating things, even there's like Flex Pitch or there's, you know, Melodyne that's like auto correcting your pitch a little bit. So like, there's all of these things and ways that quote unquote AI or computer software has been like used in people's music forever. And somehow when it's like a kind of somewhat autonomous, people don't like it. But I think if you're a creative artist and you're using AI to help you with your flow, I mean, you're just using software. That's fantastic. I would even go so far as to say if you use Suno and just one shot at a song and you love it, I mean, it would be pretty cool if you could go find an actual person to record it and sing it and maybe polish or tweak a couple of the words that might sound weird or cringey or off and, you know, have like a real person make it. I think that's amazing. Now, right now, I'm saying this all from the standpoint of AI generating music. I can, when I listen to it, I can tell it's AI generated, but that won't be the case forever. These models are getting better and better. The twang of the guitar is going to start sounding more natural. It's going to be just like how the tone of Claude when it talks has gotten so much better since the very early days. So I think a lot of things are changing there. What's interesting is the share of AI generated music has climbed from about 10,000 tracks a day on Deezer in January of last year. So just last year, 10% AI, and it is now 40, and then it went to 44%, which is 75,000 tracks a day in April. And now we're at over 50%.