The Music Business Buddy
A podcast that aims to educate and inspire music creators in their quest to achieving their goals by gaining a greater understanding of the business of music. A new episode is released each Wednesday and aims to offer clarity and insight into a range of subjects across the music industry, all through the lens of a music creator for the benefit of other music creators. The series includes soundbites and interviews with guests from all over the world together with commentary and clarity on a range of topics. The podcast is hosted by award winning music industry professional Jonny Amos.
Jonny Amos is a music producer with credits on a range of major and independent labels, a songwriter with chart success in Europe and Asia, a senior lecturer in both music creation and music business at BIMM University UK, director of The SongLab Ltd and the author of The Music Business for Music Creators.
www.jonnyamos.com
The Music Business Buddy
Episode 9: Navigating Ethical Waters in AI-Driven Music
What if AI could compose music just as emotionally compelling as your favorite artist? Join us as we uncover the transformative power of AI in music creation with insights from Emily Jackson of Horus Music. This episode dives into how innovative tools like Suno, Boomi, SoundDraw, and Google's latest advancements are reshaping the music industry. Discover how AI is not only democratizing music production but also expanding the creative horizons for music creators by generating unbiased musical suggestions. We'll discuss the longstanding role of AI in music through MIDI messaging and advanced composition plug-ins, setting the stage for today's revolutionary developments.
But it's not all smooth sailing—there are ethical complexities and intellectual property issues that need to be tackled. Can an algorithm legally mimic the style of a band like Coldplay? What rights do original artists have in this new landscape? We'll explore these pressing questions and the cautious approach of streaming platforms towards AI-generated content. Emily Jackson provides a balanced view, highlighting both the opportunities and challenges that AI brings to the table. As we look toward the future, we ponder whether original creators will reclaim their dominion in an AI-influenced world. This episode promises a thought-provoking journey through the evolving landscape of AI in music creation.
The Music Business Buddy. The Music Business Buddy Hello everybody and a warm welcome to you. You're listening to the Music Business Buddy with me, johnny Amos, podcasting out of Birmingham in England. I'm the author of the book the Music Business for Music Creators, available in hardback, paperback, ebook format. I'm a music creator with credits on a variety of major and indie labels. As a writer or a producer, I'm also senior lecturer in music creation and music business. Wherever you are, whatever you do, consider yourself welcome to this podcast and to a part of this community.
Speaker 1:I'm here to try and educate and inspire music creators from all over the world in their quest to achieving the goals that they set for themselves by gaining a greater understanding of the business of music. So the subject for today is artificial intelligence in song creativity, and I'm not here to kind of cast any judgments on any of this kind of stuff, but more so just to kind of draw some commentary on it, to simplify it and to just understand it a little bit better, because there's quite a lot to it. You know, there's a whole host of AI-powered music generator tools out there, the likes of Suno and Boomi and SoundDraw, which is now integrated into Canva, and not to mention what Google are building. So there's a lot of different players out there that are building amazing tools for music creators and for people that don't create music. There's a lot to it, so let's try and dig into it a little bit, but let's start off with some simplicities. Okay, so in my book, the music business for music creators, I talk about how ai is not the future, it is the present. It's here. It's changing many aspects of the music industry, in fact many industries. Uh, in a way that is kind of a little bit difficult to try and sort of fully calculate and consider. But AI is a subject that sits right at the top of the mainstream agenda in the music industry and there'll be a lot of changes in the next decade. So it's always good to kind of get ahead of the curve, just to kind of understand, you know kind of what things might look like in the future, and I'm not really sure we can do that without fully understanding where it is in the present. So that's what we're going to try and do today, so we're going to look into it a little bit more. And why this, this subject is, is heavily opposed by the music industry. Um, for good reason in. You know as well. So we'll look at that too.
Speaker 1:But, you know, let's just go to the basics for a minute. You know, because the role of artificial intelligence in creativity, you know, it could actually be argued that it's nothing particularly new in the context of music creation, especially for music producers. You know, there's all sorts of tools like probability tools, and especially with MIDI messaging and stuff, there's actually, quite frankly, nothing new. It's been around for a good 40 years or so, but it's the way in which it's now being used to create new music. That's changing a lot of things. So music creators, you know, have a whole host of plug-ins that are available right that can generate patterns for song composition and arrangements, and it's all achieved through research on genre-specific patterns and how they relate to sort to simple and quite advanced mathematics, and the results of these can be turned into grooves that can then be assigned to software instruments of a creator's choice, which then puts generated patterns in the hands of composers to do as they see fit. Now, that's a slightly different thing than some of the other stuff that grabs the headlines in AI, but it is all part of the same kind of subject matter.
Speaker 1:So you know, it's easy for a music creator's judgment of a song to become a little compromise when they're so heavily involved with the creativity. Love is blind when it's so close. When we create a song that we love, we don't always see. You. You know the truth behind, whether it is actually any good or not, you know. Now that might be even more true if a music creator has a song on a record or played an instrument or been involved heavily in the arrangement. Now, there's nothing wrong with those things. We need those things. They are the art.
Speaker 1:However, there is perhaps a chance that ai could generate more suitable suggestions than what a human can come up with, especially when it comes to new parts for an arrangement. Now, ai does not hold a bias. Perhaps, in the same way that a person does bad thing, but a person's performance and idea generation are often, you know, a little bit conditioned by mood or sleep or lack of it, or a personal preference or a personal feeling on that particular day. Again, that's where the art can sometimes come from. So we have to get the balance right between these things. But there's a lot to this subject. So, uh, if you remember, back in episode five, when we talked about metadata I um. I ran an interview with emily jackson from horus music and I asked for her viewpoint um during that interview on the subject of ai in creativity, and this is what she said so I think this question can be approached from lots of different perspectives.
Speaker 2:So, starting from a positive perspective, it definitely allows for more creativity. There will be endless opportunities to create music in different ways with resources that may never have been known or available to an artist before. The accessibility of those tools as well, so they're becoming easier to use and more available to the DIY musician. The tech itself isn't just reserved for big budget releases. You know things can be done on a smaller budget now.
Speaker 1:Okay, well, that sounds very positive, it sounds very inclusive. So why is it we're seeing such a huge pushback from the music industry on this subject? Let's dig into this a little deeper. Okay, let's start simply Now. Imagine if I wanted to create a song in the style of Coldplay. I could create an algorithm which would analyse 30 Coldplay songs. The algorithm would then analyse, by running pattern analysis on typical movements, typical lyrics, typical chords, typical arrangement styles, and then build something new out the end of it.
Speaker 1:Now, if we think about the input and the output here, right, so the input are the Coldplay songs. The output is a song that Coldplay didn't write. Now here comes the confusion. If I then go and say, right, I've created this song, yeah, kind of sounds a bit like Coldplay, but you know, that's what I wanted, I could go ahead and release that. Now let's just think about the intellectual property, because there is the issue. If Coldplay songs have been used to create a Coldplay style song, then Coldplay or the writers and the rights holders need remuneration. Until that's fixed, this will never, ever go away.
Speaker 1:Now it could be argued that well, you know, coldplay didn't write that song, it was just based on the typical things that they do. However, if we remove those songs from the first place, we don't get an output. So there you know. There is a problem, right, and the music industry is doing everything it can to battle this, and rightly so. And if anybody ever gets wound up by that kind of thing, then just remember that if those were your 30 songs that were being used to create new songs elsewhere, then you probably wouldn't be very happy about it. So it's good to try and get some kind of balanced, informed understanding of why this is such a big issue for so many people right now. Now a further problem. As if there aren't enough problems already, here's another one. Uh, is that that third party company that designs that algorithm that analyzes the music and then kicks out new music also have intellectual property on the software that they created to analyze the songs? Are you still with me? Ok, so if I pay I don't know, £20, £30 for a plug-in for a piece of software to analyse those songs, then the company that owns the patent to that generator are going to get paid a licence fee for that. There's no back-end collection for them, of course. I've already paid them to use their software. The trouble is, of course, we're so early in on this right now, in the mid 2020s, that there hasn't really been a system of fairness put in place yet that is recognised by all parties.
Speaker 1:But there is an overriding positive here for music creators and it boils down to this for music creators, and it boils down to this when we create music, we generate ideas and then we figure out which ones we want to keep. Now, if we just think about writing a song or creating a piece of music as those two simple facets, right. So generating ideas and then deciding what we keep, they're actually two quite different skills, aren't they? One's creativity and the other one is detectivity. So if we come up with five ideas, maybe that's enough for our song, but maybe if we come up with a giant slew of 25 ideas and we're trying to skim it down to one song, then we're probably only going to have to keep five or six of those 25 ideas. Now, that is something that ai doesn't really solve. Maybe it will in the future, but where it's at right now, it doesn't, because it's still down to us as humans to pick which of the generated ideas we like.
Speaker 1:So, for example, if I've got four chords and I go right, let me help. Let me use um, I don't know a generative midi tool to be able to create this or that style, or put it in this genre, put it in that genre. They generate the style. I then have to decide do I like that or do I not? Now, in many ways, that's not that different from working with a musician next to you oh, could you play it more like this, or could you play it more like that? When you've got a computer in front of you that's generating things. It means we can say what we really feel without having to offend people.
Speaker 1:There are advantages to all of this. Of course there are, but there might also be a downside in as much as that. From a creative perspective, it can encourage a little bit of procrastination because we're coming up with so many different ideas, our ideas, the ideas that are generated before us, in front of our eyes, on our very screens, and that can slow down the creativity process. That depends on the kind of person that you are, of course. Anyway, let's take a look at kind of the implications of this right now in the 2020s for music creators. So let's have a look at where this sits in the context of streaming. I asked Emily Jackson from Horace Music about this subject, and this is what she said.
Speaker 2:Looking at the streaming market specifically from a distribution perspective, we found that platforms are approaching AI quite cautiously. There's nothing that's been formalised really in terms of policy as yet, or any guidance on what they will or will not accept. We have found, though, that there are some companies, entertainment companies particularly who are already investing into their own technology. So, for example, there's the entertainment company who have developed a tool which can change an artist's voice in real time. Um, so, again, that can feed into creativity, but you know, you have to question how has that tool learned?
Speaker 1:how has that tool learned? Emily's absolutely right. This is the big problem for many, many music companies, for many rights holders, music creators, artists, labels, you know if, if something is examining performance as well as compositional method and then reusing that to make it available to other people, then there is a problem if it's not being identified correctly, because one of the most popular methods for creating ai powered music tools uses machine learning algorithms, particularly deep neural networks, that analyze large data sets of existing copywritten music and then generates new compositions based on that analysis. Now I think there's a key difference between something and something else here. Right, let's just split this into two different things. There is music that we go right, hey, make me a song of this style with this mood, boom, and it does it. That's one thing. Another thing is where a music creator uses AI-powered tools to co-collaborate on a new idea.
Speaker 1:Now, for me, they sit in two very, very, very different areas. In fact, even if we link that through to intellectual property, we'll see a difference. For example, the US Copyright Office requires that all works registered for copyright have a human author. If the works, elements of authorship ie originality, creativity are created by a machine, then it's not protected by copyright. Now, that's an interesting one, isn't it? Because if it is a human that uses a machine to create a song that's composed by a human, then is that okay? Well, I guess the answer right now is yes, that is okay. You see, it goes back to the difference between those two different things music driven by data sets or music that's co-collaborated by a human or decided on by a human driven by an AI powered tool. I know there's a lot to this, but I think there's a key, distinct difference between those two things there. So I asked Emily, and this is what she said about it.
Speaker 2:We need to consider how the AI learns, what educational material it learns from, so which songs it listens to, what music it's fed into to learn, and then whether that material is being paid and treated fairly in terms of royalties or licenses, and making sure that the original creators are either credited or paid and remunerated fairly.
Speaker 1:And that last point that Emily just made there is the absolute epitome of what the problem is right now. Just so we're aware, and I'm just going to roll back and play that bit again.
Speaker 2:Making sure that the original creators are either credited or paid and remunerated fairly.
Speaker 1:OK, so how can that be achieved? Well, I'm not quite sure that I have the answers right now. I'm so sorry to say that, but I'm not sure anybody does. However, there are some positive steps here moving forward. Either what Emily states starts to happen through legislative changes, through technological changes and through cultural shifts, or and I'm just going to throw this out as a couple of additional ideas maybe artists and labels begin to license their performances for AI use. Again, it would need regulating, it would need the right technology for it to happen, but if that were to happen, then that can allow access to those files on a legal level. Think of it as like a hybrid between a bootleg remix and an approved remix.
Speaker 1:Now, let's assume that that were not to happen. Let's take another path. Another path might be that songwriters, music creators, music performers begin to gain new work creating ethically clear data sets for use in copyright free music. This is something that has started to happen already, but if that increases, then that gets away from the need to remunerate rights holders. The byproduct presumably of that and I'm guessing here is that people start to get bored of that because it's not music or songs that they feel familiar with.
Speaker 1:It is then at that point that this whole subject of AI and creativity starts to dwindle, just like NFTs started to dwindle three years ago. And then we move into a different era and nobody ever talks about this again, and original creators are the only creators of original music, do you like? The sound of that? Sounds quite plausible, doesn't it? Who knows which way the wind blows. Ah, anyway, thank you for tuning in that. We'll revisit this subject in the future. We have to because there's so much to it, right, but for the time being, thank you for tuning in. Thank you for being here as part of this community, and may the force be with you.
