‘Technology is changing the world we live in, from the way we discover music to its creation, AI will bring innovation but also big questions for the industry,’ says Emma McClarkin.
Last week PRS for Music hosted an evening to explore how artificial intelligence (AI) might impact music creators today and what effect it might have on the music industry’s future.
PRS Explores: Artificial Intelligence - Melody + The Machine was led by former MEP and technology and international trade specialist Emma McClarkin, who was joined by a panel including classical singer and co-founder of creative services company FeedForward AI, Lydia Gregory, and Audio Network's chief product officer, Matthew Hawn.
The panel also included Lord Clement Jones, chair of the House of Lords Select Committee on Artificial Intelligence, and Andrew Burgess, author of The Executive Guide to Artificial Intelligence.
On the evening Emma stated: ‘Just as the UK leads in music so we should in our understanding of AI and the impact it could have.”
Here’s what we learned…
We need to move beyond archaic views of technology…
Lydia Gregory: ‘All of this isn’t helped by these sci-fi tropes that we’re all familiar with in the everyday world. We think of it like robots might be coming after us, they’re coming after our jobs. What are they going to do? But really what I want to convince you of today is that artificial intelligence technologies are already being used in the music industry in really useful ways to solve problems. But also, if we dig a little deeper into what the technology is and what its capabilities are, we can actually we can see there are opportunities to unlock new ways of working with music, both in the creative process and for solving problems.’
But be wary of overestimating what AI is capable of…
Andrew Burgess: ‘When we talk about generating music, this is where it gets a little bit controversial. What we’re trying to do is predict the next note or what’s going to happen next in the piece of music. Of course, that’s all very derivative because it’s based on what somebody’s done in the past. So that’s where we need think about whether AI is really creative or not.’
Matthew Hawn: ‘I’m a bit of a sceptic in terms of the idea of artificial intelligence. We tend to think of computers and want to ascribe human things to these machines. The anthropomorphic nature of what we try and turn into intelligence is, to me, a mistake. I don’t hear anyone saying we want artificial creativity. Artificial intelligence is taking existing learning, but that’s not what musicians do. Musicians are human and bring emotion to bear on so many things. Anything a machine learns is only a reflection, an echo of something that’s come before, it’s not truly crating something new. In that sense, it’s an instrument. It’s an instrument that can be played by someone with enormous creativity.’
The use of AI in music raises complex copyright issues…
Matthew: ‘I think we need to take some guidance from 20 years ago. We don’t need to go to computers for this, we can just go to Paul’s Boutique and the Beastie Boys cutting up 7,000 samples to create a new work. Where so you start to create a new work? When does that happen? And that didn’t need a computer to do that, those guys just were chopping up pieces, it was a human effort. But it was the same thing. So, at some point you’ve got to decide how much of the work you use, and these are all well-established parts that copyright’s very good at. So, I wouldn’t try to reinvent that just because there’s a computer involved, but I do think we’re going to have to be more granular in copyright and think about different types of copyright. They need to be in service of the creativity of the people who create them.’
It’s important to fully explore the ethical ramifications of AI…
Lord Clement Jones: ‘In a sense the essence of this is the rightful share of the creators of the value of creation. Whether it’s about copyright infringement or about other aspects of having that fair share, that itself is a very ethical question. But really the other issue, is there are choices to be made. Margaret Boden, who is one of the great gurus in this area, says, ‘it isn’t just what AI can do it’s what it should do.’ You do have a choice. The absolute crux of artificial intelligence is that it has to be our servant, not our master.’
Lydia: ‘In the music industry I think one of the challenges is that everything we can do is based on data. So there’s two choices that can be made there: the first is that the data that we have to work with today a lot of the time hasn’t been generated or collected with the aim of doing something with machine learning in mind, so there’s a gap between the aspiration of solving a particular problem and the data that’s available in order to learn that, ten it comes down to which data we want to collect. A lot of the problems we want to solve in the industry are valid problems we want to solve and they’re really exciting. The word data, I think we always think it’s got some personal component, and that’s not always true, it’s not necessarily tracking what you’re doing. Data can mean a whole bunch of different things. But there are choices about what data we’re happy to be collected.’
PRS for Music launched PRS Explores in 2016, with the aim of facilitating debate about change within the music industry. Previous topics have included The Music Modernization Act, virtual and augmented reality, blockchain, and the EU Copyright Directive.
To watch the full stream of PRS Explores: Artificial Intelligence - Melody + The Machine, please visit prsformusic.com.
Last week PRS for Music hosted an evening to explore how artificial intelligence (AI) might impact music creators today and what effect it might have on the music industry’s future.
PRS Explores: Artificial Intelligence - Melody + The Machine was led by former MEP and technology and international trade specialist Emma McClarkin, who was joined by a panel including classical singer and co-founder of creative services company FeedForward AI, Lydia Gregory, and Audio Network's chief product officer, Matthew Hawn.
The panel also included Lord Clement Jones, chair of the House of Lords Select Committee on Artificial Intelligence, and Andrew Burgess, author of The Executive Guide to Artificial Intelligence.
On the evening Emma stated: ‘Just as the UK leads in music so we should in our understanding of AI and the impact it could have.”
Here’s what we learned…
We need to move beyond archaic views of technology…
Lydia Gregory: ‘All of this isn’t helped by these sci-fi tropes that we’re all familiar with in the everyday world. We think of it like robots might be coming after us, they’re coming after our jobs. What are they going to do? But really what I want to convince you of today is that artificial intelligence technologies are already being used in the music industry in really useful ways to solve problems. But also, if we dig a little deeper into what the technology is and what its capabilities are, we can actually we can see there are opportunities to unlock new ways of working with music, both in the creative process and for solving problems.’
But be wary of overestimating what AI is capable of…
Andrew Burgess: ‘When we talk about generating music, this is where it gets a little bit controversial. What we’re trying to do is predict the next note or what’s going to happen next in the piece of music. Of course, that’s all very derivative because it’s based on what somebody’s done in the past. So that’s where we need think about whether AI is really creative or not.’
Matthew Hawn: ‘I’m a bit of a sceptic in terms of the idea of artificial intelligence. We tend to think of computers and want to ascribe human things to these machines. The anthropomorphic nature of what we try and turn into intelligence is, to me, a mistake. I don’t hear anyone saying we want artificial creativity. Artificial intelligence is taking existing learning, but that’s not what musicians do. Musicians are human and bring emotion to bear on so many things. Anything a machine learns is only a reflection, an echo of something that’s come before, it’s not truly crating something new. In that sense, it’s an instrument. It’s an instrument that can be played by someone with enormous creativity.’
The use of AI in music raises complex copyright issues…
Matthew: ‘I think we need to take some guidance from 20 years ago. We don’t need to go to computers for this, we can just go to Paul’s Boutique and the Beastie Boys cutting up 7,000 samples to create a new work. Where so you start to create a new work? When does that happen? And that didn’t need a computer to do that, those guys just were chopping up pieces, it was a human effort. But it was the same thing. So, at some point you’ve got to decide how much of the work you use, and these are all well-established parts that copyright’s very good at. So, I wouldn’t try to reinvent that just because there’s a computer involved, but I do think we’re going to have to be more granular in copyright and think about different types of copyright. They need to be in service of the creativity of the people who create them.’
It’s important to fully explore the ethical ramifications of AI…
Lord Clement Jones: ‘In a sense the essence of this is the rightful share of the creators of the value of creation. Whether it’s about copyright infringement or about other aspects of having that fair share, that itself is a very ethical question. But really the other issue, is there are choices to be made. Margaret Boden, who is one of the great gurus in this area, says, ‘it isn’t just what AI can do it’s what it should do.’ You do have a choice. The absolute crux of artificial intelligence is that it has to be our servant, not our master.’
Lydia: ‘In the music industry I think one of the challenges is that everything we can do is based on data. So there’s two choices that can be made there: the first is that the data that we have to work with today a lot of the time hasn’t been generated or collected with the aim of doing something with machine learning in mind, so there’s a gap between the aspiration of solving a particular problem and the data that’s available in order to learn that, ten it comes down to which data we want to collect. A lot of the problems we want to solve in the industry are valid problems we want to solve and they’re really exciting. The word data, I think we always think it’s got some personal component, and that’s not always true, it’s not necessarily tracking what you’re doing. Data can mean a whole bunch of different things. But there are choices about what data we’re happy to be collected.’
PRS for Music launched PRS Explores in 2016, with the aim of facilitating debate about change within the music industry. Previous topics have included The Music Modernization Act, virtual and augmented reality, blockchain, and the EU Copyright Directive.
To watch the full stream of PRS Explores: Artificial Intelligence - Melody + The Machine, please visit prsformusic.com.
