Interview with Martin Clancy
by Killian Laher

Martin Clancy’s book, Artificial Intelligence and Music Ecosystem is out now.

“Artificial Intelligence and Music Ecosystem highlights the opportunities and rewards associated with the application of AI in the creative arts.

Featuring an array of voices, including interviews with Jacques Attali, Holly Herndon and Scott Cohen, this book offers interdisciplinary approaches to pressing ethical and technical questions associated with AI.”

Can you give me a little bit of your background?

Martin Clancy (MC): My name is Martin Clancy and I have just published this book, which I’ve edited. I’ve written three chapters and conducted a series of interviews which make up another three chapters. The other contributors are leading figures in the world of music and AI. That would include robotics, global ethics, computer science, data, industry, artists like Holly Herndon, philosophers like Jacques Attali, who’s written a hundred books and has a column in Le Monde and was an economic supervisor to Sarkozy and Mitterrand. All these people are giants in their particular field. They all responded to my call to say, what do you think about AI and music and the way it’s going and how do you see the future from your area of speciality? That came as a response from my PhD, which was awarded last year. That was where I researched the ethical and financial implications of AI and music.

Where that came from was because I teach a piece of software called Ableton Live. I’m an Ableton certified trainer. I teach a new generation of DJs and electronic music makers the software that they use for live performances and making stuff. I noticed that this younger generation of musicians, when it came to dealing with technologies that had AI written on them, their behaviour changed slightly, and I can talk to you more about that. I’m in my late fifties, I left school when I was 15 and went on the road and learned about music that way. I’ve basically always made a living from music, doing various different things. I’m also a manager, so I manage an Irish artist, Jack L, and I work both as a practitioner and as an academic. I’m also now heavily involved with AI ethics, I’m the chair of the IEEE Global AI Ethics Committee. The IEEE is the biggest body in the world, it’s 400,000 people. In that capacity, I get to speak, along with the committee, to people like the UN and the EU about the arts and AI and the future and stuff like that.

Am I right in saying that you were in In Tua Nua?

MC: I formed the band when I was in school, about 14. There was myself and one other person (Ivan O’Shea) and we were what you would call that stage post-punk, new wave. What happened was I wrote a song and I was introduced to a young singer (Sinead O’Connor) and it was her first recording and that actually kind of ‘blew up’. The next thing I knew we were in a seven-piece band with violins and uilleann pipes and stuff, and that went on for about ten years.

The first signs of AI were things like the Roy Orbison hologram tour and the Beatles’ Love album…

MC: AI is an umbrella term for a bunch of different technologies, primarily machine learning. You’re right to identify the Roy Orbison and the Love album as the first indicators of a different approach and the shape of how things are certainly going now with AI. Of course, you have tribute bands, which we have of many dead singers. While that’s got nothing to do with AI, that’s definitely got to do with this idea of recreating something that is no longer real. The logical extension of that is the Abbatars and the success that’s had where they use the latest technology to build a virtual band.  What’s interesting about that is everybody’s raving about it. Artistically, economically, you can’t argue with it, but it’s not hard to imagine other versions of that being done where somebody is basically resurrected without their permission and it’s being done in a tacky kind of way and on the cheap as well. So you can definitely see where you go, we haven’t got the budget for that, but why don’t we just bring back X, use a bit of AI and see if we can make a few bob from it. That’s up to people to say, I’m not buying a ticket for that.

Maybe the purists won’t buy it, but others will.

MC: I have great faith in the common taste, I love popular culture so I think if something is genuinely crap, people leave it. Look at the fad of 3D movies, how much went into that? Or the metaverse? It’s probably too early to say that something that wasn’t invented is already obsolete!

There’s a lot of fear around AI and with good reason, but you can have overwhelming fear too and sometimes it’s more difficult to find a light and to lean into positive examples of people working with technology. I think that’s probably where the hope for the future comes, apart from all of this humancentric stuff, which is questionable. Ultimately if we as human beings say, no thanks, then the share prices of these corporations will drop and they’ll just have to get in line.

What’s driving the move in this direction? Is it commerce, is it art? Is it creativity or what is it?

MC: It’s not art, let’s be clear about that. It is innovation driven by finance, as always. We flew to the moon because there was an arms race. Even the internet for instance, was developed by the US defences, in the event that if all communication systems collapsed then how would the military be able to communicate? That was the root of the internet, which then became World Wide Web, which is built on top of it. Technologies are invented primarily for commercial reasons. But then the interesting part is they get repurposed and people go, oh my God, let’s use that for this.

I’ve not told anybody this including any members of my family, but the first album I ever had was The Wombles’ first album. I was thinking about “making good use of things we leave behind, things that the everyday folks leave behind” thing the other day. The idea is repurposing, making things new out of old things. The history of music is always about human curiosity: Here’s a piece of technology and someone going, I like the sound of that, even if it’s broken. Examples of that would be distortion. Famously, a bass player in the late fifties was working in a session band, and his amp was broken and then he went, that sounds alright. Ten years later you get Jimi Hendrix creating a new art form.

The same thing applies when you look at hip-hop and the invention of the DJ culture.  They turn their record players into a musical instrument. The same thing with glitch and computer music. Creatively, the art comes from curiosity and people going, if we turn that picture upside down, that looks better, you know what I mean? These tools are going to be great fun. The question is for those that work in the arts, how will anyone ever be able to earn any money to pay for any of it? That’s the concern. Ireland was the first country to introduce a basic support scheme for the arts in September, and that had nothing to do with AI.

Have you seen any of the text-to-image generation? This is important because this is very recent, in the last five months, this has blown up. You have a range of AI where basically text image generation, and the three main ones, one is DALLE-2, Imagen and Stable Diffusion. What this consists of is free open source software and you text in a prompt like an astronaut on a horse in the middle of a desert. We write it in the style of Picasso or in the style of Da Vinci, or Van Gogh. Immediately four images appear. The quality is astonishing. Literally offering photo realistic. I was showing my students the other day and one of them typed in Boris Johnson kissing Donald Trump and lo and behold, there’s a photo! With a bit more work it would be a perfect deep fake. This is exciting, but also potentially scary. While it’s open source, it’s been baked into 200,000 different products and services in the last three months.

“An Astronaut on a horse in the middle of a desert” – Created using DALLE-2

Things like photoshop, things like PowerPoint. These text-to-image generators will replace search, probably. It’s spreading. Coming up next is text-to-video, which is being introduced. You type in text and you’ll have a full animation of that text. Text-to-music is due to be released by the same company, in a couple of weeks. That quick. I don’t know what that will sound like, but… Three months ago, certainly at the end of the summer, when people were looking at these images, say, for instance, on the interview, you can tell that’s AI because, look, when you look at the hands, it’s all thumbs. The AI wasn’t particularly good at hands then, but it is now.

You just give it a little bit of time. That time is weeks or months, not years. That’s the part that’s very interesting. That’s new, that’s not in the book. But the book does speak about a number of those companies that are introducing these technologies. Snapchat and instagram brought out filters that are using these text-to-image generators.

Is the music business ready for this? It doesn’t have a history of being ready for advances in technology.

MC: That’s a very polite way of putting it. A couple of thoughts on that one. I don’t call it an industry, I call it a music ecosystem. There’s a particular reason for that. ‘Music industry’ suggests that there are such things as professionals and amateurs. Even though I’ve always worked in music and I’ve been paid, I take pride in a certain amateur approach to things. Even though I have a PhD in music I still don’t like the snobbery of, this is a concert hall as opposed to a club, you know what I mean? When Jack L was in the 3Arena at one point, between the band and him and the 70-piece choir there was close to 90 people on the stage, and some of them were professional musicians, who do it full time. Others would be amateur enthusiasts, others would be musicians who are gigging musicians but have other jobs.

It’s one big beautiful noise, as the philosopher Neil Diamond would say! That’s the music ecosystem, including the audience who are listening to it, and anybody who cares, particularly the listeners and the community, from amateur choirs to people who have dedicated their lives to academia, that’s all the music ecosystem. COVID showed us that the system was fucked and the system was broken because of the things that you spoke about, which were past technologies, the introduction of past technologies. The reason Jacques Attali contributed to the book is that he wrote another book in 1977 called Noise, where he said, music is a herald for social change. What that means is music is the ‘canary in the coal mine’ for new technologies. I asked him: How the hell did you predict this?

He was basically saying what happens in music, it happens there first. It goes to other industries after. When you think about Napster, and you think about how for the likes of Google, I don’t mean this critically because they’ve acknowledged it, Google were able to build their unimaginable platform on the basis of loopholes that existed in law, and through music. The other side of that is when you look at all the job losses in tech recently, there’s no longer a separation between the haves and the have nots.

I think that can be a force for positive change in the sense that I’ve seen a change in computer sciences, even in the time that I’ve been doing research.  With music research, when you put it to the computer scientist, you’re saying, why are you making that stuff? They go, well, we love music. And you go, okay, so why don’t you make it a design feature that people can make some money.  It’s not that it changes things, but a kind of moral responsibility comes in and then it can be like, how do we design something better? Any fool can design a gun, I would imagine, but it takes a bit more wisdom to want to be a doctor.

It’s too easy to say ‘it’s over, creativity is dead’. But it isn’t.

MC: As soon as you make that statement that’s true. As soon as you go, no, it isn’t, that’s how punk started, that’s how hip hop started. Let’s have a look at these tools, what are you going to do? The speed of change is fast and that’s why we see companies like Facebook, Google and TikTok, suddenly appear. When you look at when they were founded, you go, wow, that was in 2004, or that was only 2000. These new companies that are appearing, that are getting huge funding are appearing and rapidly growing, and that’s how exponential change works.

The Interview continues in Part 2.

Martin Clancy’s book, Artificial Intelligence and Music Ecosystem is out now.