Tallinn Music Week Conference 2023: Future Radio, Getting Smarter Every Day

25 October 2023 Conference

Photo: Aron Urb

Nordic Hotel Forum Conference Centre
13th May 2023

Despite the MTV era’s “killing of the radio star” or the Internet era’s rise of on-demand services, radio has been an ultimate survivor that has patiently evolved with the changes in tech and consumer habits. The rise of podcasts, connected devices and streaming services’ dependence on algorithms to curate playlists and make user recommendations, have led to changes in how people listen to and interact with radio. It has embraced new tech to maintain its relevance and remained a key medium for news, entertainment and music in many countries. Simultaneously, a new generation of community stations and the kids of 20th Century pirates have utilized the Internet to share underground vibes legitimately with little more than a laptop and a microphone. 

However, both FMs and online channels are now entering an AI era of sleek GPT hosts with no ‘ums’ and ‘erms’ waiting in the wings. Will these brave new robots ever spin emerging indie artists? How can star broadcasters and cult tastemakers future-proof themselves?

Learn about it all by this panel of radio stars. 

Cristoph Lindemann, Head of Music at PULS (DE)
Ash Kilmartin, Radio WORM (NL)
Ruben Jonas Schnell, Founder & CEO of ByteFM (DE)
Hervé Riesen, Deputy Director for FIP / Radio France (FR)
Tanja Douglass, CEO of Radio Helsinki (FI)

Yousif Nur, music and tech journalist (UK)

The panel is re-watchable for the holders of PRO and DigiPRO passes until TMW 2024 via the TMW PRO platform. PRO passes for TMW 2024 are on sale at the TMW webshop.

Yousif Nur. Photo: Aron Urb

Yousif Nur (YR): Good afternoon, everyone! Welcome to the “Future Radio, getting smarter everyday” panel. I’m joined by some very esteemed guests. The intro was pretty interesting from what I read on the website: “Despite the MTV era’s “killing of the radio star” or the Internet era’s rise of on-demand services, radio has been an ultimate survivor”. But I would like to know for myself, within the next hour, can it become the ultimate survivor with AI looming large and ever-present in our lives. With that in mind, I’d like to introduce my guests or hopefully they’d like to give a small introduction. 

Christoph Lindemann (CL): Thank you very much. My name is Christophe Lindemann. I used to be Head of Music at the public radio station in Southern Germany in Bavaria. Now I’m a music strategist for different music products that we’re offering because we restructured our radio team a little bit, which I’m going to explain later. 

Ash Kilmartin (AK): My name is Ash Kilmartin. I am the programme manager and programmer at Radio WORM in Rotterdam. WORM is a multifaceted platform for all sorts of culture.

Ruben Jonas Schnell (RJS):  Hi, my name is Ruben Jonas Schnell and I’m the founder and programme director of Byte FM that started as an online station in Hamburg 15 years ago, and now we broadcast also on FM in Hamburg and digitally in Berlin. We are commercial-free and listener-funded.

Hervé Riesen (HR): Hello, my name is Hervé Riesen. I’m the Deputy Director in France for the musical public service channel called FIP. I am also the Chairman of the Eurosonic group at EBU [European Broadcasting Union – TMW].

Tanja Douglass (TG): I’m Tanya Douglass and I’m the CEO of LiveLaboratorio, which is a company that owns Radio Helsinki. It’s a commercial cultural station and also a live music venue GLivelab Helsinki.

YR: What’s the typical demographic of your stations, who would be your typical listener?

CL: Our typical listeners would be above 30, but our job is to reach 16 to 25-year-olds.  We did not manage this with the radio, so we thought of new ways of doing that. 

AK: For us, it’s the old punks and their kids.

RJS: When we started, we didn’t want to address a certain demographic with our programme. We decided to focus on people who are interested in culture and music. However, by the time we broadcasted, we learnt that our listeners are from 20 to 60, I would say, with a priority on those between 35 and 45. A little more male than female, which is too bad.

HR: The main target is between 35 to 50. We are lucky because we have a really good 50/50 balance between male and female

TD: Ours is about 50/50 male, female. And about 25 to 55 years is pretty equal. Well, we’ve been around 22 years, so people that started as young listeners are still with us. We’re trying to keep up with the younger ones and their interests.

YR: How do you build up the loyalty from the people who listened to you from when they were 17 and wanted to be rebellious up to now, being middle-aged, as it were? How do you keep them listening?

TD:  I think they are the easy ones to keep because they have been doing it. The loyalty has been built over the years. I think the struggle is more in getting new listeners from the younger groups. But I think it’s about whether we’re doing what they’re interested in and if we’re easily found for them. The younger people are not going to find the FM station. We have to be available on different platforms and in different areas in different ways.

YR: Bringing AI into this, what are your concerns or challenges dealing with AI moving forward?

RJS: I’m not too concerned about us when it comes to AI because all our programmers are enthusiastic. That goes back to your question about what brings loyalty. We are enthusiastic about what we do and we take the listeners seriously. I think that’s the main reason why we have a strong loyalty to our programme. As long as we keep that, I think we’re all set when it comes to AI. However, when it comes to mainstream radio, then I think they have to be very concerned about AI. 

I’m expecting technology to be so far ahead already that you will have announcers that are artificial but sound human. If they talk about the same things every day, including Facebook or news bits that are automatically selected, that’s going to be working and it’s going to be sounding attractive to people who just want that, which is the majority. I think AI is going to be playing a major role. And I think the responsibility for our societies is to always make it very transparent if AI is involved so that the listeners have at least the choice if they want to listen to machines or not. I’m assuming that most people won’t mind listening to machines as long as it’s sexy to their tastes.

Ruben Jonas Schnell. Photo: Aron Urb

YR:  Well, it’s funny because we are at the advent of listening to machines. We’ve got the RadioGPT, which is another form of ChatGPT. It’s basically replacing human DJs with AI ones that actually speak without the “ums” and the “uhs” and can tailor playlists depending on your musical tastes. I don’t know what your thoughts might be on that? Or does this have you quaking in your boots or anything like that?

AK: I think that what strikes me about the idea of RadioGPT is that it sees people, radio staff, as a problem to be solved. And in fact, when you come from community radio, which is what we are, people are the reason for being there. So there’s no problem there to solve. And I think that that’s why it doesn’t have me quaking in my boots, I have to say. I think it’s the connection with other people – other makers locally and from other scenes – what keeps people coming to you and what has driven the flourishing of online community radio in the past few years.

CL: I find it really exciting to witness this kind of development that we’re seeing right now. And I find it very fascinating that we are already able to almost recreate the simple form of classic radio with machines. But my question would always be, will it make it better? Is there value beyond recreating the status quo? What do we want from it? At this point, I don’t have any answers yet that satisfy me. 

RJS: Well, like I said, I’m not concerned about us. When you’re trying to do an individual program, I think it’s not so easily copied by machines. But for mainstream radio, I think as a program director, generally if there’s a chance to save money on salaries and you’re employing a machine that’s doing the same stupid jokes, then it’s easier to apply the machine. Also for the music selection, if it’s always the same then that can be done by machines and it will.

HR:  I’m not concerned, and that’s why I’m really concerned about. The story of my radio station is that it was created more than five decades ago in 1971. We now live in a crazy period because it’s the most successful period that we’ve had. The more we have algorithms, the more we have technology, the more people are afraid about ChatGPT and the more listeners we get because of that. It’s basically a refuge. So a few years ago we claimed to go back to the basics of the old cliches of the radio. Always live and direct, nothing automatic. I think we are one of the last national stations with no musical genre format. This morning we can play Arcade Fire and after that Claude Debussy, and then Rosalia mixing Charlie Parker and Dizzy Gillespie. It has to be mostly about the sound, the mood. The modernity comes thanks to the music, but you can have modernity also in the timeless recordings. I heard an interview we had yesterday with Daft Punk for the anniversary of one of their legendary albums. They were obsessed with timeless sound and they thought it important not to confuse what is fashionable. That’s what we do.  We have the smallest overall team of Radio France. Less than 20 along with the speakers, the programmers and the administrative staff. But we have the biggest programming team. There are eight full-timers and each day they program track by track. You have to work like a DJ, but with all moods. I think you would find us totally crazy if you see how some of them work. Last time I saw one of the programmers in my team, he wasn’t inspired and so he took his guitar, started playing and said, “Okay, okay, that made me think about that track which I will play”. That’s human, manual, and totally crazy. The more ChatGPT will become popular, the more of a refuge we will become.

YN: Christophe and Ash, you two in particular mentioned the aspect of content in radio. 

CL:  When we were talking before this panel, we all agreed that we had a really hard time reaching people under 30 years of age and in significant numbers. To me, this is a warning sign. For our station, it was really essential to become creative because our network couldn’t reach young people. We’re a public radio station and a public network, and we’re supposed to reach all kinds of people in our society, and we couldn’t reach young people. Our job was especially to reach people from 16 to 25. With our radio, we just couldn’t do that. We started wondering if young people don’t need radio in classic form anymore. Because for the social aspects they have their social platforms, and for music they have streaming services, algorithms and everything. Do we still create value that plays a meaningful role in their lives? 

So we started to think of new ways of reaching young people. And of course, we started making all the classic mistakes. We had a big radio team with the music editorial staff and we had social media experts. And so everything we did, all the value we created in, for example, music journalism, we were begging the social team to play it out on different platforms. What happened was, of course, it led to really unsuccessful on-demand music content. We were unsuccessful in reaching young people through radio. And then we were unsuccessful in reaching young people on social platforms, because we said “We have an interview with Portugal. The Man, and they are super famous. You need to post this on YouTube, on Facebook, on Instagram, on the website, everywhere”. And it just didn’t work and it created big tension between the social media team and the radio team, because the social media team wanted to be successful on all these platforms and they didn’t want to post something that created 39 likes, which sometimes happened when we posted an interview with Arcade Fire or whomever. They said, “Well we want to play out what is successful on these platforms”. 

So we started to analyze what young people are looking for on social media platforms. We stopped working with Facebook because there were no young people there, just like in our radio station. We didn’t find young audiences there anymore, but we did find that young people are actually looking for music content on YouTube. We thought, okay, we need a team that creates music journalism on YouTube because people are looking for it there. We actually took money and we took people from our music editorial radio staff, took them out of radio and built little teams that created a YouTube channel analyzing musical phenomenons on YouTube and a podcast about German rap. These were people who were working in radio before, having their evening shows. We stopped doing these shows because nobody listened to them. Now we publish weekly music journalism on YouTube and on podcasts, and we reach hundreds of thousands of young people because we see the analytics on YouTube. We see who we’re reaching. And so we managed to create successful on-demand music content. You might think, okay, but radio really suffers from this. Yes, they do have less resources. We have a smaller team now. But the amazing research that these people are doing for their YouTube channels and podcasts creates amazing topics that we bring back into radio.

The hosts of these YouTube channels or podcast go back into radio and say “This is a story we’ve been following all week. It’s really interesting because of this rapper and what you might not know about this rapper is blah, blah, blah…”. They bring amazing content that they wouldn’t have done in their radio shows back to radio. I’m just trying to disrupt our thinking a little bit, and it might not be the path that is right for everybody but I think our network would have pulled the plug on our radio because we weren’t doing our job, which was to reach 16 to 25-year-olds. Now we have proved that we can actually reach 16 to 25-year-olds and we can play a meaningful role in their lives every week. But we had to reorganize everything. It was changing our thinking from being a radio station that also does a YouTube channel and that also has an Instagram account to being a content network for young people. We just produce different things for young people. So this was the path that we took to find an answer to the question of what is the future of radio.

Cristoph Lindemann. Photo: Aron Urb

YN: I love what you just said there at the very end – content network. Do you think people should change their ways of thinking? Not radio stations anymore, but rather content networks? Is this the way forward for us?

CL: Well, from an audience perspective, the content network thought is not that important. We do have our channels that we call our shopping windows where we just explain “Look, this is all the things we do, you might like this or that”. But from our team perspective, it was really important and we struggled with it. I was the Head of Music of the whole network, which was like a vague responsibility for everything musically created, like a podcast and this and that. Nowadays every product – the YouTube channel, the podcast, the radio station – has its their own strategists who are doing analytics and telling, “Well, the audience wants more of this. Be a bit more deep in your conversations, it works really well” etc. After two years of struggling to let go of our old titles and of our old way of thinking has been freeing for us.  

Radio is one thing we still do, but it’s one thing among many.

RJS: Just one thought. I think it’s very important to take into consideration what the audience wants. But at the same time, I think the quality of our radio means not always asking that, but just giving them what we find valuable. I have one particular question though sbout what you just said, which I find quite interesting. Are the YouTube channels similar to what radio showd about the same topic would have been like, or is it a totally different editorial approach?

CL: Totally different. The podcast for two years was very similar to one of our evening radio shows where they just talk about music. But they were talking about one topic in-depth. We didn’t reach enough people. I think we reached more than with our radio shows, but we didn’t reach enough. So we pulled the plug there and we’re trying to develop music journalism on TikTok right now. But for example, the YouTube channel, it’s just a 7 – 10 minute format every week where a host is explaining a phenomenon.

RJS: And do you incorporate that into the radio programme?

CL: Yes. We take the interviews that we did for that little YouTube video and we broadcast them on the radio. The host will come into the radio show and explain everything.

AK: I’d like to offer a counter example to that, coming from a completely different realm. The radio station at WORM began 10 – 15 years ago. There was already a radio project at WORM, but that disappeared in its own time. The radio project began as a hobby project of one person just before the first lockdowns in the Netherlands. It was during those times that somebody set up a little broadcast station in the office with hardly any gear whatsoever, no care at all about who was listening or where. And then suddenly everyone else in the building would pay attention. So suddenly people that are from the bar are like “Actually yeah, I want to DJ, I want to do a show”. People from the film work plants are saying things like “Oh yeah, we can get involved in this”. The theatre, the art gallery, everybody else in the building were saying “What we do we could do as a radio show”. And it really became a thing that people could say, “Well, we used to put on a club night about Colombian Cinema. We’re going to turn that into a weekly two-hour long radio show”. Two years later they’re still doing that every week because they’ve found value in that format. It’s driven by the people that make the shows rather than what we assume people want to listen to. And so for me, it always comes back to the value for the maker, the value for the artist who’s showing up to make the show. So that’s a very different approach. 

I don’t see AI wiping out any of our activities as radio makers anytime soon. Because from the economic aspect it is indeed cheaper and easier to have a machine make a show. But if your radio hosts aren’t getting paid in the first place, you’re not saving anything. So they’re going to keep showing up to do what they want to do. Where I can see AI being really interesting for community online radio is in content. Stations like Radio WORM are made and listened to by people with a commitment to experimentation in art and media and music. And so those people are interested in what AI is doing in a way that they want to question it and be critical of it and think about its wider implications socially, ethically, and politically. I think that’s where we are going to see it, much as we will see it in art galleries with contemporary artists, using it as a tool to question things. I’m expecting that that will come along at the same time that RadioGPT is becoming more widely known.

Ash Kilmartin. Photo: Aron Urb

RJS: Actually, ByteFM is a station where you have 100 hosts choosing music. Everything is hand-picked, individual, original and everybody who’s working there loves what they do and can make their own decisions.  I criticized or warned against AI, but as soon as the technical possibilities are accessible to us, I would love to put an AI curated show on the channel for half an hour or an hour a day and then try to set different specifics each day, maybe on the musical colours or the emotional output of the AI. I think that would be very interesting. I think even our audience who is interested in uniquely curated music programmes would appreciate that as well. But again, I think the very important factor is we have to always make it very transparent what is being broadcast. Is it a human or is it a machine that people are hearing?

TD:  Yeah, if you have listened to RadioGPT, it’s pretty good. So I would be worried a little bit about this area, although I don’t think we’re right there yet. But I’m into technology personally. I think it’s a good thing, but I think we have to have specific rules for it. 

As an example, I actually put my own name and also Radio Helsinki into ChatGPT. And all the information was wrong. I’m not one that usually Googles her own name or something like that, but for this purpose, I wanted to see how accurate it was. 12 years ago I lived in the States and I was a CEO at a company of healthcare. I’ve now been in Radio Helsinki for 7 years and the GPT was mixing these two jobs based on something that I have on LinkedIn. So it was total bull crap. It sounded really good, but it had nothing to do with the radio.

AK: On the RadioGPT website, saying how great RadioGPT can be for you and all of the benefits that it can offer you, it did say there can be some errors and that you should listen to every show before you broadcast it. So there’s still a job for someone.

YN:  Yeah, it’s still in its infancy, I suppose.

CL: When you say you listened to RadioGPT and it was very good and that’s why you’re worried… Why? All of us have probably in one form or another broadcast pre-recorded shows or reruns and we always have to make a responsible decisions. What do we do if a pre-recorded show is running and something major is happening in the world or something? Will we cut the pre-recorded show? RadioGPT might even one day be smart enough to read news about what’s happening right now. But if you think that we still need somebody making responsible decisions, then why are you worried if RadioGPT is good?

TD: Yeah, of course, we have format freedom in everything. I don’t know what any DJ is going to say online while they’re doing their programme or what music they’re playing. Actually, I’m more worried that they’re not saying anything because it’s risky to have opinions nowadays. Formatted commercial stations should be worried about this more than maybe our kinds of radio stations or platforms just because programming for them is already quite narrow.

RJS: This whole AI and ChatGPT technology is so brand new. I think all those flaws that you detected, they won’t happen in a couple of months from now. I think it is going to be a serious source for news and for everything and everybody. Not everybody, but a lot of people will use it and use it as a reliable source. I think that’s quite threatening. RadioGPT, it already incorporates news. They are quick. They’re possibly quicker than you can be if you have an editorial staff that has to work on that news. They’re connected to social media whenever something is being hyped. And that’s the decision-making fact, the attraction of the news. They incorporate it into their programme, which means that the programme will end up being successful and attractive to the masses. I think it’s a very dangerous route we’re taking. I don’t think it concerns us as much, but for society it’s extremely dangerous.

TD: I’m more worried that those algorithms and everything AI are trained in very certain things. I think it becomes a certain info war. How well are different opinions represented in the data? I would be worried about that. I think we’re already going toward the idea that one-way thinking is the right thing. And if you’re outside of that, you must be some crazy conspiracy theorist. Like I said earlier, that is something I’ve said to some of the people that do programmes for us. That I want to be in the courtroom because you said something crazy. But it’s so hard to even find people having different opinions right now because if you have any digital evidence about anything you’ve said and you’ve changed your opinion about it later in life, that’s going to be there. I don’t know how it’s helping the freedom of speech idea that is dear to me.

I’m more worried about that, really. it’s not just about the radio, but it’s all about the information and how AI is trained.

YN: I want to talk about the positives now for a change. I suppose TikTok also comes into the equation where people can make discoveries of older groups and so on – as you found out recently with your daughter, Herve, who I believe discovered The Cramps.

HR: Yes, from Netflix series. I asked her to repeat it three times. Because my first idea was that it’s impossible that she talked about the band I knew. Yes, thanks to TikTok and thanks to the series…  In fact, that’s what we do for our target groups. We don’t try to be fashionable, as I already said, and we don’t try to keep them, saying “you have your universe, you have your music”. It’s the contrary. You can listen to so many things and sometimes in your daily life, you can have something different. Not too far from what you like, but it’s different. You can listen to SkyRock, which is mainly a hip-hop/rap station and totally different from FIP., but it’s another complementary proposition. 

That’s interesting how they work, especially for the series. It’s really interesting to look at the musical approach for the  target users. .

YN: You work for a very traditional station like Radio France. How do they go about being innovative and keeping pushing forward, for example? What are the challenges for such a traditional platform?

HR: In the way we distribute the programmes? Yes, we do it in a traditional way, but it’s also amazing that Radio France works as a group for podcasts, for the replays, for such content. We don’t try to reinvent something, create new platforms etc. It’s simple. You have the apps, the websites and the social networks. Social networks are another thing as they’re more aimed at communication than consumption, but it works.

YN: I want to go back again to the question of TikTok and Netflix and how people discover new music through the algorithms. The same could apply to Mixcloud, Spotify, and SoundCloud. We touched on this a little bit. How do you think that can help in terms of how people can discover new music?

CL: It plays a really huge role now. I DJ, and I do it sometimes in clubs where everybody’s between 18 and 22. I study “viral on TikTok” playlists all the time. You can play anything. One of the first phenomena like that was Toto’s “Africa”. It became a huge song in really young clubs maybe seven years ago. Then you had Fleetwood Mac, and now you have a guy who’s dancing on TikTok and he’s a really great dancer. He’s a super cool hipster and he’s dancing to Boney M’s “Rasputin”. You can play “Rasputin” now and people go crazy in the club. It really changed the way young people discover music. They don’t care and don’t know if the song is from the ’70s. They just love it. I find this really interesting, and I love it because it opened up the range so much.

AK: I would say that platforms like Mixcloud and SoundCloud, where you can upload mixes after they’ve been broadcast, are super important because those are the places that point directly to the artist’s Bandcamp page. Those are the places where you might be listening to NTS, Cashmere, Kiosk or Radio WORM. It’s only when you see that someone’s uploaded the tracklist afterwards that you can then go and find that artist. That is definitely discovery and it points you to where the live gigs are, where you can pay for the music from them. That is super important. It also shows you who’s listening to what. If you listen to  internet radio, you can see from the playlist who’s listening to what else. You can see who’s touring, who’s been dropping in and sleeping on such and such a couch in that city and then going to another couch in another city. I think it’s not only for people completely outside of that scene. You can really chart how music moves that way too.

YN: How much attention do you pay to these platforms working for a radio station?

AK: I actually have a radio maker who’s trying to convince me to get our stuff off Mixcloud because they’ve invented their own peer-to-peer system that’s non-commercialized and decentralized. That explains where we’re at.

RJS: I don’t really [pay attention – TMW] at this point, but maybe I should change that because I really like what you said about the tracks that are hip on TikTok. In a way that actually democratizes music listening, because the kids don’t associate Boney M with the German charts of the 70s that are negatively associated for me, but they listen to the track and they think it’s funky. I would love TikTok viral selection to be on Byte FM, curated by an advertiser who puts things into perspective. I have been looking for a way to connect ByteFM to TikTok, but all my suggestions don’t pass my team because they think it’s such a different market. But of course, I would love a TikTok show that is hopefully hosted by somebody who is on TikTok and who is credible for the TikTok crowd, but at the same time maybe a new voice on our channel.

AK:  It seems like that could be the first show on Byte FM that’s actually curated by AI. It’s an easy step.

RJS: Ha, there you go. There you go. Starting next week, maybe.

AK: I’ll be tuning in.

CL: For us, it was a really difficult question if we should, for example, incorporate Kate Bush’s “Running Up That Hill” into our programme if our target audience is 16 to 25. And I said yes, because the song is massively popular with 16 to 25 year olds right now because of Stranger Things. But some of the old music editorial team said, “We can’t do that, it’s not our format”. When we were restructuring everything, we let other people do the radio programming and they are all young. Now Kate Bush’s song is playing on the radio station. The young people had no issue saying, “Of course we’re going to play the song. It’s great”. Our programme also became much more open and diverse. Herve knows all about that, a diverse radio program. I have to say this again because he’s [Herve Riesen – TMW] very modest, but listen to FIP, if you haven’t. That music programming is art. It will do something to you.

HR:  It’s very interesting because all our listeners always give the feedback like “Your station is amazing, we discover so many things”. But when they say that they discover new music, then new music for a kid could be Charlie Parker. I think many things have changed thanks to free access to music. There are some optimistic things coming out from the disappearance of differences between underground and mainstream, between old school and new school. The mentality has changed. We are now close to Africa. We talked this morning with you [Youssif Nur – TMW] because we both used to work a lot in Africa. I remember it was 20 years ago in Mali, in Bamako. I worked at La Chaine 2 radio station and it had a collective programme that was played over Mali, France and Switzerland, so we had to share the music programming. Sometimes colleagues from Mali asked me to play an old-fashioned track and a legendary French singer said, “No, no, no. It’s not possible because it’s the music of my father”. They asked me, “So what’s the problem?”. I said “No, it’s not the music of my father, it’s the music of my grandfather”. They replied that that’s even better. I said, “Yes, it’s wonderful”. Well, now we can do that, since we also have that mentality. 

Talking especially about TikTok, the problems are not really ours. It’s not so much about radio. Radio can use TikTok to promote its programme. But for the music industry it can be a problem. As I said this morning, last week a journalist asked me, not about radio, but about the music industry, “Are you afraid that with ChatGPT we can create in five minutes the new Rosalia, the new Beatles?”. Maybe I’m wrong, but I said, “Not really, because that takes a longer time”. In my country, when you have a huge new phenomenon, a huge new star, the major company signs them immediately and it never works out.

Hervé Riesen. Photo: Aron Urb

YN: What are your thoughts on the ability of AI to recognize songs for artists to be paid their royalties.

TD:  I have comments on this, although not conclusions. I think this is something that we definitely think about, since the company that Radio Helsinki is owned by is the Musician’s Union in Finland. So in anything that we do, we are always looking out for the musicians – whether they’re getting paid and if recognized for their work and stuff. I don’t think that technology has been the problem so far in getting paid equally for the work as a musician. We’ve had technology that could have been used in any of those streaming places in order to actually pay the people that were listened to. Technology has been around a long time, and so it’s not about the technology. I don’t know if this is a new chance for that to be corrected, so that people would actually get paid if they were listened to. And not just the 10 biggest ones. I think there’s an opportunity here, but it’s never been about lacking technology on those things. It’s about the desire of the ones that have the power –are  they going to pay it out or not. So if we want to look at the positive things, we may think this is a new chance to correct the ways.

But it’s always been there. It’s been there for decades.

CL: In public radio, we wasted millions of euros with people writing down that a song was used in this TV show for 8 seconds between the 35th and 36th minute. And they were all doing it by hand. For years now we have developed a system so that all of this works automatically. Of course, with every new platform we’re playing, we’re still behind. 

We had YouTube and suddenly it was a grey area where the music royalty rights management didn’t work as well, but now it does. And so it’s really amazing to have artificial intelligence to handle all this. Yes, it’s a big help. All these people can do meaningful things now instead of writing down little lists.

YN: Tanja mentioned the funding briefly.  Do your respective governments or the EU help in a lot of ways? Are they open or receptive to funding the newer technologies or your shows, for example? Or do you have to look for sponsorship? Because we mentioned the really random sponsors you might have seen out and about.

TD:  I can surely comment on that. We’re quite a diverse group of different radios, even though we have this in common that we share a wide cultural and musical take. But like I was mentioning earlier, Radio Helsinki is a commercial station that’s doing all these things that publicly owned or government owned stations would do. They’d have funding, but for anything that we do at Radio Helsinki, the money has to come from somewhere. Our programmes are sponsored by somebody. So if we look at a programme we might start with, we  have to look for sponsors at the same time. The point is that for digital things like AI, these changes will take money and it has to come from somewhere. I’m on the board of commercial stations in Finland and it’s been interesting to see that they’re not really pushing these new things… Why would they make a change if they don’t have to make it? 

In general, I think at commercial stations they write down things like, “I have listened to this radio station for 30 minutes this day” and whatever. Who does that? Do you actually get credible information from that stuff? No, you don’t. For example, you can’t own a company phone to answer these things. You actually have to have your phone registered to yourself and not through the company. The funding issue is definitely something that is difficult. We’ve done this from 2016. We were the first ones to do our own apps and things, but it takes a lot of money that has to come from somewhere and it’s not like somebody has to support those things. You might have a cool new artist and they’d want to do programming, but the people who are sponsoring are lagging behind. So it’s sometimes super hard to get something for a new artist because they’re known only in certain circles. They’re known and they might be even on TV already, but the companies that are paying don’t know about them yet. 

Are we doing stuff that our listeners want? Are we educating them? Are we trying to bring in new things? Those are all considerations and different motives for us to do things. We have to get the money and then we of course want to do something that our listeners want. But we also want to make sure that there’s always new stuff that they can find through us. 

Tanja Douglass. Photo: Aron Urb

YN:  I want to know what you think of artist-generated songs via AI. Because in the last week I’ve heard AI-generated Oasis’ songs. It’s absolutely true. There were a couple of Oasis’ fans who were so sick of waiting for them to reform that they formed a band called AI-CIS, and made songs with Liam Gallagher’s voice. And he responded on Twitter saying that he loved them. I believe Nick Cave did something similar? You were telling me about this, Herve.

HR: What I saw is that he received a song from a fan. A fan sent him a Nick Cave song made by ChatGPT and asked him what he thinks. Nick said “Yeah, it’s okay. It’s not so bad, I recognize myself. But that’s exactly what I don’t want to do anymore. I’m still not able to tell you what I want to do, but I have a new approach. I have new influences these days and let’s all be surprised together.”

YN: Would you ever play it on Radio France?

HR:  No. The wrong track? Yeah. No, no.

CL: We would. From a music journalism point of view, it’s really interesting what is happening right now. I hear that a lot of people in writing, for example, use it to break their writer’s block. They get a first draft and then they steer away from the AI draft and do their own thing. Maybe if Nick Cave is confused right now which way to go, he could try “Give me Nick Cave in a Radiohead kind of way or in a Kate Bush way and let’s see what comes out of this”. Maybe it gives him new ideas to be creative himself. I see lots of potential.

HR:  But if we ask ChatGPT, “Do you think an artist like Nick Cave will go to a big ceremony with King Charles?”, ChatGPT would reply “No, never.” But he did.

YN: Just very quickly, can you talk about the Swiss radio show?

HR: I saw that two or three weeks ago the Swiss public service in the French part organised a day made by ChatGPT. They told their listeners that today is ChatGPT day. They put a short footage online where we saw the hosts listening to themselves and being really surprised, saying, “It’s me, but I don’t remember saying that. No, I never said that”. It was really interesting and really close to reality. But the public was afraid and said, “It’s so close to the real channel. Please never do that again. We want reality. We want you and we want you to verify it’s you”. It was clever doing that.

TD: Great marketing. I think we’ll do that too, so they’ll want us.

YN: It reminds me of when back in the 1930s the BBC put on a narration of “The War of the Worlds” (actually CBS Radio Networks’ episode of “The Mercury Theatre on the Air”, narrated by Orson Welles  TMW). And because no one had ever heard anything like this, people actually thought it was real and the radio got complaints and had to tell afterwards that it wasn’t. So that could be the first incident of AI in radio, perhaps. So in closing, do you have any final remarks at all? Because time’s up.

AK: I mean this entirely positively. If the Radio GPT revolution comes, we’re going to need to find a different name for what we do.

YN: Finally, can we turn a radio station into an NFT?

AK: Why would we?

YN: I have my answer. Thanks, everyone.

The panel is re-watchable for the holders of PRO and DigiPRO passes until TMW 2024 via the TMW PRO platform. PRO passes for TMW 2024 are on sale at the TMW webshop.