Making sense of the Ed Sheeran situation; The million-pound fake vinyl factory; Will there be a new Oasis album?; AI pioneer speaks to TPOM; and The Ace Of Shatner!

23:16Stu:It was brought to him, wasn't it? There's no sense here. There's been no sense suggestion here that, well, he didn't put Macklemore on, didn't he? Wait, wait there, wait to hear what I'm saying. There's been no suggestion that Ed Sheeran heard those speeches on the stage and kicked Macklemore off the tour. There's no suggestion that he had any problem with Macklemore saying that on the stage. No. It's only when it was brought to him as someone saying, we will not have you in our stadium if you don't kick him off the tour. Yeah. And that's the thing. He was giving Macklemore that stage and allowing him to say what he wanted.

23:52Stu:yeah and that's what's got missed i think it wasn't ed shoo and kicked him off the tiff of doing it it was no no of course i mean like you say he he he booked macklemore knowing that this he was doing this is what macklemore does now you know it's not like a surprise you know so he booked it obviously he must be okay with that i guess what people are saying is when you know you're sometimes you came yeah you know what life's like like you might not want to um like you know to put it bluntly sometimes you've got to choose between eating two shit sandwiches and one of smaller than the other, right?

24:22You've got to eat one of them. And in this case, he tried not to eat either sandwich. And everybody said, if you're in an impossible position, you've got to do something. And it sort of felt like he was trying to go, oh, look, this doesn't do with me. And I think that rubbed people up the wrong way.

24:44Stu:That's how I read it, anyway. Well, that's it. And I suppose the only two things I thought was, one was, I think, I mentioned how Robert Kraft is a bit more nuanced than the evil villain. Macklemore is not an absolutely 100 % easy hero here because he has his own history where he once dressed up in a costume that was very, very clearly a two-stage. Who hasn't done that in the past? On stage. And was criticised for it and apologised. But then that gives you maybe less of the moral high ground to say, I'm not being anti-Semitism now and you should trust me on that. because I think what people have said is, actually, you've kind of maybe not got as much trust with us as Jewish people as you could have if you hadn't dressed up as a stereotype.

25:23Stu:But the other thing I think is, I wonder what the alternative scenario is here. Imagine if this happened, Ed Sheeran gets rung up, or he's rotas, Biden and said, we're not going to have you in here unless you kick Macklemore off. If he'd said no, Macklemore stays on the bill or we don't play your venues, what do you think would have happened? Would they have backed down? Or not? Because that's the alternative world that we're kind of wondering here, is what if Ed Sheeran had said, no, I am going to - I think that's what people wanted. They wanted to find out what would happen when someone puts their principles before everything else.

26:00Stu:And I wonder would they have backed down? Because for them, they want Ed Sheeran to play their stadium. It's not like they're the ones with leverage here and he has none. he's bringing 80 ,000 people into their very lucrative business. I feel like we're going to be talking about this. We can never know what would have happened. But that's what I think people are disappointed about. They say people may think if he had stood up, maybe the tour would be going on with Macklemore and we'd be able to discuss what was being said. I feel like we're going to be talking about this for a while because I don't think, by the way, that Ed Sheeran has exploded his career.

26:36But I think this is going to follow him around for a very long time. And I think my feeling is he'll probably change as a result of it. We'll probably see a different Ed Sheeran in the way he communicates as a result of it. And, you know, maybe that's the best outcome, you know, over time. Instead of expecting him to be exactly right off the bat, you know, over six months, two years, he changes and becomes different. And surely that's good, you know, that's an artist growth thing.

27:00Stu:Well, think of us, think of what all of us are learning about this. Like most, you know, most of us don't have a good history of Israel, Palestine, and all the things. we're all learning and no one is saying to us make a decision now but we know stew as people publishing news stories and we have had increasingly had to write things about this you know israel and palestine because of artists talking about it that it's an incredibly difficult thing to do because if you get something that is as you know even if it's factually or as far as you can tell factually correct you're still going to get bombarded with with complaints from either side because it doesn't align with their worldview.

27:38And that's just the world we live in now. It's not good, but there you go. Now, one artist that I don't think will be espousing views on Palestine from the stage, but I'm prepared to be surprised, is Oasis. And news that will surprise absolutely no one who has been online this year. Oasis is going to play some more shows next year, including, as far as my last check, about 5 ,000 shows at Manchester City Stadium in Manchester. but those tickets won't be going on general sale um what does that mean and why not and what's going on with oasis tickets for next year well basically they're saying no no public pre-sales

28:21Stu:no general on sales the only way to get tickets is to pre-register on the band's website um and you'll get if you're lucky a code for a about a specific ballot buying window they'll say right this is your window to go in and buy tickets you've got this code and it's actually i've signed up for this to test it out yeah and maybe to get tickets so um get the question right yes i've got the question i think i've got the question right but um uh so if i am successful it's like what is that like a it's like a they pull people out random they go right here's a code for you you can go and buy four tickets now between this day and this day and you don't have to queue you will just get them if you log in and use this code well Well, it's not a guarantee of getting tickets.

29:04Stu:It's a guarantee of getting into that window. You may get in there and they're all sold out. You may get in there and all the cheap ones are sold out. So it's not a guarantee of tickets. It's just your pass into the pre-sale from the band. We don't really know how it will be chosen. It could be a random ballot, but it may be based on... I mean, they're working with a company called OpenStage, a tech company we've talked about before. OpenStage has technology that enables artists and their teams to kind of really go into quite detail about fans. Like, well, which fan has been on my main list for this long?

29:37Stu:Or from a link to my store? So it could be people who bought tickets before perhaps, will be offered tickets first. Well, or people who haven't been bought tickets, which is another way of getting people to concerts. And you might say, well, I want people who didn't get a chance to see me before. I think it's different cities to the last run, so maybe that's covered anyway. So we don't quite know. And of course, they can't say, this is what's going to get you the best chance of getting in, because then everyone would go and do it. We'll quickly go and buy merch from a link from there. So there's this happening, so essentially your registration.

30:08Stu:They're also fixing the ticket prices, so no dynamic pricing. Do we know how much they are yet? Between$122 and$541, I believe. Obviously, the$541 is probably golden circle VIP business. I would want Liam to feed me chips directly into my mouth for that kind of money, to be honest. But this is two things. One, this is trying to repair something. Because the last tour, there was a couple of days of confusion and anger from fans when they got into the pre-sale and the prices seemed to have suddenly surged. So I think in the end it was not direct pricing or something else. So I think from the start here saying, right, this is how it's going to work and we're going to make a fine.

30:50Stu:And two, this is maybe a thing we're going to see more, is we're going to not just selling to anyone. We want to get you to register. Now one, if we get you to register, you're on our mailing list. This is a customer acquisition strategy, partly. You want to come to the concert, you have to sign up for the mailing list, give us your details. Which is kind of, I'm not saying that's a weird conspiracy thing, it's just what the music industry does. You get fans' details. But also it's saying, right, this idea of the first come first serve queue, it's kind of broken touts can still exploit it in all kinds of ways so this idea of like a closed garden ballot may become it's not the default yet open stadium work with quite a few big artists but it's by no means every single tour yet but i think it's maybe the way we go in future you'll to get tickets for tour you will register with the band and then they'll send you a code and that's how you'll get you and get your tickets and it seems it seems smart enough doesn't it isn't it because ticket buying for these giant shows has become a truly miserable experience, right?

31:52No one enjoys it. Everyone hates it. Okay, maybe there's some truth in the fact that, like, you know, the kind of the hassle of it is part of the process, and it sort of makes you feel happier when you get them. But I think in general, if you said to someone, hey, do you want to spend, you know, five hours watching a huge number go down in the queue in front of you before you get offered the chance to get tickets? You know, like, no one wants to choose that, right? So anything that might come up with a better system is clearly going to be better for everybody and better goodwill for the artist.

32:20And anything that stops, I mean, frankly, anything that stops ticket touts from secondary ticketing and selling these tickets at inflated prices is a good thing, isn't it? But we'll see how it goes because, you know, I mean, we presume that demand for these tickets is going to be enormous. I was talking to someone, and this is a little bit of a segue to the next part of the podcast. I was at the Ripper Barn conference in Hamburg last week. And I was chatting to someone about Oasis and the tickets and everything like that. And we were saying, they've got a lot of dates on this tour. They're doing, is it 12 nights at Manchester City Stadium, something like that?

32:58That's a lot of dates, considering that just two years previously, they've toured basically all the hits. The shows were sort of two and a half hours long. They didn't scrimp on a show. They put all the hits in. So do people really want to go back and pay these huge prices to stand in a stadium and watch Oasis sing the same 20 songs again which sort of leads us to sort of conclude that there's probably a new album coming out right because that has to be the driver for so many tickets.

33:27Stu:Do you think they want a new album from Oasis though? I don't know. I think Oasis fans would be okay with that. I mean look I mean Is it Rolling Stones play where it's like there's a tour so here's an album. Yeah I think so. It's their best album since the last few at once. A best hum for a decade, too. I mean, I find it very hard to believe that Noel Gallagher's gone for two years without writing any songs, and he's not doing High Flying Birds anymore. So, for me, it's just a, you know, a bit of pattern recognition here. But we'll see. I mean, but think about it. I think it would be maybe a new compilation.

34:00Stu:Another compilation. All the tracks with Tony on drums. I would buy that one. That's a compilation. I feel like, hang on, Black Crowes Klaxon, Black Crowes Klaxon. Oh, go on, yeah, please. They've been making new albums since they got back together. But that seems like a genuine, we want to go in the studio together. We don't just want to go make money on a stage. We actually love making music. And genuinely, I'm not buying the Kool-Aid here. They do seem very, very like we miss making music together. Yeah. Do you get that sense with Oasis at all? They've missed being in the studio? I think they want to play.

34:31Stu:I think they're enjoying playing on a stage. Yeah. I think they're primarily a live band. I'm not saying it's a cynical thing. I think they made those gigs really good last year. But I don't know if they have the burning ambition to put new music out or not. I think they will do. I think they are primarily a live band. They want to go and play live, right? And to do that, I think, you know, both of them, both the brothers outside of Oasis have stuck to a very rigid formula of write an album, tour an album, right? They've done that for over a decade. That's what they did in Oasis. I just think it's, I think obviously they can do it, the comeback, they can do it without releasing an album.

35:08I think it's a no-brainer there's an album coming out. To be fair, there weren't many points in a set

35:12Stu:where you can go to the bar. It's all killer when I saw them. So maybe they need three new songs so they won't go for We and the Bar. I think they're thinking of their slightly older fan base and their weaker bladders, which is very good of Nolan Lee. Thank you, guys. Now, as I mentioned, I went to the Raper Barn Conference and Showcase Festival in Hamburg last week. And if you've never been, it's a really great, it's a bit like sort of the Great Escape in the UK and there's lots of showcase festivals, showcase gigs from all over Europe and the world. And there's a big conference that runs alongside it with all the bigwigs in the music industry.

35:49One of the talks, so I was there, and there was a partner conference running inside Raper Blonde Conference called Music Frontiers. And that was a bit more sort of like cutting edge music stuff. And speaking at this conference was a man called Francoise Pachet, who helped pioneer generative AI music. Now, before you run a race screaming and closing the podcast player, I interviewed him afterwards because, and we'll play you this interview in a second, because it was kind of fascinating. First of all, all the AI music that you listeners out there have been so vocal about to us by email and saying that you don't like it, and if you talk to artists, they don't like it.

36:29It's kind of his fault in a way. He was part of a pioneering team working, paid for by Sony's sort of future engineering lab. So not by Sony Records, but by Sony, the major company that sits above it that makes Playstations and stuff. And they funded him for about 10 years, which sounds like a dream job, really, to sort of turn up every day in a laboratory and sort of try and collaborate with jazz musicians and try and make a computer play music. And he is a musician himself, Francoise, and he played various examples during his talk of stuff he'd done throughout his career. And there was a really fascinating example, actually, from maybe about 15 years ago, a computer program he built where it was connected to a piano and he got in a jazz pianist to play something.

37:18So this jazz pianist would play like, ding, do ding, do ding, play a little riff. And then the computer, you'd see the keys moving on the piano, the computer would try and play something that wasn't the same, but sort of in the same style back to the pianist. and the pianist would take over again. And it was sort of like, you know, having a sort of playing along with each other. And it was really fascinating to watch and how it works. Anyway, the very, the reason I interviewed him was this landmark thing he did, we reported on this, you know, I've talked about it a lot 10 years ago, the first really recognizable AI generated song, if you like, which was called Daddy's Car.

37:53And it must be noted that they trained the AI on the compositions, the songwriting, but not the recordings. So they took a bunch of Beatles songs, which are sort of, well, as he explains, sort of owned by Sony so they could get away with it. But the compositions were owned by Sony at that time. And they trained the AI on it and it then generated a song, which then they got a bunch of human musicians to go and record from the sheet music that the computer generated. and he also then later on when he went to work for spotify um did the same thing but with pete seager the famous us um sort of folk artist with the permission of pete seager's estate and they generated a whole bunch of new new quote-unquote pete seager songs um i think that what's interesting about him was and the reason i recorded a quick interview with him for this podcast was i think he's a lot more aligned with the listeners than you might expect he's very pro artist and very interested in what music is and what it can be.

38:55So we're going to throw over to me now in backstage somewhere, as a band tunes up in the background, talking to Francois Pachy. So we'll come back to you in a minute.

39:07Stu:So I would sort of like to go back. You were talking about Daddy's car and that sort of very exciting period then, and the work you did to create it. And you just spoke here at Music Frontiers on stage about how you train the AI on a catalogue of Beatles songs and you could do that because the project was funded by Sony and they owned the compositions to that. I'm interested in the sense of there was still controversy around releasing the song afterwards even though it seemed to be sort of you were working within the rules. Can you tell us about that? Yes. So when I was there, I had also, I want to say, I was working at Sony, but I also had the funding from the ERC, you know, the European Commission, a big, big project.

39:57And I decided to build something I called LSD, you know, which is a lead sheet database, which I did by myself. I paid people to do this. I didn't ask Sony. I didn't ask anyone, you know. I paid people. I gave them, like, songbooks. And I did the Beatles and Elton John and Jazz, a real book, and Gershwin, whatever. and I asked them please enter all these songs, again the score not the audio and I built and then this database I used them to do Daddy's Car, Hello World and all the research we did because we needed data so this database I built with this money from the European Commission and then we did this Daddy's Car and then we realized oh fuck that we have problems with copyright and we actually did not I knew that there were copyrights but the intent was never to exploit those copyrights and to steal anything it was just to show hey what we can do with AI is really interesting and at that point yes people in the management started to be really not at ease and I thought you know well it's just the wrong people but in fact no it was the beginning of this big wave of people worried really about, deeply worried about copyright being used for AI.

41:17And that was something interesting that you touched on a few times, which was, you said that research, like that has to be at the edges of the law, right? It has to, or near the acceptable point because you're doing something new. And then you later recorded an album with the blessings of the Pete Seeger family, using his voice, creating new Pete Seeger, inverted commas, Pete Seeger songs, this is when you were working at Spotify. But then that was also, it was released, but not pushed and promoted very hard for sort of similar reasons of concern. I mean, so that's, and that's something we're seeing today as well, this sort of anxiety around copyright and yet an enthusiasm for the technology.

42:03How do you sort of balance that up? Yeah, I think it's not rational at all, indeed. the only motivation that I had and we had in the team and especially this musician called Skiger, the French guy, Benoit Carré, was, hey, let's try to do things which are interesting musically, you know, and let's try to show that the technology we are doing is able to produce interesting, valuable musical stuff. And we didn't try to sell it, you know, to make money, just to show that we are progressing in our understanding of what makes a song interesting, that kind of stuff. So we were completely not into the game of stealing anyone's...

42:44Again, and as you said, we had the right from the family of Pizziger in that case. But at that time, it's hard to remember now, but it was 10 years ago, something a bit less, 8 years ago. It was really the beginning. People didn't know what to think, right? And now things have evolved into a kind of war. Okay, well, you have like two camps or maybe more than two, I don't know. And I'm not in that war anymore. I'm not interested in that. But I think maybe we have been naive or we have underestimated things. But I still think the problem of making a good song in the style of someone, of a copyrighted someone, is a very interesting problem.

43:28So if you can do it, then you can always ask the right to the guy. Yeah. When you think back to that experimental time where you were doing these projects based on intuition and feeling and care about music, you're a musician yourself. And then you look at now what we sort of as a culture consider AI music to be, which is huge monolithic platforms spitting out millions of plausible enough sounding songs. What do you feel about that and how it's changed? Yeah, I think the original goal that we had, and not only us, but many people. For instance, I could mention the Magenta team at Google. When they were created in 2014, maybe, the goal was the same.

44:18Like, do interesting stuff. You know, study what's an interesting sequence. So the goal was very different. So what happened is that the technology worked so well that the initial goals were shifted to industrial applications, problems of industrial applications. And these were not the goals we had really. But that's life and it happens very often. You start by a research question and then you end up with something that has a very concrete application. That's not exactly what you wanted, but that's life. That's the definition of open-ended research. you never know what's going to happen. Like I said, you were a musician, and we heard some examples on stage of AI music that you worked on.

45:00And to me, what I found fascinating was that the song sounded quite nuanced, and there was a lot of nuance in the sound and the compositions, of voices and silence and things like that. And I've obviously heard a lot of, let's call it today's AI generated music, which can feel fairly bombastic and one note in that sense. Have you played with those kind of big platforms? And what do you think that they do well, first of all? I know, of course, I play a lot. Well, I don't know. It's a good question. What do they do? So first of all, the lyrics is incredible. but the lyrics, very often you do it with ChatGPT and okay, but the lyrics are very good, I think.

45:53They are to the point, you know, you ask to talk about this and that, and it does it usually quite well. And the song is, the song, the sound is, again, is very good production-wise. It's like almost faultless now. And there is all the gimmicks of the style out there, you know, the intro, the special sounds. recently I generated something in the style of popular Greek music of the 50s, you know, and the Greek people said, my God, he even has the accent of taxi drivers in Athens, right? Which I didn't, okay. But so, yeah, it's good production-wise, it's incredible. And composition-wise, it's very, very conventional.

46:36That's what I would say. And it fits with what many people want. So that's why it was so popular. And then what could be improved? Because we've seen, for instance, in the recent week, I think it was Suno, released a version of their platform that is the wild version, which is supposed to tap into a sort of more unpredictable form of music. And it still feels familiar to the rest of the type of product that it creates. So what could they do better? And how would that work and sound? I think that, so first of all, I don't think anyone knows the limits of diffusion models, not more than we know the limits of LLMs.

47:18You know, every day or every month we are surprised by, look at what happened in math, for instance, two years ago, the mathematicians would say, oh, no, LLMs, it's bullshit. But now they are like, oh, fuck, we were wrong. So I don't know the limits of these systems. But what I think is that if you want to make an interesting song, as interesting as the song I was talking about, like by Jobim, McCartney and many, many others, then we miss the basic ingredients that we don't have in those algorithms. That's what I think. I don't have a proof of that, but that's what I think because again, good songwriting is very often not only about sampling, like going to the next, next note and next chord, but it's also about solving problems and anticipating things and stuff like that.

48:03and those algorithms by definition don't do that. So I would say, you know, I don't have a proof that they will never be able to do that but I think we need other kinds of algorithms or maybe enrich those algorithms with other things especially for controlling them, for saying, okay, I want this, I don't want that, I don't want diminished chords or I want diminished chords. No way, you know. They cannot do it because they don't know that there are diminished chords in which songs there are and that kind of stuff. Maybe one day they will know. We'll see. We'll see. It's not obvious because those technologies are, I was discussing this with a student just before, they are very good at sampling.

48:44They are not good at controlling. It's very hard to control. When you look at AI music now and with those models in mind and what people are doing with it, what do you think AI music is really for, like now, in a real sense, a human sense.

49:04Stu:What applications can you see, AI-generated music really being used for? Obviously, we're seeing people use it for fraud, but if we take that out of the equation, is it for, I don't know, people who think they can't make music, letting them make music? Or is it people remixing music that they like? Is this what it's for? Or is that just a very simple application of a complex technology. It's probably what Morricone call like functional music. I want a song for the birthday of my wife or whatever. For those occasions, it works really well. But it's also used by professionals now. I'm working on that also with a friend.

49:49Like, you know, when someone comes to you, a professional producer, and he says, oh, I want a song for this singer. and how do you say in words what you want? It's very hard. So you use Suno to create five or six versions of stuff just to help the conversation. And the guy, oh, that's the kind of stuff I want. And then you gain a lot of time. The guy say, ah, okay, I see what you mean. That kind of sound, that kind of rhythm. Okay, and then you can do the conventional way or you can extract tracks and replace and do all sorts of stuff. Actually, it's a lot of work to use AI in that kind of environment, but it can help, I would say, essentially guiding the conversation between the producer and an artist.

50:31And you know this conversation is very often very difficult because when I started, I played a bit a long time ago, I was like playing piano and the guy said, oh, I want chords which are a bit more sharp. What the fuck do you mean? There are already sharps in there, right? So for that thing, it's very helpful. It just speeds up the understanding. It solves the play it faster but slower problem. You were working at Spotify for a while, leading their development around that sort of research into technology. I'm not asking specifically about Spotify perhaps, but big streaming platforms are a little...

51:14They're trying to figure out how they should use AI-generated music. Should they treat it the same as human-generated? Is there an acceptable level? But how would you like to see streaming platforms like Spotify do something with AI music? There must be opportunities, like 24-7 infinite-generated streams. Would you like to see them doing that? And is that what they should be doing? So first of all it's true that the beginning was wild. They were like, okay, let's try to do AI and generate music. And then they realized, oh, this is going to raise some problems. So they stopped. And now they have done something which I think is quite smart.

51:57They focus on covers. Which means that they have a deal with the artist. The guy or the label say, okay, it's okay if you do covers with AI for that song. So they do things in a very restricted domain. and then everything becomes legal, which is their obsession now, is to be legal because they want to be seen as the good guys and they can't do any otherwise anyway. So they have found, I think, an interesting niche of making covers with AI. We'll see whether it works and things like that, but they have to fit with the laws now. It's no longer the jungle era that there was like 10 years ago. And when you look at the space now, And people have very strong reactions about AI in general, and they have very, very strong reactions about AI-generated music.

52:47Is there any one specific myth or belief or misunderstanding that you think has taken hold that frustrates you and you'd like to clear up? Well, the thing I said before was like, you know, AI is fed by what the works have in common. So if you complain, then you should complain, why do we do humans do we do so much things which is so similar in the first place that's one thing but then no i i i don't know what should be done i i just stick to my point if you are interested in a technologies that help doing good songs there is a lot of space you know the problem is not solved at all and today the premise shifted to something like to becoming a kind of commodity you know Everyone can do songs.

53:38That's all right. But we should not forget that music, even mainstream music, is not only about that. There are lots of... How do you make a song that survives, that will be played in 10 years or 20 years? That's not something that these technologies can do. I don't think so. You mentioned on stage that a sign of a good AI-generated song is one that you start to see people wanting to do a cover version of. and we haven't seen that yet. What other signals do you think there would be of AI-generated music sort of crossing over to a new level of acceptance? That one, for instance. Precisely that one.

54:18No, I mean, I don't want to be critical because I know people who love some songs which are completely AI-generated and I don't see the problem with that. I mean, there are legal problems, maybe. But so I think the AI is already able to do things that touches some people. But I'm talking about songs that make history, right? That are so good. And it's like in math or chess. The AI does things that humans can't do now. In almost every domain. So I'm waiting for AI to do songs that humans can't do. That's what I'm waiting for. And I'm not saying it will not happen. I hope it will happen. I'm very curious to see what it will be.

55:00But it's not there yet. And the final question then, you're still working in this space, you're doing research and making new things happen. Which direction are you looking in? Where are you trying to take your research? You mentioned that you can essentially work on your own now instead of having to have a room full of 15 people because of AI coding technologies. So that's quite exciting because you are limited by your capacity to think of things. So where are you trying to take it? I talked a bit about that. I'm trying to go in the composition domain, not sound. Production is almost perfect now.

55:41How do I create a song, even like a jazz standard? It's so hard. A good one. A jazz standard is a song that all the jazz musicians in the jam session they want to play. Because they think it's good because it's interesting harmonies and stuff like that. A good bossa nova, as good as the ones that we know. A good mainstream pop song that people can whistle, you know, because it's so nice. That kind of thing. That's where I want to go. And I'm saying that the technology we have now, Transformers and others, they are not enough to do that. They're not enough. And as you said, it's a very, very exciting time for researchers now because precisely they can work alone or with two people.

56:27You take two good researchers with Codex and Cloud, I mean, they do miracles. They do incredible things. You don't need teams anymore. And I wonder if that's where we might see this great first great AI song come from, because you don't need to know the technology now. You can be a creative person and be liberated by that, perhaps. Well, that's a debate. I mean, how far can you go with coding without knowing a bit about coding? I think you still need to know quite a lot of stuff, but I don't know. I mean, anyway, yeah, maybe people who don't know much will do things. Probably, probably. It's still, you know, it's like it's changing every day.

57:05So, you know, maybe next year will be completely different, you know, with agents and with whatever will come up, you know. But it's very exciting for people who have ideas. Is there, final, final question. Is there one thing that if you achieved it with your work at the moment, you think, okay, I've really satisfied something. I've satisfied myself by achieving that thing. a song that McCartney is still alive, but others say, oh, that's a good song. That's enough. That's enough. That's a good place to leave it. Francois Pache, thank you very much. Thanks. Okay, so there we go, Stu. You've also put a new story up about this.

57:48Just quickly, we will talk about this more, by the way, in the lock-in in a few moments, folks. So for those of you who are super fans, we will go over this in a little bit more detail, discussing it from by the bar. But what did you think about what he says about his views on AI music and what's happening now?

58:06Stu:Yeah. Do you know what I thought was interesting? I learned most about human songwriting, actually. I thought that his argument basically that AI models, current ones, can produce flawless, conventional sounding tracks. But he seems to be saying they haven't solved the real challenge of true musical composition. And I thought he was talking about this idea of songwriting is not just, well, this chord comes after this chord, this melody works in this way. It's not just a structural bit of blocks. It's about anticipating and problem solving and making these deliberate choices, isn't it? And that was interesting to me, saying that there is more to human songwriting than just what goes after what.

58:46Stu:And he seems to be saying that's what current Aeo music is. It's not solving problems, it's not doing what humans do, which is kind of almost heartening. But then he's like, I believe this is what with a soul. He didn't think it's impossible. He wasn't saying it's not possible for doing that. He's saying it hasn't done it yet. Yeah, I thought he was fascinating because he was very clear that he didn't really care about the legalities of it. He was really interested in making technology that did something. He wanted to really make a really, really good song that a human wasn't involved with. And his benchmark was like, you know, if Paul McCartney liked a song that was made by a computer, then I think that's an achievement.

59:19Because from his perspective, all songs that are truly good are additive to humanity and i think by definition what he's saying is that there are not yet any truly good ai generated songs because we don't feel that we are our music the general consensus is our lives have not been improved by ai music there's been some good attempts but nothing that really makes us go wow i'm really pleased that exists in my life and he said one more interesting thing we'll talk about this more in the in the lock-in which is he said that actually really good songwriters won't share with you that final bit of data.

59:54We can sort of use machines to sort of learn from their songs and stuff like that, but they won't share their secret sauce. So he said that Paul McCartney, if you ask him, won't explain his songwriting process to anybody. And that's not because he's protecting his intellectual properties, because he as an artist does not want to think about it or discuss it. He doesn't want to say, In case you miss up. In case, yeah. In case it stops working. In case he kills the goose that loads of gold. No. So I wonder if there's actually some things that will be forever unknowable that will only exist in human brains and people won't want to talk about it just because they don't want to stop it from working.

1:00:31Anyway, talking about musical genius, Stu, and finally, Lucy in the Sky with Diamonds, Mr. Tambourine Man, Common People by Pulp, and now The Ace of Spades. What links those four iconic tracks? This is like an only connect question from the quiz show.

1:00:50Stu:Well, what links them is they've all been covered by William Shatner. Wow. So he, I mean, the first three of those were from his albums from, well, first few back in the day. And recently he did it by one with Common People. And now he's basically, he's formed a metal band. So it's William Shatner and the Uckers, I believe. So I believe he's not quite, he doesn't want to swear. Yeah. And they, I think they played a club in Chicago because they're going to be playing a festival called Brightfest. And he's got an album coming out called something like What the F is Heavy Metal, where he's got like, I think it's like Brian May and Rob Howe from Judas Priest and like all kinds of metal luminaries on it.

1:01:32Stu:And so basically he played this club, it's on YouTube, you can go and find it. And he covered the Ace of Spades, he covered Enter Sandman as well. And if you've ever heard any William Shatner music, it's like his very, very, very, very unique vocal style with a flair for the dramatic. And the thing is, he's 95, one. Which is remarkable, by the way. He's been nearly into space in recent years, and now he's fronting a metal band. So, you know, if I'm doing that when I'm 95, I'll be very happy. And he seems to be having a brilliant time. It's not like someone has shoved him on stage and said, this will get some money, do a metal band.

1:02:14Stu:He seems to be having a time of his life playing with this kick-ass metal band, covering songs that he admitted he knew nothing about until recently. He's not like a lifelong metal fan or anything, living out a dream. He just discovered metal recently. He's like, right, for a band. Yeah. It's quite something. I will post a video to William Shatner doing Ace of Spades in another podcast, but it is really something and you must watch it because it really is. It does look like a 95-year-old man who, by the way, looks pretty good for his age, I will say. It looks like a 95-year-old man who recently discovered heavy metal and is now quasi-singing them with a really good bunch of musicians.

1:02:53More of that.

1:02:54Stu:Have you ever seen... Oh, what's it called? um what's that it's a casino in hard rock cafe in las vegas i once went to a trade show there in las vegas and i went with this it was like rock karaoke and they had a real crack musicians backing band wow okay and anyone could come up and sing in front of them you know they were like and it was all like tech conference delegates who were going um but it's a bit like that but i'm quite fond of him now like i've seen him interview with people who try and get him to political views, going back to our Ed Sheeran start. And William Shatley is very good at talking about, well, you know, when you've been thinking about space and life, you look at the Earth and say, we're all one people, aren't we?

1:03:36Stu:It's slightly kind of cliche, but it's also quite profound. He's very much like, shouldn't we all be getting along? And he could be a 95-year-old massive reactionary, and he doesn't seem to be. He seems to be very open-minded, curious, and kind of having fun. Interesting. Well, there you go, folks. I will put the link to that, blood the podcast, and check it out. That is it from us for this week. Thanks so much for listening as ever. If you're enjoying the podcast, we would love you to spread the word a bit with a like, a follow, a subscribe, and a share. All of the above, hypothetically, which really does help us.

1:04:08And thank you for doing that. And a huge thanks as ever to those of us, those of you who support us on Patreon and become super fans. If you would like to get 20 or 25 minutes more blah blah from me and Stu every week, which occurs immediately after the show, you can go to patreon.com slash the price of music and for the price I would say if you live in London for the price of half a pint every month you can get all that extra stuff alternatively if you'd like to get in touch with us about anything we've talked about today or you want to cancel Stu for anything he said earlier or any questions on the music industry then you can drop us a line at the price of music podcast at gmail.com it's the price of music podcast at gmail.com all that is a red And we have a very good letter today, Stu, in the lock-in, which is happening very shortly.

1:04:56And that's it. Thanks for joining us. See you next time.