Fireside Chat: Music and AI, Part 2: The Human Touch: Imperfections and Intent

In this five part series on music and AI, filmed at the twenty twenty-four Marketing AI conference or May Con, I sit down with composer Ruby King to discuss AI, music, and the future of creativity. This is part two the human touch, imperfections and intent. Let's try another piece that we're gonna switch to a something that's theoretically a Viennese waltz. We'll find out if it actually is or not. So what was that?

How was that? Um it was very interesting. It was quite nice. Okay. It was quite nice.

And the first thing I was noticing was again sort of going into like how was this recorded if it was by humans? And it sounded like I could have been wrong because I'm sat like away from it, but it sounded like it was a bit of crackle on the recording. Interesting. Like a little bit. And I was kind of wondering, it sounded like a recording that my old violin teacher would give me from her like cassettes, and it was like, this is what the piece you are playing, and it was recorded a very long time ago in black and white.

And like it's just the EQ as well, like the the quality of the sound. It doesn't, it didn't sound like it was recorded, it didn't sound like I was in the room with it. And that's not a bad thing. It's just it sounded like older recording. So I'm wondering if maybe the info it was fed was from the past.

Highly likely. Yeah, okay. As to whether it's a Vienne's waltz, I can't answer that because I don't have I can't be sure enough. I'm not an expert and someone will shoot me down online if I say one or the other thing, so I'm going to back out of that song slowly. Okay.

But it started faster and it got it had some really slow bits in it. That was nice. I was surprised by how many different sections there were that seemed to be trying to do a theme and variation. I'm not completely sure it did do a theme and variation because I think it might have forgotten exactly what it did, or the music wasn't memorable enough to remember. I would have to look at the actual notes themselves and listen to it quite a number of times more.

But it sounded like it was attempting to do a theme and variation. And I think waltzes are usually three, four, because one, two, three. One, two, three. And most of that was. I think one bit seemed to be six, eight.

Is it six, eight, or six? My music teacher's gonna kill me. But sort of like where it's longer, one, two, three, four, five, six, one, two, three, four, five, six. So it's still the same feel, but the phrasing doesn't stop midway through the bar, it ends at the end. If that makes any sense.

No, it makes total sense. Okay. But it it it calls back to how the Transformers model works, where it's doing things in chunks and instead of sewing it back together. One of the things that you'll notice with tools like Suno, particularly for longer compositions, they lose coherence like two and a half, three, three and a half minutes into the song. Sometimes they'll just don't go totally off the rails, and you're not sure why.

But then when you look underneath the hood, like, oh, it's because it did it's doing 30-second chunks, and it either forgot, or there was a token that got misplaced along in like chunk 13 that just throws the whole thing off. Yeah, like that was that was enjoyable listening. But another thing that because I've done violin for quite a number of years, and one of the key things I was listening to there was is this a real human playing it? Is this playable? And because it's one of the main things you can hear in music that's like violin or viola or whatever lead is the Boeings.

And most people don't bother about this, which is why I'm insufferable to watch like TV with like an Umbrella Academy when they whip out the violin, I'm like, nah! Because it was it was a bit loud, sorry. It was it or like a Sherlock, where there's not quite what he's doing, the hand motions don't match the actual audio. It's just so painful. Like, just get someone who can, please.

Please, we exist. Just the fingers, or the editor's mess it up afterwards, and like I don't blame them, but okay, anyways. Like one of the main things is bowing, and you can hear it because if it's like up and down, up and down, you can hear how the string, it sort of changes. There's a break in between the noise. If you're slurring, which means like going from one note to the other with the same bow, it's only the note that changes.

There's not really a pause in the middle. So most of that was separate bowings, especially when it was doing quite fast jumps. And at one point it was like I pulled a face at one of one of those face pullings was because it was doing something quite fast with a lot of jumping, and their right hand must have been going like that. It is very possibly possible, but that player deserves an award. Probably some aspirin.

If the composers specified that that would be how you should do it, then you would probably try and do it like that. But a violinist would naturally try and not do that much effort. Um it doesn't sound right being so separate either. If it was more together just in phrases, like a few notes in the bar, maybe just if it's three, four, then maybe six of those could have been in one bow, and then another six in another bow, and that would have still kept the phrasing, but just the way that it's performed, it's not thinking about performance rules, it's just thinking these are the notes, and this is a violent sound, therefore this is correct. Right.

So AI is essentially synthesizing from the the outcome of the data, but does not understand the techniques used to make the data. Yeah, I think so. Okay. Because there's a few times there, I think in that piece, it would have been nice if there was a slide up because it does do some jumps. And the nice thing with the violin is it's not a piano.

I just can say that I play both. But um on the piano, you can do slides, you can do glissandos, but it's easier on a violin because you you can kind of slide up to a note and add some nicer variety once you get there. Piano, it's a bit more like you can't get all those micro tones between. So it's kind of like you just got to be. Because there's defined intervals between one note and the next.

So if you are a content creator using generative AI, the instruments that have more analog between notes like a violin as opposed to being more defined segments, it'll be more of a giveaway that you're using AI because the machines are not synthesizing the technique needed to create the sound. Yeah, so it is said that the violin or strings are the closest thing to the human voice. And the human voice, we can easily tell most of the time when it is not a human voice. Like, well, okay, saying that there's a lot of like speech that is very, very good now, and you can't really tell. Um, those models are very advanced and it sounds very good.

But singing isn't quite there yet. I'm assuming because there's so many different techniques, all the way down to breathing and like where you hold your head, like it's gonna the sound's gonna be different if you're looking up to the if you're looking down, and like it's just because of the like there's so many variables. So the violin and singing, like the human voice, are dead giveaways. Well, voice more so than violin, because I think you have to be a bit more trained on violin usually, but still it is coming down to the subconscious. When you listen to that, are you thinking this is a real performer?

I can imagine being sat in a concert hall, or you going, This is a violin. I can't say anything more about it than that. Right. Okay. Let's try a piano piece.

So this one is supposedly a ragtime. That's not great. Slides in there. You'd expect the violin to come back in the same there. So this is an example where it should have stopped.

Unless it's I mean it should have stopped if that was the intention, but if this was the intention, then that would be an impressive thing for a human to do. So the prompt on this was just an upbeat bag time. Then it's just it liked when it did, it was like I'm gonna do more before you shut me down. I'm done now. Okay.

So that was an example with in particular with Suno, where it will have these they're almost hallucinations where it sounds like there's a logical stop of the song and it's like I'm just gonna keep going. Did it have was it given the time frame it had to? No. So it just it was just I haven't too much of a fun time here. You can outstab me.

If I had to guess, and this is pure speculation, the model had enough fragments of enough tokens left over to forecast from that it was like I don't know if I should stop or not. Okay. Um it was definitely it did come feel like it came to an end. But it continuing wasn't necessarily wrong. So it wasn't right for the what you said.

But if you were in a concert hall and that was played, and then there was a pause, and sometimes you do do that, which is why you're told do not clap in the middle of a movement, because you will look like an idiot and everyone will stare at you, and the musicians will remember you for the random time. True story. Then like it's kind of it sort of felt like a different movement. Right. It was it did the music didn't necessarily feel connected, but it felt like a nice different piece.

You might expect if it was meant to be the same piece, you'd expect the first piece of music to come back again, and that would be sort of a bigger end, and that would be an ABA structure, because you sort of had the A and then the B, then it just ended. It was a nice end, but it wasn't necessarily the way you'd expect a piece to be. A and B does happen, that is okay, but for the instructions you gave it, you'd expect it to go back to the A section. So it's not that it's wrong, it's just it's not it's wrong for what you asked it to do. But musically speaking, if you handed that in, it would be probably be seen as a good thing that you did something creative, like a false end, and then you continued with something that was different, and it was like an oh moment, which is a good thing.

So musically it's good, prompt-wise, not so good. Right. Okay. Um, let's try one more piece, and then we're I want to talk about how we can for people who want to use these tools, how we can get better performance out of them. So this is the last one.

This is the right time. Yeah. Good baseline. It's repeating itself. I don't know.

It's good. B section quite thick. Oh. That is a variation of theme A. That's definitely theme Af speed.

No. Left out. Slow B. That is bigger. Can't remember being well enough, but I remember A.

Oh. Is that a key change or is that just okay, back to the longer than I thought they could be, and it's remembering itself quite well. You would probably want it to be doing something more exciting with this now, because it is just feeling like it's gone back to the start. Left hand should probably be doing something more interesting. It's an odd end.

I mean it's a valid end. It's not what I would have done, but it it ends fairly convincingly. That concludes part two. Thanks for tuning in, and stay tuned for part three. See ya on the next one.

If you enjoyed this video, please hit the like button. Subscribe to my channel if you haven't already, and if you want to know when new videos are available, hit the bell button to be notified as soon as new content is live.