Episode 32: AI Vocals

The Impact of AI Vocals on Music and Creativity

In this episode of the 'Face Your Ears' podcast, hosts Justin Hochella and Rich Bozic discuss the rapid advancements in AI technology, specifically focusing on AI's ability to generate and manipulate vocal performances. They debate the implications of AI-driven tools like Deep Seek and ACE Studio, which can now replicate and alter human vocals for a fraction of the cost previously required. The hosts express concerns about AI's role in potentially eroding the human elements of music and artwork, with Rich offering a singer's perspective on the importance of human experience and toil in creating meaningful art. They also touch on historical parallels, such as the introduction of auto-tune, and ponder the future of live performances in an increasingly AI-dominated landscape.

00:00 Introduction and Greetings
00:31 Podcast Updates and Social Media
01:22 The Rapid Advancement of AI
02:58 AI in Music: A Game Changer
04:19 The Rise of AI Vocals
05:48 Concerns and Criticisms of AI in Creative Arts
07:02 AI Tools and Their Capabilities
16:39 The Future of AI in Music
20:07 Philosophical Implications of AI
26:38 Conclusion and Audience Engagement

Ace Studio: https://acestudio.ai

For Production Services, Coaching and Lessons, visit:

www.rjbmusicproduction.com

For help prepping your songs/voice for the studio:

www.bozicvoicestudio.com

----

https://www.youtube.com/@FaceYourEarsPodcast/shorts

https://www.faceyourears.com

https://www.instagram.com/faceyourearspodcast/

https://www.facebook.com/people/Face-Your-Ears-Podcast/

Transcript

Justin 0:07

Hello everyone. And welcome to another episode of the face your ears podcast. My name is Justin Hoshela and I am joined by Rich Bozek. Rich. Hello. How are you?

Rich 0:21

Hello. Good day to you, sir.

Justin 0:23

Good day.

Rich 0:24

It's a beautiful day today, it snowed slightly outside

Justin 0:27

Ah, yes. It's snowed here as well. Just a bit.

Rich 0:31

before we get into our topic, uh, I want to let everyone know that we do have a website. faceyourears. com that has all the info about our podcast and even a podcast player if you don't have a way to listen, but I'm assuming you probably have a way to listen if you're listening to this podcast.

Justin 0:57

dropped your phone in the toilet, like five seconds ago and as you were listening to this episode, run to your computer and go to face your ears. com you can check us out there.

Rich 1:11

we're on social media. look us up at face your ears. podcast. So Justin, what brings us here today?

Justin 1:23

The juggernaut that is AI continues to plow through our lives, whether we want it to or not, and. it seems that no area of life is left untouched. And we on the face of yours podcast have talked about AI. It was quite a while ago. And I think that AI has advanced something like a billion fold since, since that episode, it just seems like every day. We're. Looking at some crazy new advancement in AI technology. One of recent advancements was with deep seek, which is a just an AI sort of engine, whatever it's called. And the big hubbub about that is it is on par with open AI, which is one of the most premier. AI engines on the planet. And this deep seek engine was able to perform. At the same level as open AI, but they did it for like a fraction of the cost. And so I just mentioned that because it's like the advancements are insane. To be able to do what open AI does, but for a fraction of the budget has huge implications for all of AI and technology.

Rich 2:50

that means more of it being produced faster in

Justin 2:54

Yeah, just the advancement of it. If you can do something for a fraction of the cost, right? Look at what we do with music and technology, right? Like once upon a time, these ways of recording were so out of our reach and so expensive that, we could only dream of being able to do it. And then, one day somebody came along and made it far cheaper or at least over the years made it far cheaper. And. Here we are in a studio filled with a bunch of cool stuff, able to record far better quality than anyone was, 20, 30 years ago. So it's, it just advances, right? To be able to do it at that scale for that cheap. I think it was something like, I don't know. I'm, yell at me in the comments listeners, but I'm going to, I'm going to get these numbers wrong. But it was like, Open AI took something like, I'm just going to make up a number of 200 billion and then deep sea. Cost something like 30 million or something like that. Just a fraction of what those guys spent. And so it's incredible. And it's it's also an app that you can download to your phone for free. I should mention that

Rich 4:01

So we have access to that right away. Oh man.

Justin 4:04

I just, again, I just mentioned this to set the stage to be like, this is where we're at in February of 2025 with AI, like it's just exploding.

Rich 4:14

Actually, this is an episode that I've been dreading would come. Today we talk about, AI vocals

Justin 4:22

AI is advancing so much that we're now able to do something that I think, Rich, you and I both thought would be Was at least way off, if not even possible at all, but it's the ability to use AI to generate vocals.

Rich 4:37

Hold on now. I got to go on a little rant here for a second.

Justin 4:41

Warning, listener

Rich 4:42

Yeah what the hell, man? Like, okay, so we develop this crazy technology that's amazing to do stuff for us, to make life easier, to do all kinds of tedious stuff for us. We develop all of this stuff, and the first thing we do with it is we, Oh, let's wipe out the visual artists and the the, all the musicians and the people who write stories and the people who do graphic design. You would think that we'd be like, trying to teach the thing to handle the laundry.

Justin 5:19

to our taxes.

Rich 5:20

Instead, oh no, all of the things that are innately human, the greatest things that humanity can do that are unique to human beings, let's get a robot or an AI, an artificial intelligence to do that for us. It's just there's something weird about it. And also things on the physical front, we'll call it that.

Justin 5:41

Yeah.

Rich 5:41

You want to replace that stuff. Like first, I don't understand. I don't understand. Please help me understand, Justin. Help me understand.

Justin 5:48

I think AI is very powerful as a utility to assist in time saving tasks. That's probably a very narrow thing that I think AI can do and that I appreciate that it can do. But, I think when I see it start to take over, the things that make us human, So the creative arts, like you described, I do start to get concerned. And one of the things that I feel like AI is doing is flattening culture. And so we've got, billions of people who are able to produce something based off of centuries, really, of historical, creation. That's what AI is doing is it's basically like looking back on the centuries of what humans have done and, regurgitating something based on those parameters. And I think it's just, it's sad that, people look at that and. Harold that as like a good thing, because to me, it, it's It does flatten culture in that, I can't imagine how like a robot could help perpetuate and grow human culture, right?

Rich 7:02

Yeah.

Justin 7:02

And I don't know that's a more deep philosophical view, but what about vocals though, like to get to that topic specifically, talk a little bit before we rant about it. Cause I do have a lot of rants about AI and vocals, but what are some of the things just so people are aware what are we talking about? What are some of the things that you've seen in regards to AI vocals that. You didn't think would be possible

Rich 7:28

the first thing that comes to mind, there was a video that I stumbled upon in my algorithm. The algorithm throws me all kinds of voice related stuff. Of course, listening to my conversations all day about singing that video there was. a guy singing and recorded a track and then he's but I want my voice to sound more like a gospel singer. And so he was able to throw this AI plug in and suddenly the melody that he sang got transformed and had all of these Uh, this phrasing and tone that sounded more like a gospel singer. And then he's like, you know what? No, I'm not satisfied with this. I want it to sound more like a, like a country singer. And so he switched it up to that. It was truly impressive. It was still his voice. And then he changed it up even more and was like, I want my voice to be a woman's voice. And then he was able to change it up to be a woman's voice. And then he was able to change more parameters of it to change the tone quality even more. It was truly impressive. If you didn't know what you were looking at and just the average person was just listening, they'd be like, Oh yeah, that's, that's an R and B singer. That's a woman singing. They wouldn't blink an eye. They wouldn't know the difference. And to imagine that's still in its early stages. Cause I know some of the critiques of all that is all, it'll never capture the phrasing. It'll never be able to have that nuance, that spark that humans have. But I don't know. It was a really convincing start. Have you heard any of this?

Justin 9:06

I don't know that I've heard that where people are changing their vocal style I think I've heard of it like now you talk about it. It sounds familiar but yeah, that's the thing too with AI is like any critique. just give it a week it's it's it's moving so quickly Yeah wow. You were looking at something that changes the person's actual voice. That's crazy.

Rich 9:28

literally anyone can walk in and be like, okay, I have a song idea. I can't sing well. Let me sing it the best I can and then literally like use, you could use something to fix the pitch and then you could throw this plug in on it to make it sound like something else to improve the tone quality and the delivery.

Justin 9:48

Wow, so we've gone from like auto pitch to auto sing

Rich 9:52

Yeah, I,

Justin 9:53

is what it sounds like.

Rich 9:55

You know what though? When I see that kind of stuff, man, it starts to make me question what I'm hearing out there. Is everything that we're hearing what we think it is? Or have people already been utilizing this stuff? Because think about back in our day when Milli Vanilli got caught lip syncing. For those of you who don't know tell them about Milli Vanilli.

Justin 10:18

Vanilli were a duo, a music duo back in the 80s. It's actually a very tragic story. And anyway we'll, we won't go into that, but the big thing about Milli Vanilli is that they, they were basically these two guys that were hired cause they looked a certain way. They had a cool kind of model look but they couldn't sing. Like at all. And yeah, they would lip sync. In their music videos, of course, and in their performances on stage. And then one day the jig was up and their backing track messed up and they were

Rich 10:56

think it was like skipping live. It was, it was

Justin 10:59

Like they were caught in the act of like lip syncing and it was controversy at the time you know, at that time everyone just assumed like everybody was singing.

Rich 11:07

Didn't they have to return their Grammy, right?

Justin 11:10

so. I think they did.

Rich 11:12

But, yeah, it didn't end well with them, ultimately. But, Milli Vanilli, you're forgiven, because of all the stuff that's happening right now that just happens in plain sight, and no one blinks an eye anymore. Like I'm hearing now even rock and metal bands getting busted for using that kind of stuff live, but this is on a new level. AI vocals is on a new level. Like you could transform someone's voice completely to where it's not even, it's beyond lip syncing. It's you're basically on the spot doing a complete transformation of the voice.

Justin 11:47

One of the tools that I've just to kind of jump in on that topic, just to kind of flesh out what's out there is ACE ACE studio. it's, it's more or less like a DAW. based on a whole bunch of different singers. I a couple dozen, if not more singers. So the idea is that you go into this and you can, you know, play in like a melody or something with your MIDI keyboard or draw it in either one. Then you can type in lyrics, like whatever lyrics you want, you can type in and the program will automatically. Contour the lyrics to the notes in your melody. And from there you simply select a singer, like one of those dozens of singers that they have, and that's it. These singing voices are based off of actual real human being singers. What the AI is doing is it's taking their vocal qualities and nuances and. creating a virtual singer out of it. So you can

Rich 12:46

from scratch, you literally can not even be, you can be someone who can't even make a sound.

Justin 12:52

yeah, I mean, you don't, you're right. You don't even have to sing at all. it's almost like a word processor crossed with a synthesizer is how it comes across and they actually just released a new feature where it now links to your DAW. So if you're using logic pro or Ableton live, you can you know, have a melody or something from your DAW. Um, Talk to this ace studio application. Cause the way it was before is it was its own standalone thing. You would create your vocals, you would render it and then put it into your DAWs like a wave file, but now they, the apps will actually talk to each other and so it's basically like an instrument in your DAW

Rich 13:38

Wow. Wow. You know

Justin 13:40

crazy stuff.

Rich 13:41

Singers should have seen this coming. Now singers know what drummers must have felt like. when suddenly synthesized drums were introduced into the equation or any instrument for that matter that was starting to be replaced with the digital counterpart. I think what it is is because humans make the sound, um, there was this idea that, oh, that synthesized. But hey, man, technology. technology and look where we're at now. And now singers feel what, all those other instruments felt.

Justin 14:23

They have over 80 royalty free AI singers available. And so they have different styles of singing, So they have like a soul singer, pop, EDM singer, cinematic singer, opera singer, even have a child voice. They have hip hop, ballad, R and B, Latin pop, R and B funk, soul, Latin pop, Latin folk. On and on. crazy. Like, They have different styles now of all these different

Rich 14:51

Is there a plugin that could add in a little bit of that singer ego into it?

Justin 14:57

It's it's just like, uh, like a knob on it

Rich 15:00

level of ego, ego Level

Justin 15:02

of ego. Yeah. And there's like a Like an icon of somebody's like, head, and as you turn the knob up, the head just gets bigger and bigger.

Rich 15:13

And then you could have the level of room dryness that the singer takes into consideration. Like this room is really dry and the singer will be extra like edgy.

Justin 15:23

Yeah. Hydration level, like, how sick they

Rich 15:28

yes, yes, Yes. All singers walk around with a low level, sickness. It's like the easy out

Justin 15:34

true, It's true, I So many singers that I've dealt with, it seems like that it's like, are you ever not sick? And is, is

Rich 15:43

who deals with singers, yes, we are never not sick. We're always

Justin 15:47

there's a, always a perpetual low grade, like head

Rich 15:51

Yes.

Justin 15:51

at all times.

Rich 15:54

But if it goes well, then, of course, this is what I do, this is how, you know, how I am. But if it goes bad,

Justin 16:03

You know, There's there's another thing I just thought about. I remember seeing a YouTube video about this where another aspect of AI and vocals is harmonies. a tool that allowed you to generate layers of harmony within like seconds, it was insane. So you would upload, I think in this instance, you would upload yourself singing or you would upload a take or something. And then it would generate like, you know, three, five uh, seven part harmonies and stuff like it was wild. Um, you know, Based off of just one recording. you know, we Say all of this just to really describe like kind of what's out there today and it's mind boggling and Richie brought it up earlier where you said like this is bleeding edge technology it's very new. hasn't been around for long at all, maybe a year ish. I mean, I S I think I started seeing this really start to blow up. Like Six months ago, six, seven months ago. tell me more about your, view on this. Rich, you've dedicated your life to singing and being a vocal teacher. You've been doing this now for almost 20

Rich 17:11

years actually 25. years this year 25th anniversary

Justin 17:16

I'm sorry. The Bosick voice studio is hitting 25 years. That's right. So yeah. Like, how does this, strike you? And I'm curious too, are, are any of your students like talking about this or reacting to it?

Rich 17:28

so far, I think that It hasn't hit like the mainstream yet Um, my students aren't talking about it really that much yet. Then again I deal with people who really They want to learn how to sing. They want to be able to do it themselves. not looking into this kind of technology. I bet if we were to talk to a lot of producers and people who are trying to build songs and music from the ground up themselves and are familiar with equipment, I bet more of those people probably have heard of these things and have dabbled. Yeah, I know. Pitch correction, auto tune, all that business melodyne, all that stuff has already been in the zeitgeist with singers for a while with varying opinions on that usage. But this is something different. I haven't had yet, even like on the recording front, any client yet. It was like, Hey, can we try to change the sound of my voice completely with AI? Haven't had that come up yet. I'm waiting for that day to come. But nothing yet from the singer world that I'm hearing. I have been slowly putting on my feelers to some of my clients who are in other aspects of voice usage. So I have a client who does sync licensing stuff. And I asked her, what's going on in the sync world? How do they perceive AI in that world? I'm not only on the vocal front, but just song generation altogether. And she was saying that there are some distributors or agents or whoever handle that stuff who reject AI completely. But I know some of that infrastructure in that world is built on an old model. So I could see that being at first people resisting, but then I could see also eventually as, Companies who are using sync licensed music, becoming aware of the possibility of, oh, we could just do this all in house. We could just generate that could start to catch on just like in film, all the AI stuff that's starting to come into making a film and all of the jobs that's eliminated.

Justin 19:39

it reminds me of, you've mentioned auto tune. I think we've talked about this on the podcast before, but. When it first came out and auto tune, like people lost their minds. Like they thought it was the devil and fast forward 25 or so years later, like, cause I think it really started to hit around the late nineties, early two thousands. And so now it's almost Like, Of course we're going to have autotune on vocals. Like it's just part of vocal processing. Now it's so standard and I just uh, it makes me cringe to kind of think about like, where are we headed then with AI? Like, I think that's like the bigger fear with AI is like people feeling like they're, they're getting replaced or going to get

Rich 20:21

Yeah. On a lot of fronts, not just the music world, obviously, in other areas as well. with music, I just keep coming back. I can't shake this idea that it's just such a human thing. It's like an expression of human emotion and human thought and ideas. It's just rough to think that one day all of that may be replaced with just music. derivative kind of AI generated stuff because right now AI is all derivative from whatever knowledge you knowledge base you give to it. I think a few weeks back there was a whistleblower who was on the ground floor, I guess, in helping develop a generative AI who was a whistleblower and saying they're using intellectual property to build AI and kind of sounding the alarm bells. And the guy was like found like in his apartment dead, supposedly a suicide after doing that. That's a whole nother thing, but yeah, that's a whole nother thing. But. if you listen to what he was saying, all kinds of things, you know, however you feel about copyright, intellectual property, all that stuff, like all of that stuff gets called into question as well. Where's all of the information coming from? How's it being used? But then with voices too are plugin packages eventually where it's like by the Beyonce and Steve Tyler. Vocal plugin to sound, like these people. Is that coming? Think about the people who had to record all the samples for that, whatever company you were talking about that has those.

Justin 22:04

Ace Studio. Yeah.

Rich 22:05

They had to pay someone to generate the sound for all of that stuff to be able to be used.

Justin 22:14

it strikes me as sort of a hedonistic pursuit. And what I mean by that, by hedonistic, is it's, it's a really narrowly focused pursuit of pleasure. It's like someone just fixating on the end result or the product. And that's what AI is able to do and it's really dangerous, I think, for human beings to have the capacity to exist in a hedonistic way because it introduces an idea called The paradox of hedonism. also called a pleasure paradox and it refers to the practical difficulties encountered in the pursuit of pleasure for the hedonist, constant pleasure seeking may not yield the most actual pleasure or happiness in the longterm when consciously pursuing pleasures interferes with experiencing it. So the more you pursue that thing, the less pleasurable it becomes. know, AI vocals might seem really cool now, but the further we go with it, the less it's going to hit, if that makes sense, it's almost like the drug isn't having the same effect anymore. And it kind of becomes hollow and meaningless. And I think the reason that, you know, for musicians, whether you're playing an instrument or you're singing. The reason it's such a visceral experience is because you are doing it. You are literally feeling it in your body, especially with singing, is your, it is your body. Your body is

Rich 23:35

It's the instrument you can't put down.

Justin 23:37

Yeah. And there's this, there's something spiritual about that. it's almost like, you know, this is what my soul sounds like, when I sing. when you remove that, you're sort of removing the soul from the music. Maybe for argument's sake, you know, maybe there's an argument against that, but, my biggest concern. The thing that makes me sad is that people are oftentimes short sighted because they're hedonistic. They're like, I just want the end result. I want the pleasure out of it. the quick fix kind of thing to be happy in the short term. they don't realize that in the longterm, they're basically going to feel hollow. And, that's something uh, I think AI can never replicate, which is the experience AI can never. Have the experience

Rich 24:24

Yeah. We say that now.

Justin 24:26

until the brain chip comes out and can implant memories like in blade runner or you know, Until that you know, AI cannot have the

Rich 24:34

sure. And why would we want it to right? And there was something else you talked about, like as we were prepping for this episode that I want you to talk about for our listeners the concept of actively having the struggle and the work element of things. Can you talk a little bit about that?

Justin 24:53

Yeah, I think that that is part of the experience the experience of toil and failure and, uh, frustration and, all of those things that shape who we are. So it's kind of a, granular part of the experience is, the suffering. I think that that. Ability to suffer is quintessentially human, suffer in the pursuit of something that you love. Is a quintessentially human endeavor. I mean, Think about your own experience as a music student. You know, I was a music student. There's a lot of suffering there, you know, Late nights, uh, studying or practicing you know, having to do performances and stuff and be graded on it, like it was a lot of pressure, a lot of, difficulty and those experiences shaped who we are. As people and as musicians, and I think that if you remove that to me, it goes back to that flattening culture a little bit because like if people aren't having experiences, which are very much random and unpredictable, like that can generate just interesting things. And that. To me is a huge tenant of building out cultures. People just having the human experience in their own unique way. And if you can just go to a computer and type up like, I want this perfect thing, you completely cut out all of that toil and experience and so forth. it's it's another element. Humanity is that we as human beings learn best through experience. And so one might argue then that if you're able to do all this stuff without learning any of it, you're not really growing as a

Rich 26:36

Yeah. this was a interesting thing to delve into. Again, I knew it, I knew AI would be coming back for us again, and I have a feeling we're gonna be back again. Justin talking about this again.

Justin 26:51

I mean, There's gonna be something else, They've come for our instruments and now they've come for our voice and so,

Rich 26:57

What, what, What's left? Yeah, Well, the music creation, they've come from the music creation with Suno and all those kind of sites. What is left? Maybe there's nothing left. I guess live performance.

Justin 27:08

yeah, robots performing, who knows, holograms, I don't know.

Rich 27:13

It's always robots.

Justin 27:15

This was a great conversation. Thank you, Rich. I know we've, kicked this around you know, catching up, outside of the podcast. It was good to sort of capture some of this conversation on the

Rich 27:26

Yeah. Everyone has such varying opinions on these things and different ideas of things. What are some of your opinions guys? Feel free to leave some comments on social media and let us know what you think or shoot us a message or just. Hang around and listen to the next episode. Thank you, Justin, for joining me today.

Justin 27:49

Yeah. Thank you, Rich. This was great. Really great to get your perspective as a singer and a voice teacher. a really good conversation and I look forward to continuing this, with our audience. So yeah, like you said, please let us know what you think of this episode. Let us know what your thoughts are on AI vocals. And with that, we'll wrap things up here. Thank you everyone for listening and we'll catch you in the next episode. Bye for now.