0:01

Indie musicians can now or

0:03

maybe go against DSPs.

0:07

And TikTok combats Ai. You're

0:09

listening to the biz Tape.

0:23

Welcome to the biz Tape. You're all Things

0:25

music, business and media podcast. I'm

0:27

your host, Colin McKay with my lovely

0:30

host Coast to Coast. I'm running out of times

0:32

I can say that, Joseph Wazelski.

0:35

It's true.

0:36

I guess I could adopt it in regular US,

0:39

but I feel like it would be really weird if I was

0:41

like, oh, my host Coast to Coast is calling

0:43

me.

0:44

I feel like you should change my name

0:46

and your phone to host Coast to

0:48

Coast.

0:49

Oh my god, you want to hear something funny? I forgot

0:51

to tell you this, speaking of our relationship

0:53

as a host to hosts like kind of thing. So

0:56

I was having coffee with some people, right,

0:58

and I was I,

1:01

you know, you're my business partner. Correct

1:04

you would say that's a correct thing, but sometimes

1:07

I like to say partner. So

1:10

I was talking to this person who

1:12

is part of the LGBT community and they

1:14

were like, I was like, yeah, me and my partner

1:16

have the show we do, and she was like, oh my god,

1:19

congratulations and I was like, look, we

1:21

would make a great couple together.

1:23

I'm sure we could get some Pogue magazine

1:26

cover kind of stuff.

1:27

Oh yeah, you can.

1:28

I I could. I could clean the dishes,

1:30

you could make thirty dishes. We would have a very

1:32

good relationship. But

1:36

it was like, I was like, I cannot. I

1:40

cannot hold a candle. I cannot hold a candle

1:42

to your partner. So you know, it's

1:45

just I you know, it was really

1:47

funny, and I was like, oh no, no, no, no. And I

1:49

was like, I you know, I'm not I'm like,

1:51

I'm not a fake ally, I'm just

1:54

just sadly straight. You know.

1:57

I can just imagine you and

2:00

in the middle of Portland brew So.

2:03

That was pretty close. Yeah, you're where

2:05

I was. Yeah, right, Oh

2:10

god, folks, Now

2:12

we're getting more unhinged and you're gonna probably

2:14

see it till the end of the show. And if you haven't heard, this

2:16

is our second to last episode of

2:18

the show, one forty nine, going

2:21

into one fifty. We're really at a higher number

2:23

than that, but fin my number

2:26

says, but you're

2:29

we're gonna get a little off the cuff in the last episode.

2:31

Just to give you guys some uh stuff kind

2:34

of maybe just some stories kind of

2:36

hit some week, get canceled Mike

2:40

as a fake ally yeah

2:42

no, but like yeah, it was just really funny and I was like,

2:44

but but seriously though, like you know

2:47

I am and how I So it was just like full

2:49

circle. I was like, you know, going with

2:51

that, but yeah, it was it

2:53

was just it made me laugh so hard. I was

2:55

just like, yeah, my partner Joe, he

2:58

like we worked together all the time.

3:00

Yeah, I love it. Anyway,

3:03

Joe intends to start with some of our last

3:06

stories that we're really going to cover on the show

3:08

before we get into our really crazy last

3:10

episode.

3:11

I feel like I can. I'll go

3:13

first. Mine's pretty, uh, pretty

3:15

cut and dry, but I feel like has some interesting

3:19

elements to it. It might be

3:21

dated, honestly, like after it

3:24

would be funny our last episode, like true

3:26

episode is TikTok that

3:31

yeah, but TikTok has currently

3:33

rolled out some tools

3:36

for labels when it comes

3:38

to labeling AI created

3:40

content and more seal

3:43

yeah and more specifically

3:46

in the music realm uh Ai

3:48

created content that's like sounds

3:51

pretty much. So there's a

3:53

lot that AI can do at this at

3:55

this point in the

3:57

the our generation are

4:00

our current timeframe. We're

4:02

at this culmination of like AI

4:04

could could forever change

4:06

the way humans

4:09

live their lives. At least that's what

4:11

Elon Musk is pushing

4:14

to some extent and a lot of these big

4:16

tech people. But whether

4:19

or not that's going to be true is hearsay.

4:22

But what I feel like everybody

4:24

can kind of get around is that

4:26

it is very true that AI can create

4:29

content that can mislead

4:31

you into thinking that a human

4:34

made it in a lot of cases,

4:36

especially if it's not tagged on

4:39

with created by AI. More

4:42

recently, Colin, did you see that

4:45

crazy TikTok that was borderline

4:49

racist that a

4:51

lot of people were seeing?

4:53

It was this creator who her whole

4:55

thing is that she face maps herself

4:59

and she does a transition and she changes into

5:01

like a new character that

5:03

she creates. Yeah,

5:06

and she she did a

5:08

little TikTok swipe

5:11

when it comes to like the transition shot,

5:14

and she turned into a

5:18

black woman who the

5:20

creator is originally a white woman. And so

5:22

a lot of people had a

5:25

lot of people were making

5:28

jokes. They were kind of like, is this like

5:30

the new era? Like

5:33

is this the new era of black face?

5:35

Like it just keeps going.

5:38

That's honestly, that is a major concern

5:40

with you know, with like AI and

5:42

specifically with the strikes that are going

5:44

on in la is how AI

5:47

could disrupt a lot of diversity

5:49

and casting, specifically on the actors

5:52

but also on the writers by being like, oh,

5:54

we have this AI that

5:56

can you know, create a story

5:59

or create a character and

6:01

it looks real and they're you know, black or

6:03

Hispanic or Asian or something like that. And that's

6:06

a real concern. So I don't think

6:08

it's you know, some people acting

6:10

you know, rightfully concerned about

6:12

it. It's true, like this is a concern

6:14

that unions are fighting over right now,

6:17

you know what I mean.

6:18

I think especially to like even you

6:20

know, I've heard a lot of talk about AI when

6:23

when it pertains to entertainment

6:25

specifically. But I also

6:27

think that there's a lot of potential

6:30

neglect being brought

6:33

on when it comes to political

6:36

ads and like political statements

6:38

and.

6:39

Oh yeah for that, like you

6:41

know, a fake you know, if there was a really

6:43

fake news your favorite politician. Literally

6:46

it's literally literally news.

6:48

Yeah, it's literally the fake news generator,

6:51

like and which Colin, I don't

6:53

know if you heard this. There's like a really interesting daily

6:55

episode about like how there

6:58

was this. It was like two

7:00

years ago they came out with this episode. I think it

7:02

was like during the election craziness

7:05

two or three years ago, and

7:08

it was about how this person gets

7:10

like hired out to just cut

7:13

memes together. And

7:16

now you don't even have to hire that person

7:18

as a political person.

7:20

You could just get an AI to cut a bunch of memes

7:22

together for you and then like spread

7:24

it everywhere. And to be honest, the way that social

7:27

media is now, it's quantity

7:29

over quality, and people still

7:32

consume it. So it's a

7:34

lot of like just post every day stuff

7:38

that can be throwaway and be terrible.

7:40

It can be fake, it can look

7:42

fake, and people will still think it's real. It

7:45

is kind of the crazy thing behind

7:47

it. But according to Billboard,

7:50

in July, President Biden's administration

7:52

announced that seven leading AI companies

7:55

made voluntary commitments quote

7:57

to help move toward a safe, secure,

8:00

and transparent development of AI technology,

8:03

and one key point quote the company's

8:05

commitment to developing robust technical

8:08

mechanisms to ensure that users

8:10

know when content is AI generated, such

8:12

as a water marking system,

8:15

because that can't get

8:19

this action enables creatively

8:21

with AI to flourish but reduce the

8:24

dangers of fraud and deception.

8:27

And so they're basically kind of

8:30

like testing. There's

8:32

a couple of things they're kind of testing when

8:35

it's like giving people the option to

8:37

put the stamp of like this is AI technology,

8:40

which I do think, like I feel like it's

8:42

so buzzworthy right now, and like it's getting

8:44

so much attention online that like people and creators

8:47

are doing it any ways to get that

8:49

like hype bubble and to

8:52

get a lot of views. But once

8:54

that kind of goes out of fashion,

8:57

like AI becomes normal and

8:59

people are like I don't

9:01

care that you

9:03

know, like telling if this is AI, Yeah,

9:06

it becomes like this unsafe territory

9:09

of like oh shit, we're

9:11

gonna like think that this is like a real thing. And

9:14

so like there's like a lot of research that's kind

9:16

of being put in to figure out

9:18

if there's even a way to use

9:20

AI, to to find

9:23

out if something was created

9:26

with AI, and to like kind

9:28

of funnel that in and self police

9:31

in some ways which

9:33

have been proven to be useful,

9:36

especially when it comes to people

9:40

in schools, which

9:43

if you don't know if you have chat GPT write

9:45

your paper for you, unless you're very crafty

9:48

in having chat GPT write it. A

9:51

lot of the time it's still going to get

9:53

caught with their own

9:55

AI generated thing because guess what

9:57

that thing is using a

10:00

shat GPT.

10:02

So it's every time I hear

10:05

that, it just reminds me of like, what was it like

10:07

turning in dot com you know that kind

10:09

of stuff.

10:09

Well, that's what it is. I think. It's like it's

10:12

it's finding the percentage chance

10:14

that like this was created with

10:16

AI.

10:17

I mean, if you think about it, it's very similar to the

10:19

same technology where it's like where

10:24

we're basically it says, you know, hey,

10:26

we found this other piece

10:28

of literature that you plagiarize,

10:31

you know what I mean. So it's it's not really that far

10:34

I mean, it's a new technology. AI

10:36

is very new, but it's like at the same time, it's like the

10:38

concept is not that far off. It's the same reason you

10:40

would get caught for plagiarizing, you know what I mean.

10:43

Yeah, so yeah, exactly, yeah,

10:45

but Billboard continues. Voluntary

10:47

commitments are of course voluntary,

10:50

which is likely why TikTok also announced

10:53

that it will quote begin testing an AI

10:55

generated label that we eventually

10:57

plan to apply automatically to content

11:00

that we detect was edited or

11:02

created with AI. Sure tools

11:05

to determine whether an image has been crafted

11:07

by AI already exists and

11:09

some are better than others. But

11:11

in the June version of The New York Times

11:14

tested five programs, finding that

11:17

services are advancing rapidly but at times

11:19

falls short. So basically what's

11:22

happening is like, because it's

11:24

started so late this

11:27

because like we're already like almost

11:29

a year into AI, right,

11:32

I think a year like it's

11:34

it's it's been out, it's been it's

11:37

been AI version one

11:39

eight twenty eight thousand,

11:42

you know, at this point, and it's it's continuously

11:44

getting smarter and smarter. And so because

11:47

of that, a lot of these programs

11:49

that are starting at the at the base

11:52

ground or having a hard time like catching up

11:55

to its AI competitors because

11:58

there's so much information that it still has to process

12:00

through and it just hasn't had the time to

12:03

do it. And also of course people are shifty

12:05

and can get past it.

12:07

Speaking which can I can I interrupt

12:09

you for a second, Joe, just to see something that I've

12:11

been doing this whole time while we're talking that you're gonna

12:14

laugh at

12:16

the thumbnail of this podcast episode

12:18

is going to be generated AI. Hell

12:21

yeah, because I got to lose

12:23

anymore. I did it. I generated, but

12:25

I did music business podcasts

12:28

and I whatever the first one was. I was like, I'll do

12:30

it as long as it's not, you know, crazy offensive.

12:32

So I just think it'd be funny if you've been wondering this

12:35

whole time listening to the show, what the hell

12:37

is that? That's what it is? Anyway?

12:39

Hell are you going? Hell yeah,

12:42

I've I've actually, I mean we got

12:44

into here. You know what we're gonna talk about it

12:46

call it Colin. We almost got canceled

12:49

like weeks

12:51

ago because I I

12:54

stupidly made like I threw

12:56

in some of our video clips into like

12:58

an AI program and it chopped

13:00

it up and like it to its credit,

13:03

it got us to go viral. Like

13:05

it found those like key

13:07

moments and an hour of footage that I didn't

13:10

have time to cut

13:12

together, and I posted

13:14

to TikTok and reels and on reels

13:16

it just exploded. We

13:19

got to almost a million views. Yea,

13:22

within like I

13:25

think what was it. It

13:27

was like a million views in like

13:29

like three days or something, and

13:32

but we were getting the

13:35

way the cut happened, it made

13:37

it seem like we were making fun of gen Z, which

13:39

in the context of watching it,

13:42

it does kind of seem like we are, like,

13:44

we're just two dudes like making fun of gen Z. It

13:47

cut out. Yeah, what it cut out

13:49

was the context of like, oh,

13:52

we're actually like in agreement

13:55

with like gen Z, like not drinking as

13:57

much like in these like bars and stuff.

14:00

The entire time, I'm going like, oh

14:02

man, I wish I could drink. I'm

14:04

so cool, and people are.

14:06

Like, yeah, people took it very People

14:08

were like Colin was bullying gen

14:11

Z.

14:11

They were like, Colin is a you

14:14

know, like.

14:15

You called them nerds

14:18

ironically, and I've never heard

14:21

anyone called nerds. Anyone

14:24

called anybody nerds un

14:26

ironically, especially

14:28

online. But people believed

14:30

it. And that's that is

14:33

kind of to to what we're talking about

14:35

here, Uh, kind

14:37

of the toss up of like your voice

14:39

can get diluted when it

14:41

comes to these AI technologies

14:44

because they're so I mean,

14:46

they're making choices on

14:49

your behalf right inside. How they're

14:51

editing something I'm.

14:52

Gonna put this, uh this

14:54

ai this thing I just generated

14:56

into the chat, Joe, so you can see it, so you can

14:58

see what the audience is seeing. There's

15:01

a lot going on.

15:02

So have you been doing this for like weeks?

15:04

No, I just did this right now while we were talking.

15:08

Oh oh, this is you brought up thought

15:10

this was the funniest thing.

15:12

I was like this, wait, why

15:14

didn't we make this? We're gonna art? Can

15:16

we make this her?

15:19

Yeah, I I are episode

15:22

absolutely, I'm gonna do that. I don't know what that thing

15:24

is in the middle. It looks like some sort

15:27

of antenna or.

15:28

Like there's a lot of pop filters

15:30

happening.

15:31

There's a lot of pop filtering. There's clearly

15:33

a guitar that's been thrown in there. This

15:35

is very like Eastern

15:38

European art.

15:39

Like, yeah, I love it, though, I would,

15:42

honestly I want this on a T shirt. This

15:45

looks amazing.

15:47

Oh god, Joe, God, And

15:49

this is the power of it?

15:50

What is it? One one percentage

15:53

or one half of dot?

15:57

Yeah? Right, no,

15:59

get it? Like but seriously, we've been talking

16:01

about it the whole time, especially with TikTok. It

16:03

does take It's a lot easier

16:06

to generate than it is to

16:08

find tools to you know, find

16:10

what generation is right, because you got

16:12

to think about it. That's like an extra step. It's like

16:14

you have to understand how this stuff

16:17

is kind of generated, and then on top of

16:19

that, you have to understand how to

16:21

identify those things that are generated.

16:23

So I'm really glad to see that TikTok

16:26

is taking some

16:28

steps to try to you know,

16:30

filter the content and go this is AI

16:33

generated, this is AI generated, this is AI generated,

16:35

because it's seriously going to be a problem

16:38

as the technology gets better. Now I'm sure

16:40

some of you are laughing at the greatest

16:42

thumbnail that's ever been made, but like it's

16:45

you know, a lot of this AI stuff,

16:48

you know, they're like it's made with AI, and

16:51

like there's two things to that one

16:53

it's and this is honestly, if

16:55

you one of my favorite books, the Toyota Production Method

16:59

is uh. Automation with a human

17:01

touch is a principle in that book for efficiency,

17:04

and that's very similar to AI, where this

17:06

is automate. This is automation, like

17:08

it's automatically making this thing

17:10

right, and then with

17:12

a human touch you can make AI look

17:15

incredibly realistic. Like I'm sure if

17:17

me and Joe went into this image and

17:20

we you know, fixed up the words and

17:22

put the percentage and maybe took the logos out

17:24

and the you know, they'd be like, oh, this is just like

17:26

some art you know what I mean that like they

17:28

had made or commissioned or something,

17:31

But it's AI generated and the same thing.

17:33

Obviously the bar gets higher with video,

17:36

but like the same thing happens is

17:38

that a lot of this AI stuff is also

17:42

you know, one it's getting to a point where it's

17:44

just it's good enough to be on its own. And

17:46

two it is like with a human touch,

17:48

like a lot of people because you think about it,

17:51

like people generate these AI content

17:53

and they generate and they generate and they generate,

17:55

and then the human touch sometimes is just

17:57

as simple as that one's the best one,

18:00

you know what I mean. Like so

18:03

that that's kind of the thing with AIS. It's got a long

18:05

way to go, but with

18:07

human touch, this can

18:09

really be used to deceive people. This

18:11

can be used to literally create fake news,

18:14

as Joe was saying, and so like

18:16

it's definitely something that I think needs

18:19

to have some system to flag

18:21

or filter, not in a negative way. You know, it's

18:23

like this is innovation at the end of the day,

18:26

but at least say, hey, this is AI, you

18:28

know what I mean. So people that

18:30

use social media as their primary news

18:33

source, which a lot of people do, don't

18:35

get confused about something or

18:37

worse, you know, like we could have like

18:39

you're saying political ramifications

18:42

from this because everybody thought, you know, the

18:44

President said something or like some other

18:46

you know, legal officials said something and it

18:48

wasn't true, you know what I mean. So

18:51

and you have to think about it this too. It's like

18:55

AI is its own fledgling

18:58

interests in like industry,

19:01

and everyone's excited about it, but

19:03

the amount of experts in it are few

19:05

and far between. And just

19:08

you guys listening to the show and us having

19:10

interest in talking about AI has

19:12

already put you ahead of so many other

19:14

people, you know what I mean, that don't even think

19:16

about that AI is out in the landscape.

19:19

So that's why it's really dangerous,

19:21

is that there's a lot of people that haven't even

19:24

factored in AI as a thing

19:26

in the same way that like if you look up, you

19:28

know, everything goes back to school for me

19:30

today for some reason, like if you're looking up sources

19:33

on the internet for like a paper, you're like,

19:35

that's a non credible source, you

19:37

know what I mean. Like people you know,

19:39

have just been looking at stuff on TikTok

19:42

regularly before AI and going, this

19:44

is the gospel, this is true, this is absolutely

19:47

true. And now with AI, it's like

19:49

we can make that gospel say whatever it wants, you

19:51

know what I mean. So I think

19:53

it's interesting. I do think that we're

19:55

gonna see people one

19:59

who are you know, trying to be maleficent

20:01

with it, but also people that maybe

20:05

fearing that this will stifle kind

20:07

of AI content, you know, maybe

20:09

go into suppression. But I think

20:11

that this is a tool that's extremely powerful.

20:14

So the you know, the question of how much

20:16

or how little we use needs

20:18

to be answered, you know what I mean. So

20:20

yeah, and that's that's a question for

20:22

everybody, you know what I mean. Like then it could come

20:24

down to context, it could come down to what's

20:27

going on in life. You know, there's a lot of things going

20:29

on with that. So I don't know, right,

20:32

like it's.

20:33

Yeah, I mean, it's it's such

20:35

a it's

20:38

it's like it reminds

20:40

me so much to some extent

20:43

on the hype that was n FTS,

20:46

and I feel like they kind of bled into each

20:48

other.

20:48

Yeah a little bit.

20:49

Okay, Like with any tech thing, there's like,

20:52

you know, like the big tech things

20:54

that happened in our lifetime were

20:56

streaming, and then it

20:59

goes into bigcoin, and then

21:01

it goes into NFTs,

21:04

and now it's going into AI.

21:07

But I think what's different about AI in

21:09

terms of those other ones, and what makes it more

21:11

powerful is because of the

21:13

ramifications and the real ramifications

21:16

that it has on how it interprets mass

21:19

data and like how it like

21:22

collectively shapes

21:26

not only shapes, but can can build on

21:29

itself. It can like it

21:31

has its own I

21:33

don't want to say personality, but brain

21:36

right that it can. It can stack up.

21:38

And like I think, I

21:40

think like the hype behind those other things

21:42

that I just listed off, you

21:44

know it did it definitely seemed more like a

21:47

fad, right, Like

21:49

this is a fad. This isn't actually going to like help

21:52

the A lot of people with crypto

21:54

actually thought it would help like a bunch

21:56

of people, and like.

21:58

It didn't have you know, it didn't have a really

22:01

innate value, you know what I mean, Like.

22:03

It didn't have a it didn't have an everyday

22:05

value.

22:06

Well in the sense

22:08

that what created value for

22:11

it was that people believed it had

22:13

value as opposed to having something

22:15

intrinsic at the end of the day like not.

22:18

And I don't believe in this whatsoever because

22:20

this is like the argument people used to go back

22:22

to the gold standard, which should never come

22:24

back ever when it comes to finance. But like

22:27

at least at the end of it, like gold, like

22:29

you know, it could be used in electronics

22:32

a lot, you know what I mean, Like it has a very high

22:34

usage. And the crypto it's like, yeah, there's

22:36

nothing that this does except takeaway

22:38

resources and everything

22:40

like that, and AI like you're saying, has

22:42

an intended use. So I think in

22:44

suppression when it comes to TikTok it,

22:47

you know, it's we gotta This

22:49

is a very careful walk

22:52

through a minefield that I hope

22:54

that we're prepared for as a society to walk,

22:56

you know what I mean. This isn't like we've

22:58

solved it, we're just gonna ban everything

23:01

or like put a label on everything immediately.

23:03

It's just more of like, well,

23:05

we need to look at this and see

23:07

how we walk through this minefield and

23:09

how we assess this kind of information and

23:12

like the systems that we have, and like

23:14

what as a kind of free

23:16

market society we want to put constraints

23:18

on because it's a it's a very dangerous

23:21

you know, piece of technology in the wrong hands,

23:24

So we need to watch out for that explicitly

23:27

at the end of the day. So I

23:29

think, yeah, it's something

23:31

that I love your point

23:34

about technology kind of blending together because

23:36

there's things that we don't think about that really

23:38

like it all blends together. And then we're here

23:40

and we kind of have to take a step to be like, no,

23:43

we are at this step, right, we have to take

23:45

ourselves out of it and be like, we're at this step with

23:47

AI, so we have to figure it out.

23:49

Like one that came to mind when you kind of went down that argument

23:52

for me was the dot com

23:54

boom to like social media. Yeah,

23:57

I you know, Joe and I

23:59

for this example, are blessed to have older

24:02

siblings, so like we understand

24:05

what that kind of was different. I mean,

24:07

the dot com boom was like websites

24:09

are out there, you can use them for services,

24:12

you know what I mean, that's amazing and

24:14

it made the stock market go insane until it

24:16

eventually crashed because people were just getting

24:19

websites to have websites that didn't have a

24:21

use or a service ie crypto, So

24:25

that happened, and then that was like

24:27

around two thousand and then like

24:29

right around there is when social media really

24:32

started to come off. And if you talk to people

24:34

that are around our age that

24:36

don't have a lot of older siblings, that aren't,

24:38

you know, very historical about technology,

24:40

they think all of that blends together,

24:43

right, that we had websites, social media,

24:45

you know, and it's all just happened at once, But that

24:47

was one step we took, and

24:50

you know, the ramifications of having

24:52

websites that we could get services

24:54

through and you know, good services, vital

24:57

services, right, good things like we hadn't even

24:59

fully exp war that yet. And then it goes, here's

25:01

social media that has the power to

25:03

socially change how we talk to each other

25:05

and whatever we're and then I'm like, wait, we're still

25:07

talking about dot com. And then it's like here

25:10

is streaming music, media

25:12

whatever, and it's like we're not even on the dot com

25:14

thing. Stop doing this, and

25:16

here is cryptocurrens and you're just

25:18

like, we need a minute to talk about this,

25:21

right, So, I think it's one of those things

25:23

where you can't stop

25:25

it. And like anytime any

25:28

of these social media companies or the you know, the

25:30

government are doing this, we're already

25:32

behind. So there kind of has to be some urgency

25:34

to this, right, There has to be some urgency

25:36

to like, what are we doing, how can we do it?

25:39

Let's do it. But at the same time, I

25:41

did say it's a careful minefield, which is what makes

25:43

this all freaking crazy.

25:46

Yeah, so, yeah, I mean it's it's

25:48

truly the wild West.

25:49

I'm gonna take my tinfoil hat off now, No.

25:54

That's great, it's great,

25:56

But yeah, I mean TikTok's like

25:59

the only company that's also rolling out plans.

26:01

Streaming service Deezer has also

26:03

laid out their plans to develop

26:06

some tools to grab them.

26:09

And according to Billboard, which

26:11

is also another quote a quote from an economic

26:14

point of view, what matters most is regulating

26:16

the things that really go viral, and

26:19

usually those are the AI generated songs

26:21

that use fake voices or copied

26:23

voices without approval. And

26:26

that's a quote from CEO of

26:28

Deezer, Geronimo

26:32

Bulgaria, and

26:35

according to Dennis

26:38

like this is a crazy name. I'm just

26:40

gonna say, Dennis L who

26:43

is the Believe co founder and CEO,

26:46

quote, you have technologies out

26:48

there in the market today that can detect an

26:50

AI generated track with a

26:52

ninety nine point nine percent accuracy

26:55

versus a human created

26:58

track. So he

27:00

goes on, quote, we need to finalize the

27:02

testing, we need to deploy, but

27:04

these technologies exist. And

27:07

then someone else pointed out

27:10

that there's definitely, definitely

27:13

like a lot of kind of countering

27:15

points. Especially

27:18

Cynthia run In from Duke University

27:21

said that quote, every time somebody builds

27:23

a better generator, people build

27:25

better discriminators, and then people

27:28

use the better discriminator

27:30

to build a better generator. So

27:33

the generators are designed to be able to

27:35

fool a detector. So I'm

27:37

gonna I don't know about you, Colin, I'm going to listen

27:40

to the Duke University person over the

27:43

Believe CEO.

27:44

Again, it's all based on each other, right,

27:46

I mean this is why at the end of the day,

27:49

like it's very similar to the field

27:51

and the questions that cybersecurity deals

27:53

with all the time, like where our passwords

27:56

are, where our data is. That's what they do

27:58

is they make systems to pretch data.

28:00

Something breaks the system, and guess

28:02

what they do. They examine that

28:05

thing that broke their system, and then they adapt

28:07

and it's better. That's just how

28:09

one life kind of works. But too specifically

28:11

technology works. The problem with technology,

28:14

and especially digital technology is it's insanely

28:17

fast. So if you get

28:19

a whole you know, if TikTok

28:22

or any of these social media companies make this

28:24

whole system where it's like this is how

28:26

we identify AI and this

28:28

is how we label it, and then like literally it's like

28:30

by the time you put the system out, it's already

28:32

gone. Now people use that

28:36

in a bad way, Like people

28:38

say, then it's not even worth trying, and

28:41

that's not true. Right, This is a

28:43

constant battle, and this is definitely

28:46

something that I think, you know, social media

28:48

companies with the content that

28:50

you know, users create that is not

28:53

exactly cultivated by TikTok

28:56

will have to have literally departments

28:58

to deal with, right, We'll have to.

29:01

And I hope that you know all of them are like we're making

29:03

these investments now, but every one of them is like we're

29:05

making these investments now. So we can make more content

29:07

and make more money. And it's like at the same time,

29:10

it's like, well, you need to be getting these people

29:12

in there, and so you guys can do it differentiate

29:14

this content and so you don't accidentally like

29:17

sway people into you know, crazy

29:20

ideologies or thinking that

29:22

the world's gonna end in two days or something

29:24

like that, Right, And that's kind of the power. And

29:26

to be honest, it can be doomsday like that, or

29:28

it could be something as little as like you

29:31

know, in our show with music business, like imagine

29:33

an artist you like and then they come out

29:35

as an AI and they say like something super

29:38

you know, kind of controversial to their

29:40

fan base even you know, and they're

29:42

just like I didn't say that, and AI said

29:44

that, and there was nothing I could do about it. Right, You

29:47

basically lose the ability to control your

29:49

own pr which is kind of insane.

29:52

Yeah, it's it's

29:55

just so crazy,

29:59

honest, Like, I don't know.

30:03

I don't know, that's beyond the scope for the show. That's why

30:05

we're that's where we're ending at. We were just too

30:08

busy thinking about AI all the

30:10

time.

30:10

So Colin,

30:14

what do you have all

30:17

right?

30:17

Well, speaking sort of about AI,

30:19

but not really. We're talking about a

30:21

proposed US bill which is

30:24

very interesting as this could allow indie

30:26

artists to negotiate collectively with streaming

30:28

services, which I was like, hello,

30:31

you know this is casually going by my newsfeed.

30:34

So this bill is called the Protect Working

30:36

Musicians Act of twenty twenty three. It

30:39

was introduced on September nineteenth, so that was

30:41

where we go from today. And basically

30:44

the House wrap Deborah Ross, a

30:46

New North Carolina Democrat, updates

30:48

an earlier bill that was introduced

30:51

by Representative Ted Duluth,

30:53

a Florida Democrat, and that bill basically

30:55

focuses on the relationship between indie

30:57

artists and music streaming services. So

31:01

basically what this allows them to do,

31:03

and like AI companies will get

31:05

to in a minute, but it allows these

31:08

musicians to work together

31:11

as indie artists to collectively

31:13

bargain against streaming services

31:15

for these rates.

31:16

Yes.

31:17

So the reason why that's important is

31:19

because most indie artists use

31:22

like an independent distributor like a

31:24

tune Core or a distro Kid or

31:27

CD baby for instance, and those

31:29

people negotiate licensing deals with

31:31

the DSPs. Some even

31:34

do like you know, a digital Rights

31:36

Music Licensing Network. If you've ever heard of Merlin,

31:38

it's a little bit different. But the point is is, like there's

31:41

another company that's already made this agreement

31:43

with Spotify, with Apple whatever, and

31:45

their rates are their rates are their rates right, and

31:49

so like they can't really do much here. And

31:51

the bonus of a label, and

31:54

honestly, one of the biggest bonuses of a label

31:56

is they collectively bargain, you

31:58

know, and because labels, if you

32:00

think about it,

32:03

basically act as a union of themselves,

32:05

right they go. We have, you know,

32:07

hundreds of thousands of artists on

32:10

our roster. We have you

32:12

know, we'll probably get the next big

32:14

people because that's what we do. We promote

32:16

the next big people. So give us

32:18

a deal Spotify, give us a deal,

32:20

Apple Music, give us a deal Amazon.

32:23

And that can be in like promotion, actual

32:25

payment, upfront payment, you know

32:27

what I mean. We've talked about on the show, like Drake

32:30

was the most streamed artist of last year, Like the just

32:32

straight payments he would get right that

32:34

were just like you're a big artist, here's

32:36

like one hundred thousand dollars, you know what I mean, Like

32:38

not like, oh, you're paid for it. Was just straight

32:41

up from Spotify. The point is

32:43

is that this bill basically lets

32:45

you collectively argue on that

32:48

together, and it creates an exemption to

32:50

the US anti trust laws that

32:52

would allow them to negotiate licensing

32:54

deals with a quote dominant

32:56

online music distribution platform

32:59

or a company engaged in development

33:02

of generative

33:05

artificial intelligence. Isn't

33:07

that interesting? So the

33:10

bill basically defines quote a dominant

33:12

online music distribution platform

33:14

as an entity that operates an app,

33:17

website, or other online service that

33:19

is used by members of the public to listen to sound

33:21

recordings via digital audio

33:23

transmission and audio visual presentation

33:25

or any other means basically being like, there's

33:28

no way you can, you know, argue like Spotify

33:30

tomorrow can't be like we only do CDs

33:32

now like it's like nope, you're you're doing

33:34

it. And the main

33:36

caveat for these online music

33:38

distribution platform like what

33:41

fits in it and what does it is just

33:44

revenue. Mostly, it's saying if you

33:46

make annual revenues

33:48

of over one hundred million dollars or

33:50

more, then you can be

33:53

you know, collectively bargained with under

33:55

these you know, artists, and

33:57

these artists basically can come together. And what

34:00

what's pretty kind of cool about it and

34:03

I kind of interesting to me is

34:05

they can do it under one lawyer as

34:08

like a big class action lawsuit.

34:10

Basically, so like imagine,

34:13

if you will, like

34:16

we have like imagine like all

34:18

these scenes of music, like Nashville has

34:20

a scene, Atlanta has a scene, La has a scene,

34:22

New York has a scene. You know, I'm sure Oklahoma has a

34:24

scene, right, Like imagine

34:27

okay, see okay scene yeah

34:29

right, But like seriously, like imagine

34:32

like all of them, you know, they go to shows together

34:34

to do this. They're just all onside and they were waiting for

34:36

a record label and they just turn to each other and be like,

34:38

well, I know you, you know me, we know each

34:40

other, we know everybody else in this town. Why don't

34:42

we just go get a lawyer together, and

34:44

let's go get us some good rates on

34:47

Spotify.

34:47

You know what I mean, You're going to recite the Barney

34:50

song.

34:51

I Love you, you love me?

34:53

You know, but like, yeah, it's

34:56

it's basically that's kind of what it is

34:58

is that they can collectively bargain again them

35:00

as like one big thing to try

35:03

to get you know, basically

35:06

better rates and privileges that

35:08

the labels usually use as their like defining

35:11

thing, you know what I mean. But

35:15

yeah, it's kind of it's kind of crazy, and

35:18

so I kind of wanted to talk about it a little bit.

35:20

Is like, with these kind of

35:22

tools, do you think this weekends labels

35:24

in general? You know what I mean? Like, this

35:27

feels like what a label does collectively

35:29

bark it. So do you think this, you know, weakens

35:32

people, you know, wanting to maybe

35:34

be on a label even more than they have been in

35:36

recent years.

35:38

You know, I don't. I don't

35:40

want to say yes because I

35:42

think I think that there's my

35:45

gut is telling me that, Like, honestly, the thing

35:47

that labels have up on pretty much

35:49

everybody in the industry is their grandfathered

35:52

in in a lot of

35:54

ways. They have long standing relationships,

35:56

They have long they have a lot

35:58

of money that is shaved and pooled

36:01

to do a lot of things, and a lot of brand

36:05

awareness from not only the general

36:07

public but from

36:10

the legal space as well.

36:13

And industry

36:15

relationships too.

36:17

I do think that, Yeah, I do think

36:19

that, like, labels

36:21

have been losing power in a lot

36:23

of ways. I think people have been saying no

36:26

to label deals a lot more recently

36:29

because you

36:31

know, a lot of the stuff you can do yourself. There

36:33

there are certain areas like, especially

36:36

when you get to be a certain point

36:38

as an artist, you do kind

36:40

of want that signing

36:42

bonus. That's kind of like where people

36:44

get like, Okay, I'm ready to

36:46

have this good signing bonus right

36:49

now, right to help

36:51

out with life stuff. And

36:53

I think that that helps.

36:56

And also having like a team of people like you

36:59

know, not not every label

37:02

is the

37:05

boogeyman, you know, but

37:10

the label system

37:12

is changing, and I will give you this,

37:15

I do think that this will change the ways

37:17

labels operate. I don't

37:19

know if it's going to take power fully

37:22

away from now.

37:24

I think that's that's a very astute point because

37:26

there's one thing I also haven't mentioned here actually,

37:29

So this is at the same time

37:32

that apparently major record companies

37:34

are trying to negotiate with streaming services

37:36

for better payment models aka collective

37:39

bargaining. But one

37:41

that's interesting here is that they

37:44

want this new model that I've seen.

37:46

It's called that they have like two groups

37:48

of basically royalty groups, and one is

37:51

professional artists, and

37:53

they're defined as having a minimum of thousand

37:55

streams per month or a minimum of five hundred

37:58

unique listeners. And so they

38:01

want the differential between professional

38:03

and non professional, who are people that don't

38:05

fit that you know, catalyst of being

38:08

over a thousand streams a month and five unique

38:10

whisteners. They want a

38:12

higher rate for the professional artists,

38:14

and then they want to lower the rate

38:17

for the non professional basically,

38:20

which is interesting for a

38:22

couple reasons. One, we we had a whole story

38:24

on the show one of when I was talking

38:26

about that, I think it's close to

38:28

it was like close to like eighty or ninety percent of

38:31

musicians have like less than like

38:33

a thousand or to

38:36

one hundred, like or a thousand

38:38

streams per month, right, And

38:40

so it makes sense that it's

38:42

like in the way that

38:45

streaming is calculated, it's

38:47

calculated as a pie. Like

38:50

basically it says like it's

38:52

not just like you get this money right away,

38:54

it's calculated in big pie percentages.

38:57

So basically saying like, you know, if

38:59

Spotify may five hundred million dollars in revenue,

39:02

they go to the labels first, and they'll be like, okay, Universal

39:05

Music Group, you were five hundred, you

39:07

know you were would say half

39:10

even though maybe that's not true. But they're the biggest

39:12

label in the world, Universal Music Group. You were

39:14

half of the streams from Spotify. We're

39:17

gonna give you two hundred and fifty million dollars,

39:20

okay, and there

39:22

you go. And then it's like, now we're going to this. There's

39:25

a lot of computation and like different stuff

39:27

in there. I know I'm ignoring. And then Universal

39:29

goes the same way and they go, Okay, Drake

39:32

was let's say something crazy

39:34

was forty percent or fifty percent

39:37

of our basis. So then it goes to like one

39:39

twenty five million to Drake, right,

39:42

then they cut out based on his deal, they

39:44

take the ninety, they take it, Drake gets like the

39:46

ten to twenty. Then Drake makes like

39:49

like ten million dollars, right, So that's

39:53

how that works. So it's interesting because it's like

39:55

if you take out the other ones,

39:57

right, like you're you're making the pie

39:59

bigger from the beginning, right,

40:02

You're making like going like give us

40:04

And they literally call it a double boost

40:07

to royalty payments. And one

40:09

thing that's beneficial for you

40:12

know, professional kind of artists

40:15

quote unquote, is that you get these double boosts and

40:17

who's maybe gonna have continual streams,

40:19

continual listeners, which you notice

40:22

it's not based on total it's based on

40:24

like a continual people that have teams

40:27

who are probably signed to labels. So like, I

40:29

think that's kind of the idea, is that

40:31

in a way, it would, you know,

40:33

take this money and put it more towards

40:36

people that have this continuous

40:38

career who are relying on

40:40

it. But I think it's gonna make it way,

40:42

It would make it way harder to come up, you

40:44

know, via royal royalty payments. Because

40:46

we joke on the show all the time out royalty payments

40:48

are a joke. But at the same time, I'm like, there's

40:51

something after a while, there's something, you

40:53

know what I mean, And so it

40:56

would just be interesting, you know for these people

40:58

that get just like a couple bucks, you

41:01

know what I mean. It's not crazy, but yeah,

41:04

that's I thought that was interesting that you brought that

41:06

up though, like the collectiveness of it

41:08

and kind of like how we could have incentives.

41:10

And I think that's why they, meaning

41:14

basically the big three labels are

41:16

really campaigning for that kind of model because

41:18

there's kind of some complex math there and it doesn't

41:21

make it feel like like from

41:23

a political standpoint, it doesn't make it feel like,

41:25

oh, we're taking money from other people. It's like real

41:28

artists get paid more. Are you against

41:30

that? You know what I mean? So

41:33

yeah, I thought that was very stute.

41:37

Yeah, I man,

41:40

we'd really just as like,

41:42

collectively as the music industry, we really

41:44

just love cannibalizing ourselves.

41:47

Oh absolutely when it comes to

41:49

like which I guess it's not totally

41:51

like streaming. I mean we could go back

41:53

to like the day one argument of streaming of

41:56

like, you know what you.

41:58

Could do instead of being like you get when

42:00

you get one, just raise the fucking rate.

42:03

Yeah.

42:04

Yeah, like that's kind of that today.

42:06

To do that, we need a union and the music

42:08

industry doesn't.

42:10

Want you mean, you need to collectively bargain.

42:12

Yeah, but it could be interesting. I think

42:14

this would be really interesting if this gets through.

42:17

Again, like it's notable because

42:20

I just thought it was a very interesting argument to

42:22

make. But I don't know if it's gonna get

42:24

through. I mean, one of the congressmen on this set.

42:26

As you've seen in Congress, lots of bills aren't passing,

42:28

like the budget, so it

42:31

might be hard but basically, he says, this

42:34

has been a very bipartisan issue in the Judiciary

42:36

Committee. It's the perfect time to bring these

42:38

issues up. So maybe this could be like, you

42:40

know, if Congress is getting crapped on from

42:43

like the American people because they're not doing their job,

42:45

they could be like, we did something. Everybody's

42:48

from paying more musicians, you know what I mean,

42:50

Like we did something. But uh, yeah,

42:53

it's good. It'd be interesting because I feel like there's

42:55

some labels that would probably be like, no, don't do

42:57

this, because then that deincentivizes

43:00

labels, you know, to their Congress friends. But

43:02

like at the same time, it's

43:04

interesting to see kind of

43:06

this argument be made, and I like, I

43:09

just imagine and I would just like to see

43:11

it to see it. Like imagine like

43:14

one giant company

43:16

decides like, hey,

43:20

everybody that's an independent musician,

43:22

hit us up if you want to be part of this collective

43:25

thing, and like thousands and thousands

43:27

of musicians joined together under

43:30

this like being of like a

43:32

lawyer or two, and they're going like, now we will

43:34

collectively bargain as the thousands and thousands,

43:37

you know what I mean. I would love to see

43:39

like the independence of that, though,

43:41

I am worried about, you know, like the business

43:43

model of it, because being like, we

43:46

could just do it. It's like ten people, you

43:48

know what I mean, And you could pay me forty thousand

43:50

dollars as a lawyer, and then instead of making two hundred

43:52

dollars a year, you make four hundred, you know what I

43:54

mean, except you still owe me forty

43:57

thousand dollars. That's where I worry

43:59

about this law. But anyway,

44:04

Joe, I guess I don't know what we'll

44:06

do for the last episode. This is maybe the last time. What

44:08

have you been listening to?

44:09

Yeah, let's let's I

44:12

feel like we could do it for the last episode. But

44:16

I haven't listened yet,

44:18

but I will. But Devendra Banhart just

44:20

released a new record. It is

44:22

so beautiful. It's called Flying

44:25

Wig. I have actually listened to it because

44:27

I've was honored enough to work

44:31

on this project and I

44:33

am very proud with a lot of the stuff that

44:35

we've created, and

44:38

such an amazing team in

44:41

such an amazing like songwriter

44:44

and just being

44:46

Devendra is like super super

44:48

sweet human. So please check

44:51

out his newest record. It's

44:53

called Flying Wig, available

44:56

everywhere. But I've also been

44:58

listening to. I listened to that. You choice

45:00

of von song, which please

45:02

tell me you listen to that?

45:04

No, oh my god, yours right now

45:06

and see what I think.

45:06

Dude. Okay, I'm not going

45:09

to tell you anything, but you have to listen to

45:11

it purely for the sample.

45:13

Purely for this is the newest song.

45:16

It's the newest one. It's

45:20

called got Me started, got

45:22

me and started. Okay, well yeah.

45:27

Somebody. Yeah, I'm gonna look at somebody is gonna

45:29

be like, are you kidding me? Colin?

45:32

You don't know choice Evon's whole catalog. I'm like,

45:34

I'm sorry, man, all right, Oh

45:36

my god? Is this really a starstruck?

45:39

Is that what that is? Yeah, it's just it's

45:43

it's the Yeah,

45:46

I can't remember what that is. It's like because

45:48

it's god, I know what

45:50

this sample is.

45:52

I mean, I can just add that new Google

45:55

that Google voice search thing.

45:58

Oh what's the sample?

46:00

Yeah, hold on, let's see it.

46:02

No, I gotta I got it. It's shooting stars

46:04

that's what it is. I was like, it's something like starship

46:07

or something. Yeah, that's

46:09

so.

46:09

Funny, Cobra

46:11

Starship? Was that what you're It's

46:15

what did you say Cobra Starship.

46:17

No, I was not thinking of Cobra Starship.

46:21

God. I might have been thinking of the band

46:23

Jefferson Starship and then later Starship.

46:25

But oh nice, let's see

46:28

thing of old bands as I mean, that's

46:30

half my thing on the show. We've made one hundred and fifty

46:32

episodes. I'm sure if we averaged them

46:34

out, it probably wouldn't be later than like nineteen

46:36

ninety ninety percent of the time. But

46:40

basically, we recorded

46:43

yesterday a little bit for one episode,

46:45

and I for or two days ago,

46:47

and then I forgot that. I got to work the next

46:49

day and I was like, oh my god, today's

46:51

the day and it's the twenty

46:54

first of September. Baby. So I've been listening

46:56

to that. Yeah,

46:58

I mean that song holds up and incredibly

47:01

well, like like well all of

47:03

earth Wind and Virus Catalog does. I mean,

47:05

it's like one of the few that of

47:09

like the disco era that like

47:11

really holds up. Like you can play

47:14

this song at like a wedding or

47:16

like, you know, and with regular people around

47:19

and people aren't like, what the fuck is this? You

47:21

know what I mean? So

47:23

I was listening to it.

47:25

It's like one of my favorites. I

47:27

think I talked on the show. I mean, we've been doing

47:30

it for three years, and I'm sure I made the September joke,

47:32

but like there was a guy I used to do videos

47:35

every year for the like song, and like

47:37

his whole like joke was like

47:40

today's the day like kind of

47:42

joke, and he would do it literally

47:44

every year, and then he stopped doing it. So I was really sad

47:46

about it, but I did I

47:48

did have some interesting like song trivia

47:51

about it, which was which was fun to learn,

47:53

is that like basically the

47:55

song, like the songwriter for it

47:57

was like kind of approached by like the

48:00

lead singer Maurice and was

48:02

like, hey, you want to write the song? And

48:04

uh, he wrote the song.

48:06

And it was really funny because apparently,

48:09

like according to everyone,

48:11

everyone is always like what the hell is happening on

48:13

the twenty first of September And they literally were

48:15

like it just sounded the best, Like

48:18

there's nothing going on that day, There's

48:20

not There wasn't like that's when I was married,

48:23

you know, Like.

48:23

It was like they're just like that is

48:26

it rolls off the tongue.

48:28

And weird part of the weird thing about the song

48:30

is the guy wrote like I think he wrote

48:32

just the like he wrote the lyrics. I don't

48:34

know if he wrote the chorus. I mean, I'll give him credit

48:36

if you did. But like the

48:38

funniest part about it

48:41

was that he wrote the song apparently,

48:43

and it's you know, uh got

48:46

you know, the regular lyrics, and then the chorus

48:48

starts and he goes, man,

48:51

you know like that. And first off,

48:54

side note, this is one fight I've gotten

48:56

in with somebody pretty much every two years.

48:59

It is body Ah. It is not party

49:01

on. It is not any other It is

49:03

just.

49:06

Yeah that says party on.

49:08

There. There are people I have had arguments

49:10

with that go that's not what it says.

49:13

I go look it up, look up what the lyrics are.

49:15

And the amount of times I've looked at the Google

49:17

search results for this song is too

49:19

many because everyone's like, oh god,

49:21

you were right.

49:22

I was like, I know, but basically

49:25

like wait, Colin, why are you so upset over this?

49:27

And You're like, it's just it's my thing, Okay,

49:30

Yeah, it's my thing.

49:31

One thing that's really funny with this is

49:34

that apparently when the when the

49:36

lead writer like wrote the song,

49:39

he didn't have that part in there, and apparently

49:42

he hated that partly,

49:44

like.

49:44

Yeah, it's kind of it doesn't make sense.

49:46

It's yeah, it's uh, I'm sorry

49:48

what I've been misgendering this person

49:50

the whole time. It's like, uh, Ali

49:53

Williams writes the song, so like, uh,

49:55

they they wrote the song. I'm just gonna say they let's

49:58

just go with that. They wrote the song, and

50:00

and like it was so funny

50:02

because they have this quote that I love that it says,

50:04

quote, I just said, what the fuck does

50:06

body ah mean? And

50:09

he and Maurice essentially said, quote,

50:11

who the fuck cares? And they said

50:14

I learned my greatest lesson in songwriting

50:16

from him that not every lyric,

50:19

uh, that not a lyric should get

50:21

away in the groove, you know what I mean? Basically

50:23

being like, yeah, it's just fun you know what I

50:25

mean? Like, honestly, if I

50:27

asked you to like sing that song, that's probably

50:29

the part you remember, you know what I mean? Other than

50:32

do you remember?

50:35

But yeah, it's

50:37

so funny. Like I was just I

50:40

was just dying at that is. I was just

50:42

like, what the fuck is body? Yeah, anyway,

50:47

we're gonna leave you.

50:47

With that question audience, What the fuck is body?

50:49

What the fuck

50:52

is body? Yeah,