1
00:00:00,040 --> 00:00:03,120
Coming up today on Mini Tech 
Scrolls, a Sydney based musician

2
00:00:03,120 --> 00:00:07,080
calls out an AI performer and 
who booked them, how Calci and 

3
00:00:07,080 --> 00:00:10,720
Spotify are rigging the charts 
and a gay dating app using AI 

4
00:00:10,720 --> 00:00:14,320
men to lure in users. 
Hey everyone and welcome to Mini

5
00:00:14,320 --> 00:00:16,200
Scroll. 
The daily Internet culture 

6
00:00:16,200 --> 00:00:19,000
podcast bring you the biggest 
stories from social media, the 

7
00:00:19,000 --> 00:00:22,040
creator economy and the digital 
space every Monday through 

8
00:00:22,040 --> 00:00:25,120
Thursday and sometimes Fridays. 
My name is Annabelle and I'm 

9
00:00:25,120 --> 00:00:28,480
your Wednesday Mini Tech Scroll 
host and also the deputy editor 

10
00:00:28,480 --> 00:00:32,080
of Australian Youth Digital 
media brand Centennial World. 

11
00:00:32,320 --> 00:00:35,160
Welcome back to Minitex Girl for
another Wednesday. 

12
00:00:35,320 --> 00:00:38,480
Before we get into the stories 
this week, I feel like I have to

13
00:00:38,480 --> 00:00:41,360
note, I don't know if it sounds 
like it, but to me it sounds 

14
00:00:41,360 --> 00:00:44,800
like I'm losing my voice. 
So I'm really sorry if I sound a

15
00:00:44,800 --> 00:00:46,720
little bit croaky throughout 
this episode. 

16
00:00:46,720 --> 00:00:50,160
I do have some tea sitting next 
to me and I'm not even a tea 

17
00:00:50,160 --> 00:00:52,200
drinker, but we do what we got 
to do. 

18
00:00:52,480 --> 00:00:55,120
And to be honest, it probably 
didn't help that on my way in 

19
00:00:55,120 --> 00:01:00,080
today, I was singing along to 
Ariana Grande's probably loudest

20
00:01:00,080 --> 00:01:03,440
album, Dangerous Woman. 
So I may have made things worse.

21
00:01:03,440 --> 00:01:06,360
But yes, other than that, Please
remember to rate, review, and 

22
00:01:06,360 --> 00:01:09,840
subscribe all of the things. 
It's all super helpful, 

23
00:01:09,840 --> 00:01:12,880
especially as we're testing a 
new thumbnail and title 

24
00:01:12,880 --> 00:01:15,360
structure. 
So also feel free to let us know

25
00:01:15,360 --> 00:01:18,440
what you think of that. 
OK, let's get into the content 

26
00:01:18,440 --> 00:01:20,560
for this week. 
So the first story I have today 

27
00:01:20,560 --> 00:01:24,720
is actually a Local 1 and I do 
really enjoy covering more local

28
00:01:24,720 --> 00:01:27,400
stories every once in a while 
because I don't know if it's 

29
00:01:27,400 --> 00:01:30,120
something that you guys who are 
not based in Sydney may have 

30
00:01:30,120 --> 00:01:33,000
heard of. 
So for our first story today, a 

31
00:01:33,000 --> 00:01:35,880
Sydney artist has taken to 
social media to reveal that 

32
00:01:35,880 --> 00:01:39,800
their band was booked to play 
alongside an AI artist. 

33
00:01:39,840 --> 00:01:43,080
Like I said, this is a little 
bit of a niche one, but I do 

34
00:01:43,080 --> 00:01:45,440
also think it's quite important 
to cover these stories, 

35
00:01:45,440 --> 00:01:49,560
especially when it comes to AI 
because it's showing how AI is 

36
00:01:49,560 --> 00:01:53,960
changing slash impacting 
industries at the local level 

37
00:01:54,160 --> 00:01:58,240
and obviously in this instance, 
it is the creative space. 

38
00:01:58,440 --> 00:02:02,120
Anyway, Last week, a musician 
named Aiden took to Instagram to

39
00:02:02,120 --> 00:02:04,760
share a recent experience he had
playing a gig. 

40
00:02:04,760 --> 00:02:09,000
So he performed alongside his 
band at a local venue in Newtown

41
00:02:09,000 --> 00:02:12,160
called Bootleggers for an event 
called Mixed Bag. 

42
00:02:12,160 --> 00:02:16,320
The purpose of the Mixed Bag 
event series is to platform and 

43
00:02:16,320 --> 00:02:19,320
highlight emerging local artists
which which is obviously very 

44
00:02:19,320 --> 00:02:23,160
important and this event is 
booked by a Marrickville based 

45
00:02:23,160 --> 00:02:26,320
music Services Agency. 
Good Intent and for a little bit

46
00:02:26,320 --> 00:02:29,680
more context on Mixed Bag, 
attendees do have to pay and it 

47
00:02:29,680 --> 00:02:34,040
costs about 20 AUD each. 
So the key players in this story

48
00:02:34,040 --> 00:02:38,240
are Aiden Bootleggers, which is 
the venue, and Good Intent which

49
00:02:38,240 --> 00:02:40,720
did the booking ahead of this 
episode. 

50
00:02:40,720 --> 00:02:44,320
I did contact Aiden and Good 
Intent for comment. 

51
00:02:44,320 --> 00:02:47,240
Aiden did get back to us in time
for our deadline, but Good 

52
00:02:47,240 --> 00:02:50,280
Intent did not. 
So in Aiden's Instagram video, 

53
00:02:50,280 --> 00:02:53,880
he alleged that the artist that 
played before him was AI 

54
00:02:53,880 --> 00:02:57,000
generated or was producing AI 
generated music. 

55
00:02:57,280 --> 00:02:59,920
He said he was only aware of 
this when a member of another 

56
00:02:59,920 --> 00:03:03,240
band on the lineup had actually 
looked through the Instagram of 

57
00:03:03,240 --> 00:03:05,560
this artist and then informed 
them about it. 

58
00:03:05,760 --> 00:03:09,040
As for how Aiden felt when he 
realized that there was an AI 

59
00:03:09,040 --> 00:03:12,160
artist on line up, he says that 
he was quote shocked and 

60
00:03:12,160 --> 00:03:15,360
confused at first, not only at 
how that was allowed to happen, 

61
00:03:15,360 --> 00:03:18,200
but also at how that would 
actually work in a live setting.

62
00:03:18,200 --> 00:03:20,720
When I thought about it more, 
the confusion turned into 

63
00:03:20,720 --> 00:03:24,000
concern at the direction the 
live music scene is headed in. 

64
00:03:24,200 --> 00:03:26,960
I do think Aiden's initial 
reaction is definitely something

65
00:03:27,040 --> 00:03:30,520
we can all relate to because 
when I first saw this video, one

66
00:03:30,520 --> 00:03:34,400
of my first questions was how 
does an AI artist even work in a

67
00:03:34,400 --> 00:03:37,760
live setting? 
And Aiden does go on to clarify 

68
00:03:37,760 --> 00:03:40,240
what he means by an AI generated
artist. 

69
00:03:40,240 --> 00:03:43,720
In the video, he claims that the
social media presence of this 

70
00:03:43,720 --> 00:03:47,320
artist seemed to be dominated by
AI visuals and what he believed 

71
00:03:47,320 --> 00:03:50,160
to be AI generated music over 
top of it. 

72
00:03:50,160 --> 00:03:53,280
And it seems that the way this 
artist performed was that there 

73
00:03:53,280 --> 00:03:57,280
was indeed a human on the stage 
and they were singing over the 

74
00:03:57,280 --> 00:04:00,040
top of what seems to be AI 
generated music. 

75
00:04:00,480 --> 00:04:03,400
But let's listen to Aidan's 
video from Instagram. 

76
00:04:03,640 --> 00:04:06,920
Last night I played a gig at 
Bootleggers in Newtown, Sydney 

77
00:04:07,560 --> 00:04:10,200
and the artist that played 
before us was AI generated. 

78
00:04:10,360 --> 00:04:13,240
Our band was playing as part of 
1 of Bootleggers mixed bag 

79
00:04:13,240 --> 00:04:15,920
events, the line UPS of which 
are put together by the agency. 

80
00:04:15,920 --> 00:04:18,600
Good intent. 
The artists do not have control 

81
00:04:18,680 --> 00:04:21,160
over the line up. 
At no point was it disclosed to 

82
00:04:21,160 --> 00:04:23,200
us that we would be playing with
an AI artist. 

83
00:04:23,440 --> 00:04:26,840
As a musician it is a complete 
insult to be put on a line up 

84
00:04:26,840 --> 00:04:30,400
with someone who uses technology
that can only exist because of 

85
00:04:30,400 --> 00:04:34,080
the large scale theft of the 
work of my other fellow human 

86
00:04:34,080 --> 00:04:36,240
musicians. 
I don't know if Good Intent was 

87
00:04:36,240 --> 00:04:38,680
aware that this was an AI 
generated artist when they 

88
00:04:38,680 --> 00:04:41,200
booked them. 
If they weren't then whoever is 

89
00:04:41,240 --> 00:04:44,240
looking for them needs to do a 
little bit more research as I 

90
00:04:44,240 --> 00:04:47,240
mean just look at the account. 
It's fairly obvious what's going

91
00:04:47,240 --> 00:04:49,120
on here. 
As streaming platforms become 

92
00:04:49,120 --> 00:04:52,600
increasingly inundated with AI 
slop, live music must be 

93
00:04:52,600 --> 00:04:55,360
defended as an Ave. for human 
creative expression. 

94
00:04:55,680 --> 00:04:59,320
There is no place for generative
AI slop in the live music scene.

95
00:04:59,320 --> 00:05:01,680
The appeal of live performance 
is that it is raw. 

96
00:05:01,680 --> 00:05:03,560
It is passionate in that it is 
human. 

97
00:05:03,840 --> 00:05:06,920
AI music delivers none of this, 
while robbing the opportunity to

98
00:05:06,920 --> 00:05:10,080
do so from genuinely talented 
human musicians who can. 

99
00:05:10,520 --> 00:05:13,160
As this becomes more 
commonplace, venues are going to

100
00:05:13,160 --> 00:05:16,280
need to work harder to protect 
the very human artists that have

101
00:05:16,280 --> 00:05:18,880
kept patrons coming through 
their doors before the scene 

102
00:05:18,880 --> 00:05:21,320
that keeps them alive dies off 
entirely. 

103
00:05:21,520 --> 00:05:27,320
So this video currently has over
350 K views and over 22 K likes.

104
00:05:27,320 --> 00:05:30,160
The comments are all super, 
super supportive and I'm sure 

105
00:05:30,160 --> 00:05:33,800
that was a really reassuring for
Aiden because he told us that 

106
00:05:33,800 --> 00:05:37,200
his decision to speak about it 
publicly was because he felt, 

107
00:05:37,200 --> 00:05:40,000
quote, that if it wasn't spoke 
about, it could easily slip 

108
00:05:40,000 --> 00:05:42,880
through the cracks again. 
So he has definitely brought 

109
00:05:42,880 --> 00:05:45,240
significant awareness to this 
issue. 

110
00:05:45,280 --> 00:05:48,240
And I did also ask him if this 
kind of thing is something he is

111
00:05:48,240 --> 00:05:52,200
seeing more and more in the live
music space as a musician, or 

112
00:05:52,200 --> 00:05:54,440
whether this was like an 
isolated incident. 

113
00:05:54,880 --> 00:05:58,080
And in response to that, he said
that he hopes it's an isolated 

114
00:05:58,080 --> 00:06:00,400
incident quote. 
But if it's not addressed at an 

115
00:06:00,400 --> 00:06:03,480
early age, it could easily 
become a bigger issue with live 

116
00:06:03,480 --> 00:06:05,920
musicians finding it 
increasingly hard to compete 

117
00:06:05,920 --> 00:06:09,400
with cheaper AI options. 
And I think that is definitely a

118
00:06:09,400 --> 00:06:12,360
valid concern. 
Now, as you probably had in 

119
00:06:12,360 --> 00:06:15,600
Aiden's video, he didn't name 
the AI artist. 

120
00:06:15,600 --> 00:06:18,400
And I'm going to follow his lead
here because I don't really want

121
00:06:18,400 --> 00:06:21,600
to give traction to this artist 
necessarily. 

122
00:06:21,880 --> 00:06:25,800
But I did go have a look at this
artist Instagram profile because

123
00:06:25,800 --> 00:06:29,200
obviously I was curious and it 
definitely does seem to 

124
00:06:29,200 --> 00:06:32,840
incorporate some kind of AI to 
promote the performance at 

125
00:06:32,840 --> 00:06:35,480
bootleggers. 
The artist said a promotional 

126
00:06:35,480 --> 00:06:40,040
poster which read quote one 
human 2 robots pop music. 

127
00:06:40,040 --> 00:06:44,040
So with that it's very clear 
that the music is generated with

128
00:06:44,240 --> 00:06:48,120
AI to some degree. 
In fact this was confirmed by 

129
00:06:48,120 --> 00:06:50,960
the Sydney Morning Herald and 
I'm going to presume that this 

130
00:06:50,960 --> 00:06:55,280
person is most likely making 
this AI generated music through 

131
00:06:55,360 --> 00:06:58,720
SUNO, but there is no 
confirmation of that on the 

132
00:06:59,040 --> 00:07:02,560
artist profile. 
According to the bio of this 

133
00:07:02,560 --> 00:07:07,360
artist, they are a Sydney based 
Afro pop band made-up of 3/2 of 

134
00:07:07,480 --> 00:07:12,160
which are avatars and one is the
lyricist and producer and it 

135
00:07:12,160 --> 00:07:16,040
said in live performances the 
lyricist is the one who performs

136
00:07:16,040 --> 00:07:21,280
the band's songs on stage and he
adds his vocals to the avatars 

137
00:07:21,360 --> 00:07:23,880
voices. 
They say that down the line they

138
00:07:23,880 --> 00:07:27,400
will have more performers added 
to this rotation and it's 

139
00:07:27,400 --> 00:07:31,440
unclear if that means AI quote 
UN quote performers or human 

140
00:07:31,440 --> 00:07:34,120
performers. 
So from my description of this 

141
00:07:34,120 --> 00:07:37,400
artist's Instagram profile, I 
think I can say that it's fairly

142
00:07:37,480 --> 00:07:41,880
obvious that this artist is AI. 
So with that in mind, I think a 

143
00:07:41,880 --> 00:07:44,680
lot of people were worried about
the fact that the booking agency

144
00:07:44,680 --> 00:07:48,120
didn't seem to care or they were
just unaware of it. 

145
00:07:48,120 --> 00:07:51,560
And this, of course, along with 
Aiden's video gaining traction, 

146
00:07:51,840 --> 00:07:54,760
prompted bootleggers and Good 
Intent to post a shared 

147
00:07:54,760 --> 00:07:57,200
statement to Instagram. 
The statement was written from 

148
00:07:57,200 --> 00:08:00,480
the perspective of Rob Carroll, 
who is the founder and director 

149
00:08:00,480 --> 00:08:02,360
at Good Intent. 
And I'll read you part of the 

150
00:08:02,360 --> 00:08:05,240
statement now quote, I'll take 
full responsibility. 

151
00:08:05,280 --> 00:08:07,880
I didn't look at this particular
artist's social and their 

152
00:08:07,880 --> 00:08:10,280
representatives. 
Communications used word 

153
00:08:10,280 --> 00:08:12,280
inconsistent with genuine 
artists. 

154
00:08:12,280 --> 00:08:15,640
Their representative described 
them to us as a producer who 

155
00:08:15,640 --> 00:08:19,080
used live vocals in their sets. 
This person was booked through 

156
00:08:19,080 --> 00:08:21,680
an agency who presented 
themselves as professional. 

157
00:08:21,800 --> 00:08:24,680
I hadn't seen them live, but now
looking at their socials, their 

158
00:08:24,680 --> 00:08:27,080
content is clearly and blatantly
AI. 

159
00:08:27,160 --> 00:08:30,840
Good Intent and bootleggers have
never knowingly and will never 

160
00:08:30,840 --> 00:08:33,559
book people who use AI to 
generate music. 

161
00:08:33,559 --> 00:08:36,799
For those who don't know us. 
Good Intent is passionate about 

162
00:08:36,799 --> 00:08:39,760
supporting and doing the right 
thing by real artists. 

163
00:08:40,039 --> 00:08:42,559
We'll ensure we review our 
vetting processes moving 

164
00:08:42,559 --> 00:08:44,680
forward. 
We'd never want to be placed in 

165
00:08:44,680 --> 00:08:47,480
this position where we're 
encouraging the use of AI and 

166
00:08:47,480 --> 00:08:50,960
financially supporting the theft
from artists on which AI is 

167
00:08:50,960 --> 00:08:53,160
trained to put our money where 
our mouth is. 

168
00:08:53,200 --> 00:08:56,240
Bootleggers and Good Intent will
also make a donation to Support 

169
00:08:56,240 --> 00:08:58,040
ACT. 
So the response to this 

170
00:08:58,040 --> 00:09:01,240
statement has been quite 
positive, with many social media

171
00:09:01,240 --> 00:09:04,560
users and Sydneysiders 
appreciative that Good Intent 

172
00:09:04,560 --> 00:09:07,320
took accountability. 
And honestly, this statement 

173
00:09:07,600 --> 00:09:10,600
could be used as a blueprint for
brands slash companies to 

174
00:09:10,680 --> 00:09:14,800
address AI usage or I guess 
platforming AI when it's not 

175
00:09:14,800 --> 00:09:16,560
something that they intended to 
do. 

176
00:09:16,760 --> 00:09:19,760
It felt quite honest in that 
they like readily took 

177
00:09:19,760 --> 00:09:21,960
responsibility for what went 
down. 

178
00:09:21,960 --> 00:09:25,240
When I read this statement, my 
first thought was actually how 

179
00:09:25,240 --> 00:09:29,040
different this was to huff his 
response to their alleged AI 

180
00:09:29,040 --> 00:09:30,960
use. 
Granted, this is a different 

181
00:09:30,960 --> 00:09:34,000
situation, but I thought it was 
an interesting comparison of how

182
00:09:34,000 --> 00:09:37,640
one company immediately admitted
to wrongdoing and the other did 

183
00:09:37,640 --> 00:09:42,240
not and how that impacted how 
social media users approached 

184
00:09:42,240 --> 00:09:45,040
the quote UN quote scandal. 
In case you missed it, I did 

185
00:09:45,040 --> 00:09:48,200
report on the Hoffa situation a 
couple of weeks ago. 

186
00:09:48,200 --> 00:09:52,520
But as a reminder, a New Zealand
based model named Elijah accused

187
00:09:52,520 --> 00:09:56,120
Hoffa, which is a New Zealand 
St. wear brand of using AI 

188
00:09:56,120 --> 00:09:59,720
generated models in campaigns. 
And if you want more detail I 

189
00:09:59,720 --> 00:10:02,880
will link the episode where I 
covered it in the description. 

190
00:10:02,920 --> 00:10:06,840
But to make a Long story short, 
Hoffa really doubled down on the

191
00:10:06,840 --> 00:10:11,040
fact that they did not use AI in
the way Elijah was alleging. 

192
00:10:11,440 --> 00:10:15,600
And that may be true, but the 
way they did so was honestly 

193
00:10:15,600 --> 00:10:18,280
quite distasteful. 
Representatives from the brand 

194
00:10:18,280 --> 00:10:22,440
went on TV and also ended up 
insulting slash intimidating 

195
00:10:22,440 --> 00:10:25,840
Elijah in the process. 
It was honestly crazy. 

196
00:10:26,080 --> 00:10:29,240
Many people felt like Hoffa was 
truly ruining their reputation 

197
00:10:29,240 --> 00:10:32,320
and it was just kind of like an 
overall PR disaster for them. 

198
00:10:32,320 --> 00:10:35,040
And obviously this was quite a 
different approach to good 

199
00:10:35,040 --> 00:10:36,720
intent. 
But before I speak about that a 

200
00:10:36,720 --> 00:10:40,080
little bit more, there are some 
updates on the Hoffa situation 

201
00:10:40,080 --> 00:10:42,720
that I wanted to quickly note. 
The most recent update is that 

202
00:10:42,720 --> 00:10:45,240
Hoffa commissioned an 
independent forensic 

203
00:10:45,240 --> 00:10:47,840
investigation into the AI 
allegations. 

204
00:10:47,840 --> 00:10:51,160
There is a limited detail about 
who conducted it though, and I 

205
00:10:51,160 --> 00:10:55,480
did go to the link in their bio 
to try and learn more about this

206
00:10:55,480 --> 00:10:59,840
investigation and AI policy, but
the 404 error came up, so I 

207
00:10:59,840 --> 00:11:03,360
wasn't able to access that page.
But according to the statement 

208
00:11:03,360 --> 00:11:06,760
they recently issued, they found
quote, no evidence that Hoffa or

209
00:11:06,760 --> 00:11:10,400
the agency acting on our behalf 
used facial image material to 

210
00:11:10,400 --> 00:11:12,560
generate the AI images in 
question. 

211
00:11:12,760 --> 00:11:15,360
It also found no evidence 
supporting allegations that 

212
00:11:15,360 --> 00:11:19,160
Hoffa copied or recreated the 
likeness of any model using AI. 

213
00:11:19,160 --> 00:11:22,480
And at the same time, they also 
admitted that, quote, aspects of

214
00:11:22,480 --> 00:11:25,440
our public response fell short 
of the standard we expect of 

215
00:11:25,440 --> 00:11:27,600
ourselves. 
For that we are genuinely sorry 

216
00:11:27,600 --> 00:11:29,280
and we've learnt from this 
experience. 

217
00:11:29,280 --> 00:11:32,320
This investigation slash 
statement came off the back of 

218
00:11:32,320 --> 00:11:35,920
another statement that they had 
made that was posted on June 

219
00:11:36,160 --> 00:11:38,120
11th. 
And it basically said that the 

220
00:11:38,120 --> 00:11:41,840
specific imagery in question did
quote, generate a new image 

221
00:11:41,840 --> 00:11:45,280
using an AI design tool, but the
image was not engineered to 

222
00:11:45,280 --> 00:11:49,680
resemble any specific person. 
So they did admit to some level 

223
00:11:49,680 --> 00:11:51,760
of AI usage. 
And now they're kind of going 

224
00:11:51,760 --> 00:11:55,600
back on that. 
And honestly, it has left a lot 

225
00:11:55,600 --> 00:11:58,240
of people, including Elijah 
confused. 

226
00:11:58,320 --> 00:12:02,000
Elijah is continuing to call out
what seems to be contradictory 

227
00:12:02,000 --> 00:12:05,000
messaging in his content. 
So if anything else significant 

228
00:12:05,000 --> 00:12:07,520
drops from that, I will include 
it in an episode. 

229
00:12:07,520 --> 00:12:11,360
But All in all, comparing these 
two situations from a brand 

230
00:12:11,360 --> 00:12:15,320
perspective, it's clear why good
intent statement was well 

231
00:12:15,320 --> 00:12:18,640
received compared to Hoffa. 
Like I said, I do think the 

232
00:12:18,640 --> 00:12:22,800
Hoffa situation is a little bit 
harder to prove, but it all just

233
00:12:22,800 --> 00:12:25,560
kind of comes down to 
consistency and messaging. 

234
00:12:25,600 --> 00:12:29,600
Hafiz messaging has not been the
same across the board and also 

235
00:12:29,600 --> 00:12:32,160
the tone has differed 
significantly depending on the 

236
00:12:32,160 --> 00:12:34,840
medium. 
And on top of this, they seem to

237
00:12:34,840 --> 00:12:37,640
be quite combative slash 
defensive when it came to 

238
00:12:37,640 --> 00:12:41,800
Elijah's content, whereas good 
intent welcomed Aiden's video. 

239
00:12:41,840 --> 00:12:45,200
And I think that went a long way
for how the audience perceived 

240
00:12:45,200 --> 00:12:46,720
them. 
But yeah, I do think there was 

241
00:12:46,720 --> 00:12:49,240
just some key differences 
between the two situations. 

242
00:12:49,240 --> 00:12:51,840
But I'd definitely be curious to
hear what you guys thought of 

243
00:12:51,840 --> 00:12:54,560
these two and whether you think 
this comparison is even valid. 

244
00:12:54,600 --> 00:12:56,800
OK, let's move on to the next 
story. 

245
00:12:56,800 --> 00:13:00,720
And we have more news related to
prediction markets, which is, 

246
00:13:00,720 --> 00:13:03,120
like I said last week, my 
special interest at this point. 

247
00:13:03,200 --> 00:13:07,480
But honestly, in the way that 
Lauren had Simone scroll for 

248
00:13:07,480 --> 00:13:11,600
every time that she spoke about 
Club Chalamet on mini scroll, we

249
00:13:11,600 --> 00:13:14,200
need an equivalent for 
prediction markets. 

250
00:13:14,200 --> 00:13:16,760
Like at this point, it's mini 
prediction market scroll. 

251
00:13:17,360 --> 00:13:19,720
But anyway, I wanted to talk 
about everything that's 

252
00:13:19,720 --> 00:13:22,040
happening between Cal, she and 
Spotify. 

253
00:13:22,040 --> 00:13:26,000
So I guess this is quite a music
heavy episode now with a Spotify

254
00:13:26,000 --> 00:13:28,840
focus story, but we lean in 
SoCal. 

255
00:13:28,840 --> 00:13:32,400
She and Spotify have come under 
fire for allegedly altering 

256
00:13:32,400 --> 00:13:35,280
music chart. 
Kaushi basically has these 

257
00:13:35,280 --> 00:13:38,680
monthly betting markets on 
Spotify's top song chart. 

258
00:13:39,040 --> 00:13:42,440
And a man named Caleb Davies, 
who's actually 1 of Kaushi's top

259
00:13:42,440 --> 00:13:46,440
traders, has basically been 
sounding the alarm that there is

260
00:13:46,440 --> 00:13:50,640
bot powered manipulation going 
on with Spotify's charts which 

261
00:13:50,680 --> 00:13:53,840
ultimately messes with the 
betting market that's tied to 

262
00:13:53,840 --> 00:13:56,040
them. 
He's been posting all of his 

263
00:13:56,040 --> 00:13:59,400
like data slash receipts under a
pseudonym on X. 

264
00:13:59,560 --> 00:14:01,880
If you want to check that out, I
will also link it in the 

265
00:14:01,880 --> 00:14:04,920
description, but obviously I 
will give you a summary because 

266
00:14:04,920 --> 00:14:07,640
this story I feel like can be a 
little bit complicated. 

267
00:14:07,640 --> 00:14:11,080
So at the end of June, Caleb was
making couple tweets flagging 

268
00:14:11,080 --> 00:14:13,760
that there was some weird 
activity around streaming 

269
00:14:13,760 --> 00:14:16,400
numbers in that month's chart 
market. 

270
00:14:16,400 --> 00:14:19,320
He basically claimed that there 
were weird spikes in streaming 

271
00:14:19,320 --> 00:14:23,480
numbers that didn't make sense 
and seemed manufactured. 

272
00:14:23,480 --> 00:14:27,280
But these allegations really 
ramped up with Malcolm Todd. 

273
00:14:27,480 --> 00:14:30,080
If you aren't familiar with 
Malcolm Todd, he is kind of just

274
00:14:30,080 --> 00:14:32,880
like indie pop alternative R&B 
guy. 

275
00:14:33,680 --> 00:14:37,280
He's also brothers with Audrey 
Hobart, who has definitely been 

276
00:14:37,280 --> 00:14:40,320
gaining traction on TikTok for 
quite a while now, especially 

277
00:14:40,400 --> 00:14:43,600
after Co writing with Gracie 
Abrams I think it was, and then 

278
00:14:43,600 --> 00:14:47,240
also releasing her song Sue Me. 
But when it comes to Malcolm 

279
00:14:47,240 --> 00:14:50,600
Todd, there seemed to be an 
inauthentic surge with his 

280
00:14:50,600 --> 00:14:55,080
streams on his song Earrings. 
As Caleb explained in a tweet 

281
00:14:55,080 --> 00:14:58,800
quote, Malcolm Todd got fast in 
what also appears to be blatant 

282
00:14:58,800 --> 00:15:02,400
fraud with a 70% increase in US 
streams in a day. 

283
00:15:02,440 --> 00:15:06,280
And then he attached a 
screenshot of Spotify charts 

284
00:15:06,280 --> 00:15:10,720
that showed that Earrings had 
hit a new peak of #1 where it 

285
00:15:10,720 --> 00:15:13,360
was previously at #4 on the 
charts. 

286
00:15:13,360 --> 00:15:17,080
In other tweets, he also shared 
screenshots of charts monitoring

287
00:15:17,080 --> 00:15:19,040
streams. 
And there definitely did seem to

288
00:15:19,040 --> 00:15:23,040
be some quite dramatic spikes. 
Fans noted that it could have 

289
00:15:23,040 --> 00:15:26,480
been a result of listening 
parties, but Caleb did push back

290
00:15:26,480 --> 00:15:30,720
on that, showing what the shape 
of a legit streaming bump would 

291
00:15:30,720 --> 00:15:32,840
look like. 
Say, when there are events like 

292
00:15:32,840 --> 00:15:35,720
the Grammys or Coachella where 
the streams will kind of 

293
00:15:35,720 --> 00:15:39,400
gradually build up and then 
taper off kind of naturally. 

294
00:15:39,680 --> 00:15:42,600
According to his daughter, 
Malcolm Todd's streams did not 

295
00:15:42,600 --> 00:15:44,880
look like that at all. 
It literally just spiked out of 

296
00:15:44,880 --> 00:15:46,840
nowhere and then dropped her 
right back down. 

297
00:15:46,840 --> 00:15:49,360
On July 1, he shared a tweet 
trying to figure out the 

298
00:15:49,360 --> 00:15:52,480
likelihood of something like 
that even happening as he wrote 

299
00:15:52,480 --> 00:15:54,480
quote. 
As we await the final June 

300
00:15:54,480 --> 00:15:58,200
Spotify update, one last look at
how unlikely yesterday's earring

301
00:15:58,200 --> 00:16:00,440
surge was to have occurred by 
chance. 

302
00:16:00,800 --> 00:16:04,080
Looking at the data set of 
Sunday to Monday changes, it was

303
00:16:04,080 --> 00:16:08,720
a roughly 1 and 77 octillion 
chance of happening randomly. 

304
00:16:08,720 --> 00:16:12,040
So as you can gather from that, 
it is pretty unlikely. 

305
00:16:12,040 --> 00:16:15,320
Over the course of him tweeting 
about this, he also said that he

306
00:16:15,320 --> 00:16:19,120
had informed Calci and Spotify 
about the potential manipulation

307
00:16:19,400 --> 00:16:23,480
and publicly called on Calci to 
hold trading on the market and 

308
00:16:23,480 --> 00:16:26,520
hold off and paying out until 
someone actually investigated 

309
00:16:26,520 --> 00:16:29,440
what was going on. 
In one tweet, he was theorizing 

310
00:16:29,440 --> 00:16:32,520
that there was a Poly market 
trader inciting songs into 

311
00:16:32,520 --> 00:16:36,280
playlist to farm streams. 
As far as how Cal she initially 

312
00:16:36,280 --> 00:16:38,800
responded to Caleb's claims, 
Cal. 

313
00:16:38,800 --> 00:16:41,920
She did end up paying out the 
winners and Caleb shared that 

314
00:16:41,920 --> 00:16:45,320
they actually emailed him about 
it as well as he wrote in 

315
00:16:45,320 --> 00:16:47,600
another tweet. 
Quote Cal, she did pay out the 

316
00:16:47,600 --> 00:16:51,000
market based on fraudulent 
results right after sending me 

317
00:16:51,000 --> 00:16:54,000
an e-mail stating that there are
many plausible reasons that 

318
00:16:54,000 --> 00:16:57,960
Malcolm Todd's timely surge was 
not due to artificial boosting 

319
00:16:58,000 --> 00:17:00,400
and obviously he pushes back 
against that. 

320
00:17:00,520 --> 00:17:04,079
But here is I guess the twist. 
Spotify later confirmed to why 

321
00:17:04,119 --> 00:17:08,440
that yes, they did find evidence
of artificial streaming and this

322
00:17:08,440 --> 00:17:10,440
was the statement that was given
to the publication. 

323
00:17:10,440 --> 00:17:13,400
Quote, all streaming services 
face ever changing stream 

324
00:17:13,400 --> 00:17:16,119
manipulation. 
Spotify has best in class 

325
00:17:16,119 --> 00:17:18,640
detection and mitigation 
practices for manipulated 

326
00:17:18,640 --> 00:17:22,079
streams, and we don't pay out 
associated royalties. 

327
00:17:22,319 --> 00:17:25,640
A spokesperson for Calci did 
also say that they're in contact

328
00:17:25,920 --> 00:17:28,640
with Spotify to investigate this
issue. 

329
00:17:28,720 --> 00:17:32,960
Spotify has since asked Calci to
remove the company's logo from 

330
00:17:32,960 --> 00:17:36,960
its platform and website, and 
Calci has ultimately added a 

331
00:17:37,080 --> 00:17:40,040
disclaimer ascending that its 
products quote have not been 

332
00:17:40,040 --> 00:17:43,640
endorsed by Spotify. 
From Spotify's end, they have 

333
00:17:43,640 --> 00:17:47,680
ultimately adjusted the charts 
to account for the discrepancy, 

334
00:17:47,920 --> 00:17:52,560
culling over 500 1000 artificial
streams which ultimately would 

335
00:17:52,560 --> 00:17:55,960
have taken Malcolm Todd's song 
down to 4th. 

336
00:17:56,520 --> 00:17:59,240
In saying that though, many 
publications have been knowing 

337
00:17:59,240 --> 00:18:02,720
that there is no suggestion that
Malcolm Todd or his team were 

338
00:18:02,720 --> 00:18:07,160
involved in any attempt to boost
streams and numbers on Spotify. 

339
00:18:07,440 --> 00:18:11,240
It seems like the going theory 
is that potentially someone who 

340
00:18:11,240 --> 00:18:16,560
was betting on Malcolm Todd's 
song to be #1 paid for, but to 

341
00:18:16,560 --> 00:18:19,480
string the song where they would
probably get a bigger payout 

342
00:18:19,880 --> 00:18:23,760
from their winds on prediction 
markets as opposed to the cost 

343
00:18:23,760 --> 00:18:26,120
they would have to pay for these
bots. 

344
00:18:26,200 --> 00:18:29,960
And of course, this all comes at
a time of rising concern around 

345
00:18:29,960 --> 00:18:33,800
the growth of prediction markets
like calcium Poly market and the

346
00:18:33,800 --> 00:18:37,400
potential for insider trading, 
which is something that these 

347
00:18:37,400 --> 00:18:41,760
platforms are trying to limit. 
But of course, when we look at 

348
00:18:41,760 --> 00:18:45,680
this from a musician's 
perspective, it is another layer

349
00:18:45,680 --> 00:18:48,320
of upsetting slash concerning 
because it's just going to make 

350
00:18:48,320 --> 00:18:51,800
it harder and harder for smaller
artist to actually chart when 

351
00:18:51,800 --> 00:18:55,280
foul play like this is in the 
mix and is an option. 

352
00:18:55,280 --> 00:18:58,680
If the numbers keep being 
inflated to this degree, it will

353
00:18:58,680 --> 00:19:01,960
become a lot harder for anyone 
to actually breakthrough in a 

354
00:19:01,960 --> 00:19:05,320
meaningful way with limited 
financial resources and backing 

355
00:19:05,320 --> 00:19:07,920
behind them. 
OK, let's get into our final 

356
00:19:07,920 --> 00:19:10,440
story for today. 
And this was actually a request 

357
00:19:10,640 --> 00:19:14,960
from a Spotify comment to cover 
this new dating app Goose. 

358
00:19:14,960 --> 00:19:18,520
I actually did see this on X as 
well, so thank you very much for

359
00:19:18,520 --> 00:19:20,280
the reminder. 
There has been a lot of 

360
00:19:20,280 --> 00:19:23,880
discussion around this app 
online thanks to a piece by 

361
00:19:23,880 --> 00:19:28,400
Wired that claimed it was a si 
up meaning psychological 

362
00:19:28,400 --> 00:19:30,520
operation. 
This is typically used in media 

363
00:19:30,520 --> 00:19:33,640
or like commentary to describe a
phenomenon where there's 

364
00:19:33,640 --> 00:19:37,520
manipulation of information or 
like media or anything like that

365
00:19:37,760 --> 00:19:41,040
to implement emotion or public 
opinion, that kind of thing. 

366
00:19:41,040 --> 00:19:45,120
So right now Goose is an invite 
only dating slash quote UN quote

367
00:19:45,120 --> 00:19:48,560
lifestyle app for gay men. 
As the website reads, quote 

368
00:19:48,560 --> 00:19:51,960
Goose is the gay app not just 
for dating me guys through the 

369
00:19:51,960 --> 00:19:53,840
life. 
You already have places, 

370
00:19:53,840 --> 00:19:56,720
friends, interest stories and 
what's happening nearby. 

371
00:19:56,800 --> 00:20:00,000
In the piece Why it does go into
detail about the marketing 

372
00:20:00,000 --> 00:20:04,160
strategy of this app Which? 
The journalist EJ Dixon claims 

373
00:20:04,320 --> 00:20:08,440
involved using AI generated men 
to get people to join the 

374
00:20:08,440 --> 00:20:11,320
platform. 
In a thread on X, EJ Dixon does 

375
00:20:11,320 --> 00:20:15,080
share quite a succinct summary 
of what she found in her 

376
00:20:15,080 --> 00:20:17,560
investigation, so I will 
definitely link that below. 

377
00:20:17,920 --> 00:20:21,520
But she explains that she first 
found out about Goose through 

378
00:20:21,520 --> 00:20:23,880
some friends that had told her 
that they'd received messages 

379
00:20:23,880 --> 00:20:26,480
from quote UN quote random hot 
guy accounts. 

380
00:20:26,760 --> 00:20:29,640
And she notes that these 
accounts were all new and 

381
00:20:29,640 --> 00:20:32,920
conventionally attractive. 
Quote all the messages used 

382
00:20:32,920 --> 00:20:36,320
approximately the same verbiage 
and all the guys look similar. 

383
00:20:36,560 --> 00:20:38,600
They featured similarly 
aesthetic post. 

384
00:20:38,720 --> 00:20:41,280
They also all followed each 
other and commented on each 

385
00:20:41,280 --> 00:20:42,920
other's accounts. 
Weird. 

386
00:20:43,200 --> 00:20:46,720
So I started running guys photos
through Google Gemini and other 

387
00:20:46,720 --> 00:20:49,920
AI detection software. 
Most or all of the photos were 

388
00:20:49,920 --> 00:20:52,360
apparently generated using 
Google AI. 

389
00:20:52,360 --> 00:20:56,400
So essentially you have these AI
generated profiles messaging men

390
00:20:56,400 --> 00:20:58,680
to join the app. 
It's literally giving those hey 

391
00:20:58,680 --> 00:21:01,480
girlie pyramid scheme messages 
to me. 

392
00:21:01,560 --> 00:21:04,800
Anyway, EJ continues by noting 
that she found more than 20 

393
00:21:04,800 --> 00:21:07,880
accounts that followed the same 
kind of pattern, adding that 

394
00:21:07,880 --> 00:21:11,520
they also followed the Goose Co 
founder, which is another level 

395
00:21:11,520 --> 00:21:14,560
of SAS. 
And to add even more fuel to the

396
00:21:14,560 --> 00:21:18,080
fire, she found that Goose was 
actually looking to hire quote 

397
00:21:18,080 --> 00:21:22,280
UN quote ambassadors who would 
be able to run Instagram 

398
00:21:22,280 --> 00:21:25,920
accounts, seemingly as a part of
its marketing. 

399
00:21:25,920 --> 00:21:28,880
Using what seems to be fake 
accounts like this to drive 

400
00:21:28,880 --> 00:21:32,120
people to a product is obviously
unethical and it's also just not

401
00:21:32,160 --> 00:21:36,400
allowed, but it has left a lot 
of the gay community upset, 

402
00:21:36,400 --> 00:21:40,400
which is totally understandable 
as EJ continues quote. 

403
00:21:40,760 --> 00:21:44,080
It's also pissed off a lot of 
members of the gay community who

404
00:21:44,080 --> 00:21:47,040
have already criticized the app 
for a lack of diversity in its 

405
00:21:47,040 --> 00:21:49,960
marketing and promoting toxic 
body image standards. 

406
00:21:50,480 --> 00:21:53,600
On one hand, I'm flattered that 
I'm their target audience. 1 

407
00:21:53,600 --> 00:21:57,080
user told me but the need to 
essentially bake gay guys into 

408
00:21:57,080 --> 00:22:00,480
signing up feels really sketchy.
I did also have a look into the 

409
00:22:00,480 --> 00:22:04,760
sentiment online and I did see 
one Reddit user on VR slash gay 

410
00:22:04,760 --> 00:22:08,400
Reddit share how the app 
essentially made them spiral. 

411
00:22:08,760 --> 00:22:11,800
They said quote. 
I couldn't even tell what was 

412
00:22:11,800 --> 00:22:14,600
real and what was fake anymore. 
Amongst my Instagram friends. 

413
00:22:14,800 --> 00:22:17,440
I had to go through and remove 
everyone I never met in real 

414
00:22:17,440 --> 00:22:19,200
life. 
I thought this was a problem 

415
00:22:19,200 --> 00:22:22,840
that only happened to boomers. 
No one is immune in the AI era 

416
00:22:22,840 --> 00:22:25,520
clearly. 
And interestingly on the R 

417
00:22:25,640 --> 00:22:29,680
technology Reddit I have seen 
many users also say that other 

418
00:22:29,680 --> 00:22:33,560
dating apps do a similar thing. 
This is not really something 

419
00:22:33,560 --> 00:22:35,640
I've ever seen or really heard 
of online. 

420
00:22:35,640 --> 00:22:39,520
I'm definitely curious if this 
is something that you guys are 

421
00:22:39,520 --> 00:22:42,920
familiar with. 
But as one user recounted quote,

422
00:22:42,920 --> 00:22:45,520
that's basically how dating apps
have worked for years. 

423
00:22:45,520 --> 00:22:49,480
But before AI, they generated 
profiles by farming Instagram to

424
00:22:49,480 --> 00:22:52,520
send you a bunch of women who 
weren't actually using the app. 

425
00:22:52,520 --> 00:22:54,840
And there were quite a few 
comments like that. 

426
00:22:54,840 --> 00:22:57,760
So I'm gathering that this is 
something that happens to men 

427
00:22:57,800 --> 00:23:00,560
who are trying to date women. 
In saying that though, I do 

428
00:23:00,560 --> 00:23:04,320
think it is a little bit more 
harmful or I guess egregious 

429
00:23:04,320 --> 00:23:10,040
when it comes to a queer face 
slash app using AI or fake 

430
00:23:10,040 --> 00:23:13,120
accounts in this way. 
Because this is an example of 

431
00:23:13,120 --> 00:23:17,400
the use of AI to capitalize on a
marginalized community that has 

432
00:23:17,400 --> 00:23:20,160
historically found it hard to 
build these kind of 

433
00:23:20,160 --> 00:23:23,520
relationships or have a safe 
space to do so. 

434
00:23:23,920 --> 00:23:27,200
And it feels especially 
hypocritical when it is around 

435
00:23:27,200 --> 00:23:30,640
an app that is built on the 
premise of finding relationships

436
00:23:30,640 --> 00:23:33,080
and friendships. 
Even like I said, the queer 

437
00:23:33,080 --> 00:23:37,120
community has historically 
relied on these spaces out of 

438
00:23:37,120 --> 00:23:42,560
necessity, so trust and 
authenticity really matter here 

439
00:23:42,560 --> 00:23:44,760
a lot. 
And the lack of disclosure that 

440
00:23:44,760 --> 00:23:48,360
this account is essentially 
operating as an ad and is fake 

441
00:23:48,360 --> 00:23:52,520
is what definitely pushes this 
into something quite deceptive 

442
00:23:52,600 --> 00:23:54,640
because there is a chance that 
people can believe they're 

443
00:23:54,640 --> 00:23:57,440
talking to a real person. 
Anyway, I'm definitely curious 

444
00:23:57,440 --> 00:24:00,480
to hear what you guys think 
about this story, especially if 

445
00:24:00,480 --> 00:24:02,840
you are someone who has dabbled 
in the app. 

446
00:24:02,960 --> 00:24:05,760
And I'm also curious if this is 
something you have noticed with 

447
00:24:05,840 --> 00:24:09,840
other apps like AI generated 
accounts sliding into DMS or 

448
00:24:09,840 --> 00:24:12,360
trying to get you to download a 
product or an app. 

449
00:24:12,360 --> 00:24:14,640
I do feel like this is becoming 
increasingly common. 

450
00:24:14,640 --> 00:24:17,080
Well, that is everything for 
mini tech scroll today. 

451
00:24:17,080 --> 00:24:20,560
I really hope you guys enjoyed. 
Lauren will be back for a 

452
00:24:20,600 --> 00:24:23,720
regular mini scroll tomorrow and
I will be back on the mic next 

453
00:24:23,720 --> 00:24:26,440
week, presuming I don't continue
to lose my voice. 

454
00:24:27,360 --> 00:24:28,800
But thank you so much for 
listening. 

455
00:24:29,000 --> 00:24:31,120
Bye. 
Centennial World Podcast Network

456
00:24:31,120 --> 00:24:33,960
operates on Gadigal land. 
We would like to acknowledge the

457
00:24:33,960 --> 00:24:37,200
Gadigal of the Eora Nation, the 
traditional custodians of this 

458
00:24:37,200 --> 00:24:40,760
land, and pay our respects to 
elders both past and present.

