Latest / Mini Scroll / MINI TECH SCROLL: TikTok building new US app, racist AI slop goes viral + Meta AI is spreading rumours

Transcript

MINI TECH SCROLL: TikTok building new US app, racist AI slop goes viral + Meta AI is spreading rumours · Mini Scroll

• 0:00
  Hey everyone, and welcome to Mini Scroll, the daily Internet
• 0:02
  culture podcast bring you the biggest stories from social
• 0:05
  media, the creator economy, and the digital space every Monday
• 0:08
  through Thursday and sometimes Fridays.
• 0:10
  My name is Annabelle and I'm your Wednesday Mini Tech Scroll
• 0:13
  host and also the deputy editor of the youth digital media brand
• 0:17
  Centennial World. Welcome back to Mini Tech
• 0:20
  Scroll. We are back in the studio for
• 0:22
  today. I am very happy to be back and I
• 0:25
  wanted to thank you all for being so kind about last week's
• 0:29
  episode even though it was from home.
• 0:32
  Before we get into this week's stories though, I actually have
• 0:36
  a book recommendation for you guys.
• 0:38
  I finally got around to reading Get Rich or Lie, Trying Ambition
• 0:43
  and Deceit in the New Influencer Economy by Simeon Brown.
• 0:46
  It was published back in 2022 so I am a little bit late to the
• 0:50
  game, but I thought this would be a great recommendation for
• 0:54
  the pod because if you like a lot of what we cover here at
• 0:57
  Mini Scroll and Infinite Scroll, I think you would find it a
• 1:01
  interesting read. It is a nonfiction book, but it
• 1:04
  talks about everything from plastic surgery to
• 1:06
  cryptocurrency to live streaming and kind of how so many social
• 1:11
  media users succumb to the, I guess, capitalistic engine that
• 1:16
  drives social media these days in pursuit for influenced in
• 1:20
  where many end up lying to others and themselves about the
• 1:25
  process. Other than that, Please remember
• 1:28
  to write, review and subscribe. And if you want to support
• 1:32
  Centennial in other ways, feel free to check out the buy me a
• 1:35
  coffee link in the description. OK, let's get into the stories
• 1:39
  now. The first story for today is the
• 1:42
  weekly US TikTok ban update. That is honestly against my
• 1:46
  will, but it has been reported that TikTok is developing a new
• 1:51
  app for US based users ahead of the potential sale.
• 1:54
  I'm not going to get into the nitty gritty of the band since I
• 1:56
  literally did a refresher of it last week since Trump shared
• 2:00
  that there was a potential buyer on the horizon.
• 2:02
  So definitely feel free to check out last week's mini tech scroll
• 2:06
  if you are interested in that. But the update for this week is,
• 2:09
  according to reporting from The Information quote, the company
• 2:13
  TikTok has developed a plan to launch the new TikTok app, known
• 2:17
  internally as M2, to US app stores on September 5, according
• 2:22
  to two people with knowledge of the matter.
• 2:24
  So US TikTok users will see only have to download a new app to
• 2:28
  continue using the platform, but the current version apparently
• 2:32
  will be available until March next year.
• 2:34
  And for reference, there are reportedly around 170 million US
• 2:39
  based TikTok users. So the prospect of all of them
• 2:42
  moving across is honestly rather unlikely, I'd say, but I'm sure
• 2:47
  a lot of them will. So as I've mentioned before, the
• 2:51
  Trump administration has prolonged the TikTok ban from
• 2:54
  coming into force. The latest extension size in
• 2:57
  mid-september, which is set to be a couple of weeks after the
• 3:00
  new app is set to launch according to the current
• 3:03
  timeline that's being reported. Obviously, a lot is up in the
• 3:07
  air about the app itself, from questions around whether it will
• 3:10
  even happen to how it will operate, and of course,
• 3:13
  timelines. So while I acknowledge that we
• 3:15
  are largely operating on hypotheticals, this whole thing
• 3:19
  raises some really interesting questions.
• 3:21
  First off, it remains unclear if users outside the US will be
• 3:25
  able to download and use this new version of TikTok.
• 3:28
  And in a similar vein, we don't know if this new app will
• 3:31
  continue to show content from around the world or it will only
• 3:35
  be users of that app, which is presumably going to be US based
• 3:40
  content creators. I'm curious though, if you live
• 3:43
  in the US, let me know if you would rather to continue to get
• 3:46
  fed international content or you don't really care.
• 3:49
  I will say though, I think the prospect of having AUS specific
• 3:53
  TikTok is really interesting and it kind of reminds of China
• 3:57
  having its own TikTok equivalent being Doyen.
• 4:00
  Doyen doesn't show TikTok content and TikTok doesn't show
• 4:04
  content from Doyen so obviously this is a different political
• 4:07
  and cultural context, but I am wondering if this new app would
• 4:10
  follow a similar kind of format. Another thing I am curious about
• 4:15
  is how this TikTok based app will play out when it comes to
• 4:18
  things like community guidelines and free speech, especially if
• 4:22
  they carry over from Tik Tok's existing guidelines.
• 4:26
  If the platform does end up being US currently, I think it
• 4:29
  would allow for a lot more control and potentially
• 4:32
  suppression on the app. Obviously if this is the case,
• 4:36
  it really limits the user experience when it comes to the
• 4:39
  free flow of communication and ideas.
• 4:41
  As someone who lives outside the US but also is invested in US
• 4:48
  politics, culture, and society. Since I literally did American
• 4:51
  studies at university, I will say I think it is fairly common
• 4:56
  knowledge that the US can be not always, but it can be pretty
• 5:00
  insular when it comes to culture and just a general understanding
• 5:03
  of the world. So if this new app is restricted
• 5:07
  to US based content, I don't love those prospects
• 5:10
  necessarily. Another thing, sorry, I feel
• 5:13
  like I'm just going on a ramble, but while I don't think other
• 5:16
  countries have the capacity to be doing what the US is doing, I
• 5:21
  can't help but wonder if this is something that other countries
• 5:23
  would want to do. Maybe out of national security
• 5:26
  fears or a desire to control the content that is shown on the
• 5:30
  platform. But with that, I'm wondering if
• 5:32
  this is something that we will see become more popular like
• 5:36
  having apps per country as things like cacking and data
• 5:39
  theft and mining become key methods to gather information.
• 5:44
  Like I said, I don't think many countries would have the same
• 5:47
  power or capacity to do so, but it is an interesting trend that
• 5:52
  could be starting. And I guess my last question
• 5:55
  around the development of this kind of app is what it means for
• 5:59
  Internet culture. Like I said, we don't have a lot
• 6:01
  of information about what this would look like, but Internet
• 6:05
  culture is so US focused and US driven, and the US is a massive
• 6:10
  market. So having an isolated TikTok app
• 6:14
  could really change Internet culture as we know it.
• 6:16
  And I think it puts a lot of international creators run the
• 6:20
  line because you can have creators who are based outside
• 6:23
  of the US who have a majority US audience because they are such a
• 6:27
  big market. I'm curious to see what that
• 6:29
  means for potential brand deals and how brands and other
• 6:33
  companies who are very online navigate these developments
• 6:37
  currently. Though, to finish this off, as
• 6:39
  I've said, it's also up in the air, but it is definitely a
• 6:42
  development that is important and something we should take
• 6:46
  note of. The next story for today was
• 6:48
  actually suggested to us in a comment and I've been seeing
• 6:50
  more and more discourse about it ever since then.
• 6:53
  But basically I wanted to cover how there has been a seemingly
• 6:57
  dramatic rise in AI generated racist slop on TikTok.
• 7:01
  Luckily, this isn't actually something I had seen on my for
• 7:04
  you page, which I'm honestly very grateful for, but I have
• 7:08
  seen many people talk about the kind of content recently.
• 7:11
  In fact, Media Matters, which is a nonprofit Research Center and
• 7:15
  kind of media watchdog which is based in the United States,
• 7:19
  recently a piece covering this issue.
• 7:22
  According to the Pieces author Abby Richards, many of these
• 7:26
  videos are created with Google's VO3, which is a text to video
• 7:31
  generator. As she writes quote, the
• 7:33
  publicly available AI generator currently has a limit of eight
• 7:37
  second videos and Google reportedly plans to integrate
• 7:40
  the video generator into YouTube Shorts.
• 7:43
  The videos that Media Matters has identified on TikTok all ran
• 7:46
  a maximum of eight seconds long. They also had a video label in
• 7:50
  the corner, used hashtags or captions, or had usernames
• 7:54
  related to VO3 or AI. Now, I don't feel like it's 100%
• 7:59
  necessary for me to give specific examples of what these
• 8:03
  Tik Toks entail because they are quite upsetting.
• 8:06
  I will say though, that so much of this content is disgustingly
• 8:10
  racist, pulling from extremely harmful stereotypes and painful
• 8:14
  parts of history. And with that being said, it's
• 8:16
  obviously all very dehumanized. According to the Media Matters
• 8:21
  report, there's also increasing numbers of discriminatory and
• 8:25
  just problematic content directed towards various ethnic
• 8:28
  minorities and other racial groups, specifically Asian,
• 8:31
  Latino, and Jewish people. Now I did some searching and
• 8:35
  scrolling to see the types of content that is being circulated
• 8:39
  on the platform because like I said, it's not something that
• 8:42
  was organic to my feed, which like I said, I am grateful for
• 8:46
  and I believe is a green flag for my algorithm.
• 8:49
  I will say though that it was extremely easy to find.
• 8:52
  You could virtually search up VO3 AI on TikTok and these
• 8:56
  videos will come up. There were also no content
• 8:59
  warnings or restrictions on virtually anything.
• 9:02
  So that is clearly a horrific failure on behalf of the
• 9:05
  platform. I will also say that I think
• 9:08
  it's also a failure on behalf of the tools that many of these
• 9:11
  users are using like VO3 that they should be even allowed to
• 9:15
  generate this sort of content. There definitely needs to be
• 9:18
  more restrictions on that. Unfortunately, so many of these
• 9:21
  videos had a lot of views and engagement.
• 9:24
  For example, one anti-Semitic AI generated clip I saw had around
• 9:28
  1 million views and almost 100K likes.
• 9:30
  Also I would like to add that the only disclaimer put forth by
• 9:34
  TikTok on a video that I saw was one AI generated video that was
• 9:39
  essentially poking fun at the Holocaust.
• 9:41
  A little window in the corner did pop up that read quote,
• 9:45
  learn facts about the Holocaust and that directed uses to about
• 9:49
  holocaust.org, but that was literally the only disclaimer I
• 9:53
  saw when I was researching for this episode.
• 9:55
  So the commenter that prompted us to to cover this story
• 9:59
  specifically mentioned the rise of harmful content against black
• 10:03
  women, which is something that I also very quickly noticed.
• 10:06
  I came across so many videos that were also seemingly made by
• 10:10
  VO3 that fed into some of the most harmful stereotypes, and I
• 10:14
  didn't see many comments that even condemned this kind of
• 10:17
  content when I was scrolling through, which was really
• 10:19
  disappointing. Obviously I can't speak to the
• 10:22
  experience of seeing this kind of content as a black woman and
• 10:26
  how it would impact you personally, but I can only
• 10:29
  imagine how confronting and upsetting it would be to be
• 10:34
  seeing your likeness Co opted and then subsequently used to
• 10:38
  feed into really harmful tropes that would ultimately impact how
• 10:41
  you move through the world. So with that being said, I did
• 10:44
  want to include a couple of Tik Toks posted by various black
• 10:48
  creators speaking to their opinions and experiences seeing
• 10:53
  this kind of content pop up on there for you page.
• 10:55
  The first one I saw was by TikTok user Neo Sol NIA and she
• 11:00
  said that she believes black women in this scenario are being
• 11:03
  used as quote UN quote puppets, whether that be to support
• 11:07
  causes that they don't support or just to show division in the
• 11:11
  black community. Right, I need y'all to tell me
• 11:13
  if I'm tripping or not. Y'all don't find it weird that
• 11:16
  the 92% the black women said that we were going to sit back,
• 11:20
  not involve ourselves in any of this stuff going on because we
• 11:23
  did our part during the election.
• 11:25
  We're not going to overextend ourselves for people who did not
• 11:28
  care to vote with us. Y'all don't find it weird that
• 11:32
  all of a sudden we're starting to see like all these weird AI
• 11:35
  videos either A using our likeness to express support for
• 11:42
  different things that we are not in support of, that we didn't
• 11:45
  say we were in support of, or be using our likeness to further
• 11:51
  degrade us by perpetuating harmful stereotypes.
• 11:56
  And I'm like, who is this for? I feel like they're just trying
• 11:59
  to use black women as puppets. Like if they can't willingly get
• 12:03
  us to be involved in this, they're going to figure out a
• 12:05
  way to to put us in this against our will.
• 12:08
  Another TikTok that I saw was describing how these Tik Toks
• 12:12
  are used to create a really harmful narrative against black
• 12:15
  women. Is it me or somebody using AI
• 12:18
  characters to create a narrative about black women?
• 12:21
  Take a look at this. Back in my days, we wouldn't do
• 12:23
  these hairstyles. You keep worrying about my hair
• 12:26
  because you don't got none over to stop it there because they're
• 12:28
  about to have the little girl cuts in which is not realistic.
• 12:31
  Now take a look at this next video.
• 12:33
  You are the most ugliest woman I've ever seen.
• 12:36
  Girl. Shut yo Dusty expired once
• 12:38
  again. I had to stop it because it was
• 12:39
  about to get profane. My question is, who's creating
• 12:42
  these caricatures? Because that's what they are, a
• 12:45
  black woman. Some people may try to dismiss
• 12:47
  this as playful or just social media.
• 12:49
  I actually think that it's quite harmful.
• 12:52
  Why do black women always have to be depicted as angry, as
• 12:55
  loud, as ignorant? Obviously there are so, so many
• 12:58
  issues with the fact that this kind of content is circulating
• 13:01
  online. I believe it is really important
• 13:04
  to talk about this for a number of reasons.
• 13:06
  Obviously, intersectional oppression and discrimination is
• 13:10
  something that is always important to bring up because
• 13:13
  obviously somehow so many white people can't seem to comprehend
• 13:17
  it. But aside from this, it's also
• 13:20
  important because it seems like this kind of discriminatory AI
• 13:24
  slop is not going to slow down because the platforms are
• 13:27
  literally doing virtually nothing to stop it.
• 13:30
  For example, this kind of content has been a major problem
• 13:33
  on reels for a while now, and it is something that we actually
• 13:36
  covered in our AI slop deep dive.
• 13:39
  In that video, we made the point that platforms have done a
• 13:43
  really, really bad job at moderating this kind of content
• 13:46
  because these days everything comes down to engagement.
• 13:50
  We all know how creators are rewarded for quantity and not
• 13:54
  quality, and obviously AI videos are just so easy to churn out.
• 13:58
  So the commodification of content strikes again, and for
• 14:02
  the worse. Also, I think another important
• 14:04
  factor to note with this story is how these discriminatory
• 14:08
  clips against various minority groups can be used as a form of
• 14:12
  propaganda, even though many users making this kind of
• 14:16
  content would claim that it's a joke.
• 14:18
  Even though it's definitely not funny.
• 14:20
  I feel like the two creators I referenced earlier definitely
• 14:24
  touched on this, and we can see how AI content feeding into
• 14:27
  stereotypes only upholds dangerous narratives about
• 14:31
  various minority groups. And throughout history this is
• 14:34
  being done time and time again, obviously not through AI but
• 14:38
  through other methods. Off the top of my head, I think
• 14:40
  cartoons are probably one of the best examples.
• 14:44
  Whether they were anti black in the United States or
• 14:46
  anti-Semitic in Nazi Germany, it kind of serves the same
• 14:50
  function. And arguably this AI generated
• 14:53
  slop is the 2025 version. Anyway, I'm definitely curious
• 14:57
  to hear what you guys think about the proliferation of
• 15:00
  racist AI slop on social media platforms and what we can do to
• 15:04
  stop it. I thought it was a really
• 15:07
  important story to touch on and I hope I covered everything and
• 15:10
  did it justice. And for our last story for
• 15:13
  today, this is a bit of a shorter one, but it is also
• 15:16
  about AI. Every story is either about the
• 15:18
  TikTok ban or AI these days, but basically a band is going
• 15:22
  somewhat viral for claiming that Meta AI made-up a fake
• 15:26
  controversy about them. The band called Dutch Interior
• 15:29
  posted a video to Real Tick Tock with the title text reading
• 15:33
  quote AI lied about our band. The video now has accumulated
• 15:37
  450K views on reels and around 93 K on TikTok at the time of of
• 15:43
  recording. Basically the band claims that
• 15:46
  after posting a video that was essentially about a sound guy
• 15:49
  quote UN quote crashing out the auto populated search bar at the
• 15:53
  top of the video read quote Dutch interior band controversy.
• 15:57
  So they made a follow up video responding to the generated
• 16:02
  search and they included a screenshot of what meta AI said
• 16:05
  the controversy was and I'll read it for you now.
• 16:08
  Dutch Interior, an indie rock band from the Netherlands, was
• 16:11
  involved in a controversy surrounding their name and its
• 16:14
  alleged association with white supremacy.
• 16:17
  In 2020, the band announced that they would be changing their
• 16:20
  name due to concerns that it was being misinterpreted and linked
• 16:23
  to far right ideologies. The band stated that their name
• 16:26
  was actually inspired by a phrase from a poem and had no
• 16:30
  connection to extremist views. Despite this, the controversy
• 16:33
  led to the bands music being removed from some streaming
• 16:36
  platforms and sparked wider debate about the importance of
• 16:39
  cultural sensitivity in the music industry.
• 16:41
  In at the video they claim none of this was true, even down to
• 16:45
  the fact that they are not from the Netherlands.
• 16:48
  I did do a quick search to see if I could find a Dutch interior
• 16:53
  band that was based there, but I couldn't really find anything.
• 16:56
  This was the only band that was coming up and I saw no real
• 17:00
  connection to the Netherlands apart from the fact that Dutch
• 17:03
  as a language is spoken there. So I'll insert the band's
• 17:06
  explanation now as well so you can hear it from the horse's
• 17:09
  mouth. Meta AI invented a fake
• 17:12
  controversy about our band and then promoted it on our
• 17:14
  Instagram to thousands of people.
• 17:16
  We're a band called Dutch Interior, and we recently posted
• 17:19
  a reel about a sound guy crashing out on us during sound
• 17:22
  check. For some reason, Meta AI added a
• 17:24
  suggested search to the top of the comments section called
• 17:27
  Dutch Interior Controversy. When we noticed the link, we
• 17:30
  clicked on it, only to find that AI had invented a story about us
• 17:33
  being from the Netherlands we're not, and how our name has ties
• 17:36
  to white supremacy. It's actually based on a series
• 17:39
  of paintings. The AI then went on to claim
• 17:42
  that our music was removed from streaming platforms because of
• 17:45
  this controversy. Again, none of this is true.
• 17:48
  We just we don't know how this can be allowed.
• 17:50
  Our video got 10s of thousands of views, which means thousands
• 17:53
  of people who have never heard of us could have been exposed to
• 17:55
  this misinformation. Now we're even hesitant about
• 17:58
  posting this. Who knows what the AI will make
• 18:00
  up in the comments of this video?
• 18:02
  This video has sparked a lot of conversation with many people
• 18:05
  talking about just how harmful this kind of auto populated
• 18:09
  search function is, and we have also talked about this before,
• 18:12
  but mostly in the context of TikTok.
• 18:14
  Creators have long found fault with the search bar feature on
• 18:19
  TikTok, with many questioning its accuracy and even its
• 18:22
  utility and reliability. There have been numerous
• 18:26
  instances on TikTok of the search bar hallucinating context
• 18:30
  and starting to fuel rumors that ultimately led to virality.
• 18:34
  And what makes it so much worse is that many Gen.
• 18:36
  Z's, including myself, use social media as a search engine
• 18:41
  these days. Well, I think this is probably
• 18:43
  more true for TikTok than it is Instagram.
• 18:46
  It feels so insidious that these platforms would lean into this
• 18:50
  function without what feels like any guardrails or checks and
• 18:54
  balances to make sure that they aren't spreading misinformation.
• 18:57
  Especially because much of this AI generated search and auto
• 19:01
  population is meant to be a form of fact checking.
• 19:04
  On top of this, the platforms are so secretive about how these
• 19:09
  AI search functions even work and how auto populated search
• 19:13
  bars decide what pops up. Aside from this conversation
• 19:16
  though, I will say that many people are skeptical of the band
• 19:20
  and whether they are telling the truth about their experience.
• 19:23
  Some have left comments claiming that the whole story is fake and
• 19:26
  that they are sharing this video and making it up for clout.
• 19:29
  I will say when I went to their profile to see for myself, the
• 19:33
  initial video was up but nothing populated in the search bar for
• 19:36
  me. This could be because my
• 19:38
  Instagram needs to be updated or because I'm based in Australia
• 19:41
  that there are different search results and functions.
• 19:44
  On top of this, when I searched up Dutch interior band
• 19:47
  controversy for myself, the text that they shared in the video
• 19:51
  didn't come up for me. Aside from the possibility of
• 19:55
  this being a hoax, there is a possibility that things were
• 19:59
  changed on Meta's end to adjust the misinformation.
• 20:03
  Ultimately though, I think it's fair to give the band the
• 20:06
  benefit of the doubt and assume they aren't lying.
• 20:09
  But even if they were, I think it has sparked a number of
• 20:13
  important conversations that we should be having around AI on
• 20:16
  social media and. The power of these auto
• 20:20
  populated search bars to spread misinformation.
• 20:23
  If it didn't happen to them though, it definitely will and
• 20:27
  has happened to other creators and users on the platform.
• 20:31
  So I don't think it's harmful to be having these conversations.
• 20:34
  And at the end of the day, we all know that AI does tend to
• 20:37
  make up information or context from time to time.
• 20:41
  Anyway, that is everything from me today.
• 20:43
  I really hope you enjoyed this week's episode of Mini Tech
• 20:46
  Scroll. See you next Wednesday and
• 20:49
  Lauren will speak to you tomorrow.
• 20:51
  Thanks. Bye.