Latest / Mini Scroll / MINI TECH SCROLL: The "Clip Economy" obsession, Ground News AI drama, Elon Musk vs OpenAI + YouTube targets deepfakes

Transcript

MINI TECH SCROLL: The "Clip Economy" obsession, Ground News AI drama, Elon Musk vs OpenAI + YouTube targets deepfakes · Mini Scroll

• 0:00
  Coming up today on Mini Tech Scroll, the rise of the clipping
• 0:03
  economy, Tik Tok's problem with an AI news platform, Elon Musk
• 0:07
  loses his case against open AI, and YouTube expands deepfake
• 0:11
  protections for users over 18. Hey everyone, and welcome to
• 0:15
  Mini Scroll, the daily Internet culture podcast bring you the
• 0:18
  biggest stories from social media, the creator economy, and
• 0:21
  the digital space every Monday through Thursday and sometimes
• 0:24
  Fridays. My name is Annabelle and I am
• 0:26
  your Wednesday Mini Tech Scroll host and also the deputy editor
• 0:30
  of Australian youth digital media brand Centennial World.
• 0:34
  Welcome back to Mini Tech Scroll for another week.
• 0:36
  God bless. I do not really have any updates
• 0:39
  slash requests from you guys apart from the regular.
• 0:42
  Please remember to rate, review and subscribe or check out our
• 0:46
  Patreon for more behind the scenes parasocial content if
• 0:49
  you're so inclined. But yes, let's just get right
• 0:52
  into the content for today. The first thing I wanted to
• 0:54
  cover for this week's episode is the conversation going around
• 0:57
  online at the moment about Clippers and the so-called
• 1:00
  clipping economy. I will say I have covered this
• 1:03
  before in mini tech scroll towards the end of last year and
• 1:06
  I will say I said the term clipping economy and I was like
• 1:09
  is that even a thing? And now everyone's talking about
• 1:11
  it So we are ahead of the game. But since this is absolutely
• 1:16
  everywhere for me and I presume some of you guys, I thought this
• 1:19
  would be worth revisiting and sharing some new analysis and
• 1:22
  thoughts. Of course, so much of the
• 1:24
  conversation on social media about clipping culture seems to
• 1:27
  be coming from a couple of pieces of content that were
• 1:30
  recently published on the topic. So this includes a podcast
• 1:34
  episode published by the Galaxy Brain podcast, which is like
• 1:38
  associated with the Atlantic called how short form clips took
• 1:41
  over the Internet. And that was published in late
• 1:43
  April. The Verge then issued a piece
• 1:46
  titled the clipping on May 6. And then perhaps the most viral
• 1:51
  is this article by Vulture with the headline quote, the feed is
• 1:54
  fake. And that was published around 5
• 1:56
  days ago now. So it is still very much an
• 1:59
  active conversation. The gist of these articles and
• 2:02
  the resulting commentary content on the topic is that a new part
• 2:06
  of the creator economy has kind of emerged, which is the clip
• 2:09
  economy. And I'm sure you would have all
• 2:12
  noticed how much social media is dominated by clips or teasers of
• 2:16
  more long form content these days, much of them coming from
• 2:20
  like YouTube videos or video essays and live streams.
• 2:23
  And upon first glance, you may think that these clips are being
• 2:27
  posted by fans hoping to kind of maybe boost their favorite
• 2:31
  creators or live streamers or just share a funny moment.
• 2:35
  But as many of the commentators have noticed, these clips are
• 2:39
  actually way more strategic and usually posted by dedicated
• 2:43
  Clippers from seemingly random accounts who are in fact paid
• 2:47
  for doing this work, which is not disclosed to viewers, which
• 2:51
  is a problem in and of itself. So to give you a little bit of
• 2:54
  context, in the piece by Vulture, the journalist whose
• 2:57
  name is Lane Brown, monitored a variety of Discord clipping
• 3:01
  service where announcements, I guess, slash opportunities for
• 3:06
  clipping campaigns had been shared.
• 3:08
  And he mentioned that Justin Bieber's Coachella performance
• 3:11
  was one campaign he spotted. Even though he did reach out to
• 3:14
  Justin's team and representatives, they did not
• 3:18
  respond to multiple calls or emails.
• 3:20
  But As for why someone as famous as JB would potentially pursue a
• 3:25
  clipping campaign, Brown notes that this is just simply how the
• 3:29
  Internet and virality works today.
• 3:31
  Quote. A lot of what clipping does is
• 3:33
  help good artists find the audiences who would have liked
• 3:36
  them anyway by accelerating the early excitement just enough to
• 3:39
  push them past the algorithmic threshold that decides who gets
• 3:43
  discovered and who doesn't. And I definitely do buy into
• 3:46
  that. I do think the algorithm would
• 3:48
  push out these promotional clips to people who have expressed
• 3:51
  interest in these topics or artists in the 1st place.
• 3:55
  But what is interesting though is that it's ultimately changing
• 3:58
  the way we behave and communicate online because it's
• 4:01
  been proving that clipping works for these companies and
• 4:04
  celebrities as a strategy to find viral success.
• 4:07
  So now the average user is basically doing the same thing
• 4:10
  on a smaller scale. Like we are literally all in our
• 4:13
  bot era just pumping out content.
• 4:15
  And Mia Sarto for The Verge made a similar argument to this where
• 4:19
  she noted how clipping is a way for companies, slash managers or
• 4:23
  whoever really to get around the black box that is the algorithm.
• 4:27
  As she writes quote, deploying Clippers allows companies to
• 4:30
  gamble on content at scale without paying a network of
• 4:34
  contractors upfront. Why bet once when you could bet
• 4:37
  50 times? And before we get into how all
• 4:40
  these Clippers are paid and how the economy is kind of flowing,
• 4:44
  I do think we have to know that clipping has obviously been
• 4:47
  around for years now. Gamers have obviously been
• 4:49
  clipping for ages, but this technique really gained
• 4:53
  popularity kind of out of the gaming space in 2022 after
• 4:57
  Andrew Tate had fans post clips of his podcast across social
• 5:01
  media, which made him seem extremely influential and
• 5:04
  potentially more influential than he actually was.
• 5:07
  And I'm not trying to undermine how harmful he is by any means,
• 5:10
  but the amount of content that was circulating of him was
• 5:13
  arguably inorganic. Since then, we've seen other
• 5:16
  manosphere creators follow suit, and perhaps the most recent and
• 5:20
  prolific example of that is of course, Clavicular.
• 5:22
  Hundreds, if not thousands of clips circulate of him online.
• 5:26
  Everything I know that Clavicular is literally against
• 5:29
  my will, but when you go look at his streams, there are far less
• 5:33
  viewers on his streams than the clips.
• 5:36
  In saying that, it's also being reported that much of
• 5:38
  Clavicular's success has been bankrolled by Kik because it
• 5:43
  benefits the platform too if he goes viral in this way, whether
• 5:46
  it is short form content or long form content.
• 5:49
  For context, Kik shared clipping stats for March to April of this
• 5:53
  year, noting that 1737 Clippers made more than 309,000 videos,
• 6:00
  and Clavicular, perhaps not surprisingly, was the most
• 6:03
  clipped, with 69,000 video clips from his live stream published
• 6:07
  across social media in just that one month period, and that
• 6:11
  accumulated to over 2.2 billion views.
• 6:15
  And the VAG piece does note that much of this content is slop.
• 6:18
  Like it literally is. Not.
• 6:19
  Adding anything to our social media quote clip is by and large
• 6:24
  add nothing of substance to the original work.
• 6:26
  No analysis, no response, not even filters and music and cards
• 6:30
  that fan edits often have. It's the most boring kind of
• 6:33
  content spliced purely for the algorithm.
• 6:35
  And when it comes to paid Clippers, it's hard to separate
• 6:38
  the aesthetics from those often very young editors who turn up
• 6:42
  content. With that, let's get a little
• 6:44
  bit into how these Clippers are being paid and how the economy
• 6:47
  is operating in general. So there seems to be certain
• 6:50
  platforms slash services where companies can list campaign
• 6:54
  opportunities. So like, you can kind of think
• 6:56
  of it as like a marketplace where Clippers can then opt into
• 7:00
  these campaigns. The way these people are paid
• 7:02
  though is kind of where things get sauce in my opinion because
• 7:06
  these Clippers are paid based on the amount of views their
• 7:09
  content gets. The piece for The Verge gave an
• 7:12
  example from the service clipping.net, which is one of
• 7:15
  the most popular, that offered Clippers $150.00 for every
• 7:19
  100,000 views for companies and brands.
• 7:23
  This is obviously a good model because they're only paying the
• 7:26
  Clippers for what they get and how the content performs.
• 7:30
  But from the video editors perspective, it is a little bit
• 7:33
  ominous since they're not getting paid necessarily for the
• 7:36
  time they put into the work. They're only getting rewarded on
• 7:40
  the outcome. And that definitely feels like a
• 7:43
  bad omen slash sign for any kind of creative industry that is
• 7:49
  operating in the digital media space.
• 7:51
  Obviously though, these Clippers do certain things to try and
• 7:55
  guarantee that their clips go viral.
• 7:58
  And according to Clipping Net, TikTok and YouTube Shorts are
• 8:02
  the best platforms to post clips on because they have the highest
• 8:05
  chance of going viral. And there are, of course, other
• 8:09
  clipping platforms aside from Clipping Net.
• 8:11
  I did touch on this last time I covered this phenomenon, but Mr.
• 8:15
  Beast has also jumped on the bandwagon, launching a similar
• 8:18
  clipping marketplace called Viro last year.
• 8:20
  OK, now that we've covered like the origins of clipping and how
• 8:25
  it's currently operating, I think it's worth talking about
• 8:28
  the people that may be Clippers. There seems to be a general
• 8:31
  consensus that the people who are making clips are young
• 8:35
  people who are kind of looking for a way to just make extra
• 8:37
  cash. This does seem exploitative in
• 8:40
  some ways because young people are often the ones who are
• 8:43
  looking to build up their career and build up their finances.
• 8:47
  So they may be more kind of desperate, for lack of a better
• 8:51
  term than someone who is a little bit more established.
• 8:54
  And because clipping relies on paying per view, it often is
• 8:59
  kind of a race to the bottom and relies on people to do work for
• 9:05
  potentially little to no reward. In a recent piece for the
• 9:08
  Garbage Day newsletter our favourites here at Centennial
• 9:12
  World, Ryan Broderick investigated a group running a
• 9:15
  clip sharing operation out of Rwanda.
• 9:18
  He spoke with 1 user in this group and as he wrote quote, the
• 9:22
  user identifies online as a student at the University of
• 9:25
  Rwanda. But they told Garbage Day that
• 9:27
  sharing video clips is their job.
• 9:29
  They said they often get paid to post videos and that the bulk of
• 9:32
  those offers come from AI companies.
• 9:34
  Obviously this is a little bit different to something like
• 9:37
  clipping.net, but I thought it was really interesting because
• 9:40
  we don't have the exact demographic breakdown on who the
• 9:44
  Clippers are, but there is pretty widespread speculation
• 9:47
  that a significant portion are based in the global S where the
• 9:52
  US dollar pay rates might go much further than they would in
• 9:56
  the US, but that doesn't necessarily mean they reflect A
• 9:58
  livable wage. I have kind of touched on these
• 10:01
  themes in mini tech before as well, specifically in relation
• 10:05
  to content moderators, where I reported on human moderators who
• 10:09
  are forced to watch hours of harmful content for big tech
• 10:13
  companies who do not really compensate them appropriately.
• 10:18
  And clipping is arguably maybe a softer version of that same
• 10:23
  structure because it feels more passive and almost like a side
• 10:26
  hustle. But there is still that kind of
• 10:28
  element of extraction of a labour there because the labour
• 10:32
  is made cheap by geography for the Clippers who are based in
• 10:35
  the global South and the company slash clipping services
• 10:39
  obviously benefit from that financially, but there is also
• 10:42
  less of an accountability factor.
• 10:44
  Let's move on to the next story. And I wanted to cover everything
• 10:46
  that has been going on with ground news on TikTok.
• 10:50
  So this back and forth has really been going on for
• 10:53
  probably about the last two weeks, but it only recently came
• 10:57
  across our desk, so we thought it would be interesting to cover
• 11:00
  today. Before we get into the
• 11:01
  discourse, though, let me give you some context on Ground News.
• 11:05
  So Ground News bills itself as quote, a platform that makes it
• 11:09
  easy to compare new sources, read between the lines of media
• 11:12
  bias, and break free from algorithms to summarize their
• 11:16
  mission statement in one sentence.
• 11:18
  It's basically about developing a better way to navigate an
• 11:22
  increasingly polarized and partisan media landscape.
• 11:26
  And while this is true across the world, I do feel that ground
• 11:29
  news is quite US centric. So how this works is that if you
• 11:33
  go onto the ground news website, you'll see all the major news
• 11:36
  stories with a little bar indicating the bias associated
• 11:40
  with the coverage. This little bar kind of gives
• 11:43
  you the percentage of the coverage that might be
• 11:45
  left-leaning, right leaning, or just the centre.
• 11:48
  Then when you click on a story, you can toggle between the left,
• 11:52
  right and center views to kind of see the difference in
• 11:55
  reporting and talking points. And there is also a bias
• 11:59
  comparison section that provides insight by ground AI and how
• 12:04
  ground news calculates. The bias is decided based on an
• 12:08
  average rating of quote three independent news monitoring
• 12:12
  organizations, which take into consideration things like
• 12:15
  wording, story choices, and political affiliation.
• 12:18
  So ground news has been around for quite a while.
• 12:20
  It's not something I regularly check, but I have heard of it
• 12:24
  before. However, the discourse on TikTok
• 12:26
  really started earlier this month when TikTok creator at
• 12:29
  Total Hypocrisy took to the platform, calling out a fellow
• 12:33
  creator for partnering with Ground News.
• 12:35
  She took issue with the partnership because of the way
• 12:38
  Ground News uses AI, and I'll insert a bit of her video here.
• 12:42
  Let me talk about my experience with Ground News so that you can
• 12:46
  kind of be the judge of who this publication is.
• 12:50
  At the end of February, I was contacted by Ground News and I
• 12:54
  was offered this influencer marketing opportunity.
• 12:58
  And just to kind of cover how they explained Ground News to me
• 13:03
  was it says to provide you with some context next about us
• 13:06
  Ground News is not a news publisher.
• 13:09
  So they are not a news publication.
• 13:11
  They are not bringing you first hand news.
• 13:13
  We are on a mission to bring transparency to the media by
• 13:16
  offering both sides of the story.
• 13:18
  We do not publish or censor news.
• 13:20
  We compile it into one place for individuals to draw on their own
• 13:23
  conclusions. I passed and I told them it's
• 13:26
  not the right fit for me. Here's why.
• 13:28
  The article was published in the Columbia Journalism Review in
• 13:31
  September of 2025. Talking about ground news, this
• 13:35
  might be kind of cut off at the top, but it says the tagline of
• 13:38
  ground news is quote. See every side of every news
• 13:41
  story. It aggregates articles from more
• 13:43
  than 40,000 free and subscription outlets, then uses
• 13:48
  AI to publish a daily average of about 30,000 summaries of news
• 13:54
  events from these articles. I do not want to partner with a
• 13:58
  company that just uses AI to process what they believe to be
• 14:03
  the summary of the day's news. Ground News is not a
• 14:07
  publication. It is not hiring journalists.
• 14:11
  It is not abiding by the journalist creed.
• 14:13
  It is not even close. It is an organization that uses
• 14:17
  AI to summarize information and then charge the people that want
• 14:22
  to use their publication in quotation marks to see what is
• 14:26
  going on. She definitely raises some valid
• 14:29
  concerns in her Tik Toks, and there have been other criticisms
• 14:32
  of ground news that I've seen over the years, particularly
• 14:35
  that it legitimizes extremist reporting and fringe opinions.
• 14:39
  Nevertheless, the conversation that started off the back of her
• 14:42
  video was about how AI should be used in media slash journalism
• 14:47
  and what it means for the audience. 1 TikTok user who
• 14:51
  lined with total hypocrisy noted how promoting an AI tool that
• 14:55
  summarizes news is unethical because AI can be inaccurate and
• 15:00
  biased as we all know. She also argued that the name
• 15:03
  ground news was misleading in and of itself because they don't
• 15:07
  often news, it's rather an AI generated summary.
• 15:11
  One of my big issues with this ground news is the fact that
• 15:14
  it's calling itself Ground News, even though at the same time it
• 15:18
  does not claim to be a news source.
• 15:21
  It's just compiling data and presenting a summary to you.
• 15:25
  So innocent What? What's the problem with that?
• 15:27
  Well, you know that people are going to go to you as a news
• 15:29
  source. You know that it's going to
• 15:30
  trust the summary it's giving you.
• 15:32
  Even though, yes, people could and should go and verify the
• 15:36
  information that they're reading, they wouldn't be using
• 15:38
  a summary tool if they were going to go to all of those
• 15:42
  sources and check for accuracy. They knew that when they named
• 15:46
  it ground news. It's manipulative.
• 15:47
  They understand how a human brain works, and they're using
• 15:50
  it to their advantage. Meanwhile, other TikTok users
• 15:53
  have kind of pushed back against this camp, essentially
• 15:55
  questioning why ground news is problematic if they don't bill
• 15:59
  themselves as a news publisher, and if the audience is using it
• 16:03
  as a news aggregator of sorts. These users argue that engaging
• 16:08
  with AI in its capacity as a summary tool is probably one of
• 16:13
  its better uses, and that the creators who say otherwise have
• 16:17
  kind of misrepresented what ground news is.
• 16:20
  When there's a large amount of information like I don't know,
• 16:23
  30,000 news articles per day on a given subject, it might be OK
• 16:28
  to use AI slim down. Get the general idea of what
• 16:34
  those articles are saying. It might be a good use for it if
• 16:37
  you're going to use it at all that come on, guys, that's a
• 16:41
  pretty good use for it. To point out.
• 16:44
  I don't know what Ground news calls blind spots, which are
• 16:48
  portions of a story that are not covered by a left publication or
• 16:54
  a right publication. She also seems to really want to
• 16:58
  drive home the fact that Ground News is not a publication.
• 17:00
  They don't publish, they're not real journalists, yet they don't
• 17:04
  claim to be. They're using AI in possibly the
• 17:09
  like most acceptable way that any use of AI could ever be
• 17:12
  used. I thought this debate raised
• 17:15
  some interesting questions around what actually counts as
• 17:18
  an acceptable use of AI and whether we should be engaging
• 17:21
  with it at all in spaces like journalism and media
• 17:23
  consumption. Because there does seem to be a
• 17:26
  growing wave of what I guess I would call like anti AI purists
• 17:32
  who think that consumers should be refusing to engage in AI full
• 17:37
  stop. And that is definitely a
• 17:38
  legitimate take. But at the same time though,
• 17:41
  this position is definitely becoming harder and harder to
• 17:44
  hold when AI is just literally baked into every platform and
• 17:48
  form of technology, which we literally talk about all the
• 17:51
  time. A mini tech scroll, and this is
• 17:53
  actually also something that I touched on a little bit last
• 17:55
  week when we were talking about the galbosification of AI.
• 17:58
  So I thought this whole conversation was actually quite
• 18:01
  intriguing to see where some users see the line.
• 18:05
  Like, is using AI to aggregate and summarize news inherently
• 18:09
  bad? Or is the problem kind of more
• 18:11
  about transparency, like whether the creators are being upfront
• 18:15
  about what ground news actually is when they take the
• 18:18
  sponsorship? And I feel like that latter
• 18:20
  question is not necessarily something that has popped up
• 18:23
  enough maybe in the discourse on TikTok.
• 18:26
  But I'm definitely curious to hear where you land on this
• 18:29
  debate and what you think about it.
• 18:30
  My next two stories are updates, so they're a little bit shorter.
• 18:35
  First, Elon Musk has officially lost his case against Open AI.
• 18:39
  So I did touch on this a couple weeks ago, But for some context,
• 18:43
  Elon Musk and Sam Altman were entangled in a legal battle
• 18:47
  where Elon Musk was essentially accusing Sam Altman of betraying
• 18:51
  Open AI's original mission as a startup.
• 18:54
  Basically Elon claims that Sam Altman and his Co founder
• 18:58
  tricked him I guess into giving Open AI outputs of $30 million
• 19:04
  before moving away from their original status as a non profit.
• 19:07
  Because when Open AI was launched it was kind of
• 19:10
  positioned as a company to create and develop AI for the
• 19:15
  good of humanity. Since then, Open AI has gone on
• 19:18
  to attach a for profit business to its original nonprofit and
• 19:22
  has also accepted billions of dollars from investors fame,
• 19:26
  mostly including Microsoft. So in the legal battle, Elon
• 19:30
  Musk was reportedly seeking 150 billion in damages from Open AI
• 19:34
  and Microsoft. He was also asking for the
• 19:37
  removal of Altman and his Co founder Greg Brockman from Open
• 19:41
  AI. And I spoke about this the last
• 19:43
  time I covered it, but it very much seemed like a way to punish
• 19:47
  Sam Altman and Open AI ahead of a potential IPO later this year,
• 19:52
  kind of jumping on the bandwagon of Open AI's ongoing
• 19:55
  reputational struggle and PR nightmare that it continues to
• 19:59
  be in. Anyway, to probably no one's
• 20:02
  surprise, it was reported that after less than two hours of
• 20:05
  deliberation, the jury decided against Elon Musk, saying that
• 20:10
  he had waited too long to sue. Elon Musk obviously took to
• 20:13
  Twitter to share his thoughts, saying that he does plan to
• 20:16
  appeal right in. Quote regarding the Open AI
• 20:19
  case, the judge and jury never actually ruled on the merits of
• 20:22
  the case, just on a calendar technicality.
• 20:25
  There is no question to anyone following the case in detail
• 20:28
  that Altman and Brockman did in fact enrich themselves by
• 20:31
  stealing a charity. The only question is when they
• 20:34
  did it. I will be filing appeal with the
• 20:36
  9th Circuit because creating a president to loot charities is
• 20:40
  incredibly destructive to charitable giving in America.
• 20:44
  Open AI was founded to benefit all of humanity, so even though
• 20:47
  it might not be in Elon Musk's interest to launch an appeal, he
• 20:53
  has enough money to do so. So I definitely wouldn't be
• 20:55
  surprised if this is something that we continue to see in the
• 20:58
  news cycle. The next update is that YouTube
• 21:00
  is rolling out its likeness detection technology to anyone
• 21:03
  over 18 years old. So YouTube has been slowly
• 21:07
  rolling out this AI deepfake detection tool for a while now.
• 21:10
  It's basically like their existing copyright system that
• 21:13
  will flag copyrighted content, but instead of that it's kind of
• 21:16
  scanning for your likeness and face.
• 21:18
  The company originally introduced the likeness
• 21:21
  detection system in December 2024 and that was in partnership
• 21:25
  with The Creative Artist Agency and it was seemingly first
• 21:29
  shared with a select group of creators.
• 21:31
  Since then it was expanded to other celebrities and public
• 21:35
  figures, but a couple of days ago it was announced that this
• 21:38
  feature will now be available to creators over the age of 18.
• 21:42
  Just very quickly, if there is a possible match, it will flag the
• 21:46
  content it believes to contain your identity and then creators
• 21:50
  can then view it and decide how to go forward with that.
• 21:54
  Reportedly to enable likeness detection though you will have
• 21:57
  to submit your government issued ID to YouTube, which is
• 22:01
  definitely controversial. Obviously this confirms your
• 22:04
  age, but there is long been concerns about sharing a
• 22:08
  document like this with tech platforms.
• 22:10
  We've spoken about this many times.
• 22:12
  One of the worst instances that we've spoke about this happened
• 22:15
  with the T app, which was a women focused dating platform, I
• 22:19
  guess, where women could share their experiences with certain
• 22:22
  men and to verify their account they had to share their license
• 22:27
  and it ended up all over fortune.
• 22:28
  In saying that, the reaction does seem to be mixed online.
• 22:32
  So obviously many social media users are very happy to see a
• 22:36
  platform taking initiative in helping their users control
• 22:40
  their likeness online. But others are noticing some
• 22:44
  pitfalls. For instance, the detection
• 22:46
  doesn't automatically mean removal and also that there are
• 22:50
  still protections around parity and satire content.
• 22:54
  So there are limits there and for now it only does of your
• 22:57
  face. So voice is not included, but
• 22:59
  that could come down the line. Well, that is everything from
• 23:02
  mini tech scroll today. I hope you guys enjoyed.
• 23:05
  Lauren will speak to you guys tomorrow and I will see you next
• 23:09
  week. Thanks for listening.
• 23:10
  Bye. Centennial World Podcast Network
• 23:13
  operates on Gadigal land. We would like to acknowledge the
• 23:16
  Gadigal of the Eora Nation, the traditional custodians of this
• 23:19
  land, and pay our respects to elders both past and present.