Latest / Mini Scroll / MINI TECH SCROLL: Viral AI model scandal in NZ, Polymarket's shady creator deals, Siri AI announced + OpenAI's IPO

Transcript

MINI TECH SCROLL: Viral AI model scandal in NZ, Polymarket's shady creator deals, Siri AI announced + OpenAI's IPO · Mini Scroll

• 0:00
  Coming up today on Mini Tech Scroll, the New Zealand based
• 0:03
  brand facing AI model allegations, polymarkets, shady
• 0:07
  creator partnership practices and some big tech AI updates.
• 0:11
  Hey everyone, and welcome to Mini Scroll, the daily Internet
• 0:14
  culture podcast bring you the biggest stories from social
• 0:16
  media, the creator economy and the digital space every Monday
• 0:20
  through Thursday and sometimes Fridays.
• 0:22
  My name is Annabelle and I'm your Wednesday Mini Tech Scroll
• 0:25
  host and also the deputy editor of Australian youth digital
• 0:28
  media brand Centennial World. Welcome back to Mini Tech Scroll
• 0:32
  for another week. Before we get into the stories
• 0:34
  for today though, I did just want to give a quick apology for
• 0:38
  missing last week's episode. There was some health
• 0:41
  complications within my immediate family, so I was just
• 0:44
  helping with that. But everything is OK and we are
• 0:47
  back for another week. And actually, I did not plan
• 0:51
  this when I got dressed today, but I am wearing Cowboy Carta
• 0:54
  match and I realized exactly a year ago today was when I was at
• 0:58
  the Cowboy Carta tour. So take me back.
• 1:01
  It's OK, we're doing Mini Tech Scroll now.
• 1:04
  Before we get into the stories though, Please remember to rate,
• 1:07
  review and subscribe. OK, my first story is a little
• 1:11
  bit more of a regional one. I wanted to talk about
• 1:13
  everything going on with the New Zealand St. wear brand Huffa.
• 1:17
  So this has been over my for You page for the past couple of
• 1:22
  days. And it is a little bit of a
• 1:24
  cross section between Internet culture and AI, but that is
• 1:27
  exactly what we do best here at Mini Tech Scroll.
• 1:30
  So as I've already mentioned, Huffer is a New Zealand based
• 1:33
  brand selling items like puffer jackets, knitwear, snow pants
• 1:37
  and jackets as well as some more regular basics.
• 1:41
  It has a pretty solid Instagram following with just over 80 K at
• 1:44
  the time of recording, and also it is a pretty well known brand
• 1:48
  in New Zealand. Over the past week though, Hoffa
• 1:51
  has found itself in the centre of controversy, and it all
• 1:54
  started after a social media user named Elijah, who also
• 1:57
  happened to be a model for Hoffa in the past, called out the
• 2:01
  brand for seemingly using AI models.
• 2:04
  We did reach out to both Hoffa and Elijah ahead of recording
• 2:07
  this episode, and Hoffa did not respond in time.
• 2:11
  But I did get to have a chat with Elijah and I will be
• 2:13
  including quotes from our conversation throughout this
• 2:15
  story. So Elijah has posted a few
• 2:18
  videos about the situation on his social media, with the first
• 2:21
  one on TikTok being six days ago.
• 2:24
  In it, he opens by noting that he had been blocked by Hoffa
• 2:28
  after commenting on one of the post quote.
• 2:30
  That AI model actually goes kind of hard, and what kind of
• 2:34
  prompted him to comment this was that he believed that Hoffa was
• 2:38
  using AI models in its photo shoots after he saw a picture
• 2:41
  posted by the company that was an ad that looked eerily similar
• 2:44
  to him and his brother, and they had both modelled for Hoffa
• 2:48
  before. That video is currently sitting
• 2:50
  at around 180 K views on TikTok, so it's definitely gained some
• 2:54
  traction and I'll insert the video here so you can hear it
• 2:57
  directly from his perspective. So I just got blocked by an ex
• 3:01
  client for commenting. Yo that Arkansas model actually
• 3:05
  goes kinda hard. This all started when I was
• 3:07
  scrolling past an ad and I noticed an image that looked
• 3:11
  very similar to me but as I zoomed in it wasn't me.
• 3:14
  And then I thought it was my brother cuz my brother also has
• 3:17
  modeled for this client. But then it was neither of us.
• 3:20
  It was almost like both of us combined.
• 3:22
  And that's when I realized it was AI.
• 3:24
  Now the whole premise of this video isn't the Shamer brand for
• 3:26
  using AI models, but being super shady about it.
• 3:30
  Denying that you're using the likeness of previous models,
• 3:34
  even current models that you're using, and then blocking one of
• 3:38
  your models that used to represent you because he called
• 3:43
  you out on it, I do not tolerate at all.
• 3:45
  I believe if you're going to be using AI models, then stand 10
• 3:50
  toes deep and let the people choose whether they agree with
• 3:54
  your use of AI models or not. In our conversation, I did ask
• 3:58
  Elijah how long he had been skeptical of Huff His post, and
• 4:01
  he had said it's been on his mind for about two months before
• 4:04
  making the video and bringing it to the public.
• 4:07
  He told me that because the New Zealand modelling industry is
• 4:10
  just so small, he often can recognize the models in
• 4:13
  campaigns, making it easier for him to potentially identify AI
• 4:18
  usage. But of course that isn't
• 4:20
  necessarily enough for him to make a post about it.
• 4:23
  He told me a story about one of his friends in the industry who
• 4:26
  believed that an AI image of her was posted on the HAFA Classic
• 4:31
  Instagram page. She showed Elijah the original
• 4:33
  picture and what she believed to be the AI version, where he then
• 4:37
  told me quote. I was like, OK, now that is
• 4:40
  obvious enough that I sat on it and was going through the
• 4:43
  Instagram as time went on and I was like, this is just getting
• 4:46
  worse. Which ultimately led him to
• 4:49
  share that original comment, which he positioned as an
• 4:51
  opportunity for the brand to clarify or even poke fun at the
• 4:55
  photo with him. In saying that, that does seem
• 4:58
  to be a consensus among Elijah's comment sections of the TikTok
• 5:03
  videos, with many users sharing the anti AI sentiment with a
• 5:07
  rightful concern about what it means when a models likeness is
• 5:11
  used without their consent. Elijah also in the call did say
• 5:15
  that he is very appreciative of the community that has stood up
• 5:19
  behind him, where he's actually had some pretty big creators and
• 5:22
  even a parliament member reach out to him to offer support and
• 5:25
  assistance. The members of his community
• 5:28
  have also noted that it felt quite inappropriate for a brand
• 5:32
  to block a user based on one comment, especially after they
• 5:35
  had collaborated previously. And to make matters worse,
• 5:39
  Elijah's also actually alleged that the brand has blocked his
• 5:42
  management as well as him. I just remember commenting what
• 5:45
  I commented. If I was being honest, I was
• 5:47
  trying to be a little cheeky, maybe a little sarcastic, but
• 5:51
  ultimately I was trying to find out where they stood, you know,
• 5:54
  kind of inviting them to comment and maybe be a bit playful,
• 5:58
  saying yeah, AI is the way for it, ha ha ha ha.
• 6:01
  Instead I got blocked and they deleted my comment.
• 6:04
  I miss as my agent and they told me two of the agents at RID 11
• 6:08
  who I represent had been blocked as well which was a massive red
• 6:12
  flag for me. At the time of researching, it
• 6:14
  does appear that the brand is deleting comments on other
• 6:18
  Instagram posts. I found that when I looked at
• 6:20
  certain Instagram posts, the numbered comment count under the
• 6:24
  photo did not match the amount of comments I could read, which
• 6:28
  is generally an indication that comments are being deleted.
• 6:31
  After Elijah's initial video saying that he had been blocked
• 6:35
  from Huffer's account, he has gone on to make various follow
• 6:38
  up videos on the situation, seeking to offer more context.
• 6:42
  And he expressed specific concerns about how Hoffa would
• 6:45
  seemingly post real models and then AI models, making it harder
• 6:49
  for audiences to decipher between them.
• 6:51
  And at the time of recording, Hoffa has mostly responded to
• 6:55
  these claims through media comments to various local
• 6:58
  outlets. And like I said, we did reach
• 7:00
  out, but they did not get back to us in time.
• 7:02
  But I will go through some of those comments shortly.
• 7:05
  Elijah has also confirmed with us that the brand has contacted
• 7:09
  his agent directly. He also said that a man from the
• 7:12
  brand called him directly on a Saturday morning, which he
• 7:15
  characterized as a form of potential manipulation.
• 7:18
  In the conversation, he implied that he and his child had been
• 7:21
  threatened due to the backlash. So I guess that is more of the
• 7:25
  private response. But let's get to the public
• 7:27
  response from Huffer. In a comment to the New Zealand
• 7:30
  Herald for a piece published on June 4th, the managing director,
• 7:34
  Kate Berry categorically denied the use of AI.
• 7:37
  And I'll read to you exactly what she was quoted to say,
• 7:40
  quote, no faces of our models have been changed other than in
• 7:43
  the normal computer assisted design.
• 7:45
  No one is using models that have been paid by us and changing
• 7:49
  them. In saying that, she did deny to
• 7:51
  comment on specific images and that, quote, computer design
• 7:55
  support has been used in our industry for 30 years.
• 7:58
  Kate Berry also confirmed that Elijah had been blocked from the
• 8:01
  Instagram page for breaking community standards, although
• 8:04
  she did not say which standard specifically.
• 8:07
  Elijah did then go on to pick up on one of the comments that Kate
• 8:11
  Berry made to an outlet in a TikTok video.
• 8:14
  And the comment was specifically to the Post New Zealand.
• 8:18
  In a call with the journalist, she reportedly said, quote,
• 8:21
  maybe he should just think about why he's not getting work from
• 8:24
  other people. And then she also doubled down
• 8:26
  on how computer assisted design is used in all aspects of
• 8:30
  business. In the video, he did say that he
• 8:33
  agreed to some extent with what she was saying and that he would
• 8:35
  much rather be focusing on his modelling career than doing all
• 8:39
  of this online. He also explained though, that
• 8:42
  the situation had actually gone a step further, where he had
• 8:45
  been approached by lawyers and that his Instagram account had
• 8:48
  been disabled. Obviously there is no way for us
• 8:52
  to say for sure if Elijah's Instagram account was disabled
• 8:55
  due to the situation, but the timing is definitely suspicious
• 8:59
  and I did ask him on our call about the lawyer, but of course
• 9:03
  he couldn't really talk to us too much about that.
• 9:05
  Maybe he should think about why he's not getting work, is what
• 9:09
  Kate Berry, managing director of Huffer, commented about my viral
• 9:14
  video. And to be honest, I don't
• 9:17
  completely disagree with her. Like, I would much rather put my
• 9:19
  time and energy into my modelling career and other
• 9:22
  ambitions I had rather than going around in a circle with
• 9:25
  you guys. OK, so now I've been threatened
• 9:28
  by lawyers, which is as much as I can say, but I will say I
• 9:33
  really, really don't appreciate that.
• 9:35
  So moving forward, if you could kindly so tell me who this,
• 9:40
  this, this, this, this and this are, then I will happily take
• 9:45
  down the videos that I posted on social media.
• 9:48
  I'm not trying to take down Huffer.
• 9:50
  I truly believe that's that's not good for anybody.
• 9:53
  Just like to reiterate that my intentions were and still are
• 9:58
  that I was just sharing my video to voice my concern of the use
• 10:02
  of AI and the use of likeness in the modeling industry.
• 10:06
  But more bought broadly, the creative industry threatening me
• 10:09
  with lawyers and threatening me with emotional manipulation
• 10:13
  through calling me on a Saturday morning is not helping you.
• 10:17
  And remember, don't tell a lie about me and I won't tell the
• 10:22
  truth about you. Now, anyone that listens to
• 10:24
  Minitech scroll regularly knows our stance when it comes to AI
• 10:28
  models and usage in this kind of way.
• 10:30
  We have spoken about this time and time again.
• 10:32
  It obviously takes opportunity and jobs away from real people,
• 10:35
  but aside from that, it is deceptive to sell a physical
• 10:39
  item, being clothing in this case, potentially on a seemingly
• 10:43
  AI generated person. And on top of this, AI often
• 10:46
  leans into very harmful biases and beauty standards that we do
• 10:50
  not need to see reflected in models and photo shoots of this
• 10:54
  nature. Many of these points have been
• 10:55
  pointed out in the conversation around Hafa and Elijah, but
• 10:59
  there also seems to be a lot of takes around how this has been a
• 11:02
  problematic response from the brand.
• 11:05
  APR expert on Tiktok named Caroline actually made a Tiktok
• 11:09
  explaining why she believes that Hoffa was not really well
• 11:13
  equipped for this situation and how this can seemingly be a
• 11:18
  recurring problems for brands if they don't know how to address
• 11:21
  it properly. And all inside part of our
• 11:23
  TikTok now. A New Zealand sweetwear brand is
• 11:26
  currently trashing its reputation and has exposed a
• 11:28
  really big PR problem that most businesses have today or they'll
• 11:32
  have very, very soon. The brand is Huffer and this is
• 11:35
  an AI trust PR problem that they are treating honestly like a
• 11:42
  1980s PR corporate playbook. And I'm just been watching the
• 11:45
  last few days that executives dig that hole deeper and deeper.
• 11:49
  Part of the brand is pushing back.
• 11:51
  The more that they try to suppress the story, the guiltier
• 11:53
  they look, even if they are telling the truth.
• 11:56
  This, the response has been very, very poor.
• 12:00
  If you've followed me before you will have heard me talking about
• 12:03
  why AI is basically the next big PR problem brands big and small
• 12:08
  are going to face. This is not just a communication
• 12:11
  scrub, it is a fundamental failure to well understand a how
• 12:13
  social media works and how reputation problems can start to
• 12:17
  really catch fire. It's also a governance and a
• 12:19
  planning failure, which a lot of brands I think still have the
• 12:22
  same attitude for. Like Huffer didn't have a
• 12:24
  protocol for AI because no one really thinks to build 1.
• 12:28
  Like we're all still figuring AI out that there are huge risks
• 12:31
  that come with that opportunity for efficiency, leveraging, blah
• 12:35
  blah blah blah. And I know we literally say this
• 12:37
  all the time, but This is why brands need someone on the team
• 12:41
  to be in the weeds of Internet culture and tech to just simply
• 12:44
  understand how responses and business decisions will land
• 12:48
  among the general public. In saying that, I was actually
• 12:52
  curious about what Elijah was hoping for the outcome of this
• 12:55
  situation to be, and he told me that he hopes for that just to
• 12:58
  be more regulation and public discourse around AI usage and
• 13:02
  the protection of likeness when it comes to creative industries.
• 13:06
  He also said that he hopes that Hoffa will issue some sort of
• 13:09
  apology, whether that be public or private to him and other
• 13:12
  models that they seemingly used AI on.
• 13:15
  As he told me quote in regards to Hoffa, a public apology to
• 13:19
  the models that are involved with this and an apology to me
• 13:21
  for putting me through this emotionally.
• 13:24
  In regards to AI as a whole, I would love to open up discussion
• 13:27
  with the general public and create a public forum on what
• 13:30
  regulations, rules and potentially laws we can put in
• 13:32
  place. He also did mention Denmark's
• 13:35
  law from last year that granted citizens intellectual property
• 13:39
  rights over their unique biometric identities being like
• 13:42
  their faces and voices, which I have covered on the pod before.
• 13:46
  But I am really curious to hear what you guys think of this
• 13:49
  story, whether it be the usage of AI models, the brand's
• 13:52
  response, how a IJA has addressed it on TikTok.
• 13:57
  And he also did note that if you want to stay updated with this
• 14:00
  story, he will be sharing those updates on his Instagram.
• 14:03
  So I'll put the handle on the screen as well as link it in the
• 14:06
  description if you want to stay across it.
• 14:08
  Let's move on to the next story for today.
• 14:10
  And we need to talk about how Polymarket is being pretty
• 14:13
  deceptive with how it's compensating creators and
• 14:16
  partnerships. So we have spoken about
• 14:18
  Polymarket, I think one time on mini tech scroll before.
• 14:22
  And if you missed that episode, you definitely have probably
• 14:25
  heard about Polymarket online. But in case you haven't, I will
• 14:28
  give you the rundown. So Polymarket is a prediction
• 14:32
  market, so users can essentially make bets on future outcomes,
• 14:35
  which are often tied to real life events.
• 14:38
  And these bets are made with cryptocurrency.
• 14:41
  So for context, some of the topics that users can bet on are
• 14:45
  whether the ceasefire in the Middle East will hold or various
• 14:50
  US domestic political issues. And Polymarket exists on the
• 14:54
  idea, or at least this is how it positions itself, that the quote
• 14:57
  UN quote wisdom of the crowd is more powerful than traditional
• 15:01
  sources of information. In a piece for days, Tom Waite
• 15:04
  explains how Polymarket works. So I thought I would just read a
• 15:08
  quote out for you. Quote.
• 15:11
  Basically users vote on yes or no questions, betting against
• 15:14
  each other by buying so-called contracts.
• 15:17
  Someone who believes that the existence of aliens will be
• 15:20
  confirmed by 2027, for example, would buy contracts representing
• 15:23
  the yes option. Depending on the likelihood of
• 15:26
  the event taking place, the cost of a contract fluctuate.
• 15:30
  Unlike typical bookmakers where you bet against the house and
• 15:33
  they set the odds accordingly, contracts work more like
• 15:36
  financial shares, with opposing users matched up to essentially
• 15:40
  bet against each other. The platform makes money via
• 15:43
  trading fees, so clearly you're not betting against the house as
• 15:47
  you would in a typical gambling setting.
• 15:49
  You are taking the position against others and I think that
• 15:52
  is significant because platforms like Polymarket make money by
• 15:56
  the amount of active users trading on the platform, no
• 15:59
  matter the outcome of these events.
• 16:01
  So prediction markets like this are controversial to say the
• 16:06
  least. There are growing concerns about
• 16:08
  insider trading within prediction markets.
• 16:10
  In fact, Kulshi, which is a Polymark competitor, said today
• 16:15
  that they'll start to collect employer information from users
• 16:19
  who are attempting to place bets so they can try and limit
• 16:22
  insider trading. But aside from that, there is a
• 16:25
  valuable question to ask about the ethics on betting on real
• 16:29
  life events, especially when many people are experiencing the
• 16:32
  devastating effects of certain outcomes in real time.
• 16:36
  It essentially is gamifying harmful phenomena and making it
• 16:40
  about personal financial gain, as opposed to focusing on what
• 16:45
  the realities of these events may even look like for a person.
• 16:49
  Not to mention how a platform like this can ultimately
• 16:51
  encourage and help sustain gambling addictions.
• 16:54
  Despite these concerns, Polymarket has only gained
• 16:57
  popularity over the course of this year after being relaunched
• 17:00
  at the end of 2025. But its rise to popularity
• 17:04
  really began with the 2024 election, with reportedly 3.2
• 17:08
  billion wagered on the outcome. There have been calls around the
• 17:11
  world to ban or limit access to Polymarket in the Australian
• 17:16
  context. The Australians Communication
• 17:18
  and Media Authority determined that Polymarket had offered,
• 17:22
  Quote, a prohibited and unlicensed regulated interactive
• 17:26
  gambling service to Australian users in breach of the law.
• 17:29
  In the US, the executive branch of government doesn't really
• 17:32
  seem to be as fast about Polymarket.
• 17:35
  Last year the Trump administration dropped two
• 17:38
  investigations into Polymarket and aside from this service
• 17:41
  feeding into more right wing circles, which is obviously
• 17:44
  Trump's political base, it should be noted that Donald
• 17:47
  Trump junior is an investor in Polymarket.
• 17:50
  So that might help explain why a less decisive stance has been
• 17:55
  taken. And Speaking of Donald Trump's
• 17:57
  political base, according to reports from The Wall Street
• 18:00
  Journal, these prediction markets have become particularly
• 18:03
  popular among young men, where there are many male creators
• 18:07
  promoting the platform to expand its user base.
• 18:10
  It's also worth noting that Polymarket has actually also
• 18:13
  collaborated with even more quote UN quote legitimate
• 18:16
  companies and online platforms. The last time we spoke about
• 18:20
  Polymarket on the POD was actually when the prediction
• 18:22
  market partnered with Sub Stack. And this kind of partnership
• 18:27
  definitely lends itself to a Sun level of legitimacy that is not
• 18:30
  the same as creator marketing. Mia Sato for the VAG actually
• 18:34
  covered creator collaborations with Polymarket a couple of
• 18:38
  months ago. And I thought how she put it was
• 18:41
  really perfect as she wrote quote, If TikTok influencers
• 18:44
  sell you a beautiful, aspirational lifestyle, the
• 18:46
  prediction market influences are selling you something more
• 18:49
  ruthless. The dream of monetizing life
• 18:51
  itself, of profiting when something happens to another
• 18:54
  person. This fervent energy behind
• 18:56
  prediction markets is fueled by its own content ecosystem, which
• 19:00
  is often backed by Polymarket and Khalshi's marketing teams.
• 19:04
  But that brings us to the main bit of the story.
• 19:06
  Recent reporting by Politico alleges that the company's chief
• 19:10
  marketing officer, Matthew Modaba used a personal PayPal
• 19:14
  account to send money to content creators after they've like,
• 19:17
  posted about the platform, which is obviously not transparent or
• 19:20
  appropriate. It's just super shady.
• 19:22
  So in the article, the journalist found that Modaba
• 19:25
  used a personal PayPal to send at least $350,000 to various
• 19:31
  creators, including far right influences.
• 19:34
  And this was between the beginning of 2024 and February
• 19:38
  2026 as it reads quote. But that Psalm is almost
• 19:41
  certainly an undercount. Modaba used his personal PayPal
• 19:44
  account, which is registered to an e-mail for a salad spot he Co
• 19:47
  founded to send over $2.5 million to more than 800 people
• 19:51
  during the 14 month span. And across this period, creators
• 19:55
  had posted about the platform just under 500 times on X and
• 19:59
  did not clearly disclose whether it was a paid partnership or
• 20:02
  not. I should note here that X only
• 20:04
  relatively. Recently added the ability for
• 20:07
  users to label post as paid partnerships in early March I
• 20:11
  think it was, but there are obviously ways to make that
• 20:14
  clear without having this kind of label, obviously.
• 20:16
  Politico did reach out to Polymarket for comment and this
• 20:19
  is what the spokesperson had to say.
• 20:21
  Quote. We routinely collaborate with a
• 20:23
  diverse range of independent organizations, partners and
• 20:26
  content creators spanning the political spectrum and
• 20:28
  constantly monitor, evaluate our progress and make the necessary
• 20:32
  adjustments in order to achieve our core mission of providing
• 20:35
  the most accurate, transparent and data-driven market insights
• 20:38
  to a global audience. If you want more details about
• 20:41
  the influences involved, definitely read the full
• 20:44
  political report because it is super interesting.
• 20:48
  But in saying that, I thought it was really important to chat
• 20:50
  about Polymarket once again since it has continued to make
• 20:53
  the rounds online since we first discussed it on the pod.
• 20:56
  And I thought this story was important to cover from a
• 20:58
  creator economy angle, obviously, considering we are
• 21:02
  seeing paid partnerships being funneled through personal
• 21:04
  accounts and also the failure to disclose ads.
• 21:07
  And while these aren't necessarily like new phenomena,
• 21:10
  this has happened for years. I think it's especially harmful
• 21:13
  when you think about what these creators are promoting and how
• 21:16
  many people, especially young boys, could fall victim to this
• 21:19
  kind of marketing and just see it as a way to make money.
• 21:23
  But I also think it's important because it's kind of an example
• 21:26
  of how we're seeing tech shape the news and the quote UN quote
• 21:30
  truth, which is what Polymarket positions its platform as.
• 21:35
  And this is obviously in a very tumultuous era where we're
• 21:38
  seeing the traditional political binary being harder to qualify
• 21:43
  and what is real and fake becoming just completely blood,
• 21:47
  as we already touched on in today's episode.
• 21:49
  But yes, I would definitely be interested to hear your guys's
• 21:52
  thoughts on Polly Market and this story as a whole.
• 21:55
  OK. And our last topic for today are
• 21:57
  some AI updates from Big Tech. This is like a two in one story,
• 22:01
  but I just wanted to mention them all quickly because I do
• 22:04
  think these are some of the biggest tech stories that have
• 22:06
  broken in the last week, but you've also maybe already heard
• 22:10
  of them, but I just thought I would mention it.
• 22:11
  So the first one is that Apple is really overhauling Siri.
• 22:15
  Apple held its annual worldwide developers Conference this week,
• 22:19
  and probably one of the biggest, if not the biggest announcement
• 22:22
  was the introduction of Siri AI, which is being described as
• 22:25
  Apple's attempt to kind of rebuild and reintroduce its
• 22:29
  Voice Assistant. That has honestly been not the
• 22:32
  best for quite some time. Basically, Siri's capabilities
• 22:35
  will be expanding. Siri will be more conversational
• 22:38
  and engaging, and will no longer need to hand off queries to like
• 22:43
  third party AI chat bots like Chat, GPC and Clod.
• 22:47
  This is how it's explained on the Apple Newsroom blog quote.
• 22:50
  This new version of Siri is built on Apple intelligence,
• 22:53
  allowing Siri to draw on personal context understanding
• 22:57
  and help users find what they need in the moment across
• 23:00
  messages, emails, photos and more.
• 23:02
  For example, users can ask Siri to find a restaurant
• 23:05
  recommendation a friend message them about.
• 23:07
  So if there's a hotel confirmation number from an old
• 23:09
  e-mail, I'll pull up photos with friends and family from a recent
• 23:13
  trip. And I want to add here quickly
• 23:15
  that Apple uses Google's Gemini AI to kind of underpin much of
• 23:19
  the foundation of Apple intelligence.
• 23:21
  So just something to think about that I don't really think it's
• 23:24
  necessary to get into all the new features with Siri AI.
• 23:28
  I've said before, I don't think it's the most conducive to like
• 23:30
  an independently hosted podcast, but probably the biggest change
• 23:35
  is how Siri will handle conversation.
• 23:38
  The old version of Siri would essentially just take one
• 23:40
  command and then kind of forget about it immediately.
• 23:42
  But Siri AI will be able to keep the thread going and we'll
• 23:46
  remember what you asked and follow up and as well as we'll
• 23:49
  be able to kind of read across all your apps to give you a more
• 23:52
  comprehensive answer. Of course, there are also some
• 23:55
  limitations to the launch that I've been seeing people talk
• 23:57
  about. Some of them is that the Siri AI
• 24:00
  will only be available in England at first and that it
• 24:03
  also will not be available in the EU, and then it also won't
• 24:07
  be available on certain Apple devices.
• 24:09
  Other than that, the discourse I've been seeing around Siri AI
• 24:12
  has actually been really interesting because there has
• 24:14
  been a lot of talk over the years and we've spoken about it
• 24:16
  on the podcast before, about how Apple is kind of falling behind
• 24:19
  in the AI race. Instead of focusing on like the
• 24:22
  model development, Apple seems to be focusing on delivering
• 24:26
  these functions to their users. But now it kind of seems to be
• 24:30
  playing catch up by introducing a whole bunch of new features
• 24:33
  related to AI, which kind of makes sense like from a business
• 24:36
  angle, like that is what is getting the industry like
• 24:39
  excited and funded. But from a consumer first angle,
• 24:43
  obviously can optimize a user's experience.
• 24:45
  But so many consumers are just like so anti AI where it's just
• 24:50
  so polarizing, if not negative. So it might also turn people
• 24:53
  off. But Apple does just really have
• 24:56
  chokehold a lot of us. I say, reading off my MacBook
• 24:59
  Pro right now. My last little tidbit I wanted
• 25:02
  to mention quickly is that Open AI has filed for an IPO.
• 25:05
  We have been talking about this for a while, so I thought I
• 25:07
  would just give the update on Monday, Open AI shared that it
• 25:10
  had filed preliminary paperwork with the US Securities and
• 25:13
  Exchange Commission to become a publicly traded company.
• 25:17
  And then they did announce this through a blog post.
• 25:20
  But obviously, this is very interesting coming up the back
• 25:22
  of the like the Elon Musk lawsuit, which was all about how
• 25:25
  Open AI was moving away from its original charitable mission,
• 25:29
  instead becoming like a fully fledged for profit business.
• 25:32
  And a lot of people said at the time that Elon Musk was going
• 25:35
  ahead with this lawsuit fastly as like kind of a revenge thing,
• 25:39
  but secondly, to kind of build up a negative attitude towards
• 25:44
  ChatGPT and Open AI ahead of the IPO.
• 25:47
  And also this week, it's been reported that Open AI is
• 25:50
  continuing to work on its Quote UN quote super app with coding
• 25:53
  tools and AI agents ahead of the IPO.
• 25:56
  As the Financial Times put it, quote.
• 25:59
  These changes, which will give greater prominence and resources
• 26:02
  to Open AIS coding product codecs, reflect a growing
• 26:05
  conviction within the company that the future of AI lies not
• 26:08
  in chat bots that answer questions, but in agents that
• 26:11
  perform task for users. As one Senior Open AI employee
• 26:15
  told the Financial Times, quote chat is dead.
• 26:18
  So I guess this kind of strategy shift shows how Open AI is
• 26:22
  becoming increasingly aligned with Anthropic, probably its
• 26:26
  major competitor. Because after Anthropic started
• 26:29
  focusing on designing products for businesses, it did have a
• 26:33
  significant uptick and growth, which is also part of its IPO,
• 26:37
  which it recently registered for as well.
• 26:39
  Which is also very interesting because like seeing all these
• 26:42
  companies move towards being publicly listed, it definitely
• 26:45
  indicates A willingness to embrace the current investor
• 26:50
  enthusiasm around AI. But like who knows how long that
• 26:53
  will go on for and whether the bubble will burst.
• 26:56
  But I guess that's something we'll only know down the line.
• 26:58
  Well, that is everything for Mini Tech Scroll for today.
• 27:02
  I really hope you guys enjoyed that did feel like a long
• 27:04
  episode. I don't know if it actually will
• 27:06
  turn out to be a long episode, but recording it, it felt like a
• 27:10
  long time. So I really hope it's good,
• 27:13
  otherwise Lauren will speak to you guys tomorrow.
• 27:15
  Thanks for listening. Bye.
• 27:17
  Centennial World Podcast Network operates on Gadigal land.
• 27:20
  We would like to acknowledge the Gadigal of the Eora Nation, the
• 27:23
  traditional custodians of this land, and pay our respects to
• 27:26
  elders both past and present.