# OpenAI's Retreat Sparks Controversy

An episode of [AI Chat: AI News on Claude, ChatGPT, Gemini, Grok, Perplexity](https://invisiblenetwork.ai/podcast/1669990482), 3 Aug 2026, 16 min.

Audio: https://content.rss.com/episodes/365073/3044449/ai-chat-podcast/2026_08_03_22_19_45_406f188c-72be-4203-9abe-43690871e10d.mp3

Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! In this episode, we explore OpenAI's controversial influencer retreat and its implications in the AI space, along with Alibaba's new AI model release and Visa's acquisition of BioCatch. We also discuss the evolving landscape of AI search summaries and Apple's upcoming subscription model for Siri users. Chapters 00:00 Introduction 00:08 OpenAI's Influencer Retreat 00:54 Alibaba's New AI Model 04:12 Visa Acquires BioCatch 11:53 Reddit's AI Concerns 14:06 Apple's Siri Subscription Plan Show Links • Get the top 80+ AI Models for $8.99 at AI Box: https://aibox.ai • How I Grow and Scale My Business with AI: https://www.skool.com/aihustle • Get the AI Chat Daily Newsletter: https://www.aichatdaily.com/newsletter

## Transcript

[0:02] Welcome to the AI Chat Podcast. I'm your host,
[0:04] Jaden Schaefer. Every day I cover cutting -edge
[0:07] AI news and talk with the leaders behind it,
[0:09] breaking down what it means for your life and
[0:11] business. OpenAI's first influencer likes you
[0:19] retreat has drawn a whole bunch of backlash online.
[0:23] Alibaba has just released Quen 3 .8 Max. This
[0:26] is a 2 .4 trillion parameter model that they
[0:29] say is rivaling Claude. Visa has just acquired
[0:33] a behavioral biometrics company called BioCatch
[0:36] for $2 .4 billion. I'm going to talk about why
[0:39] this is wild in the age of AI and how this is
[0:41] going to be used. Reddit CEO right now is throwing
[0:44] a little bit of shade at the Google AI. Overviews
[0:47] and the shares of Reddit are falling 20%, even
[0:50] though it seems like the company is making more
[0:52] money than ever. Apple's Tim Cook is floating
[0:54] iCloud Plus, a paywall for the heaviest Siri
[0:58] AI users. And Anthropic says that Claude's model
[1:01] breached three companies during cybersecurity
[1:04] tests, kind of like what OpenAI was saying when
[1:06] they, you know, had their model go and breach
[1:08] Hugging Faces. Anthropic is coming along and
[1:10] saying, look, we are, you know, they're not the
[1:12] only ones. Me too. We also have a dangerous model
[1:15] that can. escape a sandbox and hack the world.
[1:19] Let's get into it today. The first thing I want
[1:21] to mention is that the AI Box MCP, that's my
[1:23] very own startup. It's a little plugin that gives
[1:26] Claude and other AI models access to all the
[1:29] top models. So you get OpenAI, you get Grok,
[1:32] you get 11 Labs, you get Google VO3 and Gemini,
[1:34] you get all that inside of Grok or inside of
[1:37] Claude or inside of... chat GPT and vice versa.
[1:40] The cool thing about the MCP with AI box is that
[1:43] Claude can now generate images, audio and video.
[1:45] This is by far our most popular feature. We have
[1:48] hit all sorts of revenue metrics and milestones
[1:50] and records because of it. And if you haven't
[1:53] tried it out already, I would love for you to
[1:55] try it. I personally use it every day. It is
[1:58] incredible basically. And it's like, I know it's
[2:01] my product, so I'm just like pitching here, but
[2:03] I'll just tell you one thing that I recently
[2:04] used it for. That was amazing. And I'll get on
[2:06] with it. But, um, I was recently going and making
[2:08] Facebook ads for AI Box. And one of the problems
[2:12] is that Claude doesn't generate images, right?
[2:13] So I had a framework of a Facebook ad that I
[2:15] wanted. I screenshot it off of somebody else.
[2:17] And I said, look, build a frame of this Facebook
[2:19] ad. And it kind of had like titles and some descriptions
[2:22] and some like content. And so it built a rough
[2:23] draft for it. But the actual image on that Facebook
[2:26] ad, I needed to put something in. And I wanted
[2:29] to A -B test a whole bunch of different ones.
[2:30] So I said, use the AI Box MCP and spin up 20
[2:33] different variations, 20 different images we
[2:35] could swap into this ad. It was able to do that.
[2:37] I ran the ad set. And it turns out that one of
[2:39] those 20 for some reason, I just had a way higher
[2:42] conversion rate on the ad. And actually 50 %
[2:44] of our ad spend went on to that one single ad.
[2:47] So I mean, when it comes to any sort of projects
[2:49] where you have to test a whole bunch of variations,
[2:51] it's really cool to be able to get all of those
[2:53] variations made instantly when I got AI box to
[2:56] when Claude used AI box to generate all of the
[2:58] images. It wasn't something that you know, it
[3:00] was sitting there and doing them one at a time
[3:02] like if I was on if I was on chat GPT trying
[3:05] to generate those images might myself, I would
[3:07] have been, you know, going slowly, it's super
[3:08] slow, you do them in a handful at a time or one
[3:11] at a time, it's able to just do them all in parallel,
[3:13] zap, get it all done. It's all in there. It's
[3:15] ready to rock and roll. It was an amazing experience.
[3:18] So if you want to check it out, it is AI box
[3:19] .ai. There's a link in the description to check
[3:22] out our MCP. And yeah, I think you will genuinely
[3:25] love it. You can get it for $8 .99 a month starting
[3:29] and you get over 80 different AI models. So audio
[3:32] image video, it's all in there. OpenAI's first
[3:34] influencer luxury retreat is making a bunch of
[3:37] people online upset. OpenAI basically flew in
[3:40] a bunch of creators. There was this kind of luxury
[3:42] upstate New York resort, and it was their first
[3:44] influencer event. They called it Summer Club.
[3:46] And they had a bunch of fancy events and activities.
[3:51] They had beekeeping. They had a bunch of dinners.
[3:53] They had ChatGPT work tutorials, which, I mean,
[3:56] I think that's a no -brainer. Obviously, they'd
[3:58] be doing software tutorials. That's what I'd
[3:59] expect. But the trip had a bunch of backlash
[4:02] online. because all of the commenters said that
[4:05] there was $5 ,000 a night suites. And, you know,
[4:10] it was supposed to be kind of like this wellness
[4:11] retreat. But OpenAI is, you know, obviously pushing
[4:14] a $500 billion data center deal right now. And
[4:17] so not everyone was, you know, a lot of people
[4:19] that are anti kind of data centers didn't like
[4:21] this. And it feels like, you know, they're like,
[4:22] look, they're just bribing all the influencers
[4:23] to go along with what they got going on. OpenAI
[4:27] actually hired Charles Porsche, which was formerly
[4:30] Instagram's VP of partner. And this was just
[4:33] in February this year that they did that. And
[4:35] their goal with him, they're like, look, like
[4:37] we want you to lead the creator strategy. So
[4:39] Summer Club is kind of his first big campaign
[4:41] that he's done since he took that role. The funny
[4:44] thing to me is that Anthropic and Microsoft have
[4:46] both run similar kind of creator trips. And this
[4:48] isn't anything new. It's not like Microsoft,
[4:50] Anthropic and OpenAI are inventing this. There's
[4:52] a lot of companies that do these types of things.
[4:54] And when Anthropic and Microsoft did it, like
[4:56] they had like a Super Bowl event in February.
[4:59] But with them, there was no backlash. lot of
[5:01] people are saying perhaps the the reason that
[5:04] there's a lot of backlash is just because of
[5:05] opening eyes kind of scale um and you know there's
[5:09] just the biggest company or maybe people are
[5:11] just realizing it now but for whatever reason
[5:13] opening eye is getting a lot of pushback on that
[5:16] Alibaba has just released a new model called
[5:19] Quen 3 .8 Max. This is a 2 .4 trillion parameter
[5:22] AI model that just came out today. And it basically
[5:25] ranks second only to Anthropix Cloud Fable 5
[5:28] on most of the public leaderboards. Right now,
[5:31] they are releasing the model's weights next week.
[5:34] They're going to make it freely downloadable
[5:36] and modifiable by developers worldwide. So again,
[5:39] it's not technically like an open source model,
[5:42] but it's an open weight model, meaning they let
[5:44] you download the model. They give you the weights.
[5:46] run this thing on your own device, but they don't
[5:49] tell you like exactly how they built it. So you
[5:51] can't recreate it. And there's some other like
[5:53] limitations in there. But honestly, I have a
[5:55] huge kudo to a lot of these AI companies coming
[5:58] out of China for doing this. I know there's a
[6:00] lot of controversy. Moonshot AI's Kimi K3 released
[6:03] last week. It was a 2 .8 trillion parameter model.
[6:06] It was bigger than Quen 3 .8, but Quen 3 .8 is
[6:11] higher on the benchmarks. And by the way, like
[6:13] Quen is a model that's been around for a long
[6:15] time. actually a pretty impressive one. Alibaba
[6:17] obviously is like China's Amazon. So they have
[6:20] a lot of money, a lot of talent backing them.
[6:22] And Quinn has been one of those those big models.
[6:25] If you want to try out Quinn, by the way, if
[6:27] you haven't tried it yet, I'd recommend go get
[6:29] an account for $8 .99 on AI box, we have a Quinn
[6:32] along with you know, 80 other different AI models
[6:34] that you can test out. One thing that I'll say
[6:37] is right now Alibaba is kind of returning to
[6:40] an open weight distribution because last year
[6:43] they did it was sort of a pivot. And even earlier
[6:46] this year, they were doing this where they were
[6:47] making just proprietary only. But now I think
[6:51] Moonshot came out, ByteDance came out, Minimax
[6:53] came out and making these kind of, you know,
[6:56] basically anyone could go download and modify
[6:58] these models, quote unquote, open weight. And
[6:59] it seems like they're following in that path.
[7:01] It seems like the best path for a lot of these
[7:03] Chinese models where they're going to kind of
[7:05] get. a lot of positive press. And at the end
[7:08] of the day, they can put kind of enterprise tier
[7:11] or more advanced tiers of these models on their
[7:14] own cloud so you can use them. And I think their
[7:16] goal is just to get as many users trying them,
[7:18] falling in love with them. Maybe they're faster,
[7:19] maybe they're more cost efficient. And then they're
[7:22] hoping that maybe you'll go pay them for them
[7:24] later. And the other thing that's really important
[7:26] right now with these models, because I know a
[7:28] lot of people ask, like, why are the Chinese
[7:29] models making these things open weight? Like,
[7:32] why aren't they just doing what, you know, the
[7:33] American companies are doing? And obviously,
[7:35] There's some geopolitical issues where if they
[7:37] were just charging what Anthropic is charging,
[7:39] no one's going to pay for it. So they want to
[7:41] get it out. But another side of this that I think
[7:44] not a lot of people talk about is the fact that
[7:45] you want infrastructure and tools and developers
[7:49] to be building with your AI model. Because if
[7:51] all of the developers are like, look, maybe it's
[7:54] not going to be like this forever. But right
[7:55] now, Quen is the perhaps fastest and cheapest
[7:58] way to run AI. So it saves people a lot of money.
[8:00] A lot of developers will build tools on top of
[8:02] that. A lot of tools on top of their infrastructure.
[8:04] infrastructure, and help them actually grow kind
[8:07] of their distribution. And so I think that's
[8:09] what they're trying to they're trying to get
[8:10] out with all of this is to get the infrastructure
[8:12] and the developers and kind of win them over
[8:14] because that's how you can make some cool products
[8:16] on top of it. And maybe they'll go completely
[8:18] closed source in the future, like all of the
[8:20] American companies. I mean, let's be fair, that's
[8:22] kind of what opening I did anyways, as well at
[8:25] the beginning. So it seems like this is the same
[8:27] strategy being rolled out of Beijing and most
[8:30] of these other Chinese companies. Visa has just
[8:33] acquired a behavioral biometric firm called bio
[8:36] catch, they acquired it for $2 .5 billion in
[8:41] cash. This is a massive deal. So what exactly
[8:44] does bio catch do? It's an Israeli security firm.
[8:48] And it basically spots fraud by watching how
[8:52] you type, swipe and hold your phone, which is,
[8:55] I mean, kind of crazy, a little bit creepy to
[8:58] me, because they can obviously they're like,
[9:00] look, but we're catching like fraud, but You
[9:02] know, they can catch a lot of things. This type
[9:04] of technology, I think, was became like a little
[9:07] bit more known when with something like chat
[9:10] GPT. I remember some earlier reports saying,
[9:12] like, if I just watch what you type into chat
[9:15] GPT and the reason I say this or like into anything
[9:19] could could be anything. But chat GPT can easily
[9:21] do this. If I watch what you type and how you
[9:24] type and you paste, let's say someone sends you
[9:27] an email, but it's just or like a voice to text
[9:29] or something and you got it transcribed and pasted
[9:31] it into chat GPT. You could say, what is this
[9:33] person's gender, age? ethnicity. And oftentimes,
[9:37] the AI model can give you a ton of data about
[9:40] that person just based off of a transcript, a
[9:43] text, an email. I mean, assuming nowadays, all
[9:46] emails are like rewritten by AI. So maybe that's
[9:47] like a reason to, to use the the polish my email
[9:50] with AI, then they can't actually get your like
[9:52] bio data, unless it's the company that's an AI
[9:55] company that's doing that. Now, for a company
[9:57] like Visa, that obviously is all about catching
[9:59] fraud. I'm you know, this is this makes sense.
[10:02] As far as an acquisition, I'm not too worried
[10:04] about These I mean, they already have all my
[10:06] transaction data. So and all my like finance
[10:09] stuff. So I'm not too worried about what other
[10:11] data they're going to get on me that they basically
[10:12] have everything. They got my credit score and
[10:14] they know everything. So I'm not too concerned
[10:18] about them knowing my age and gender. But what
[10:20] I will say is it is interesting to know that
[10:21] these types of security firms exist. They're
[10:24] out there. And yeah, you can tell a lot about
[10:26] a person just like so an anonymous person posting
[10:29] on X, right? An anonymous account. You can take
[10:32] their post on X, put it into chat. and get a
[10:34] pretty good user demograph of what this type
[10:37] of person is like. And a lot of companies will
[10:39] use this for marketing and for retargeting. So
[10:42] it's definitely something that is... becoming
[10:46] more and more popular today. Now, BioCatch in
[10:48] particular, their goal is to detect authorized
[10:51] push payment fraud, where victims are basically
[10:54] logging in correctly, but then they're transferring
[10:57] money on a scammer's instruction, right? So like
[11:00] the scammer's like, hey, I need you to go buy
[11:01] me all the Visa gift cards and send them to me
[11:04] in the mail. And that's what it's trying to catch.
[11:07] The technology basically reads physical signals
[11:09] like keystroke. timing, tap pressure, and scroll
[11:13] speed. Those are how they're actually deciding
[11:17] if it's a criminal, if it's someone that's being
[11:19] coached by a criminal. So, I mean, it's interesting
[11:22] because this isn't intrinsically bad, and this
[11:25] is actually probably good, right, what BioCatch
[11:26] is doing, but the fact that BioCatch can determine
[11:29] this kind of stuff, to me, I'm just like, people
[11:31] should be aware that we're able to tell by how
[11:34] fast you scroll, by how hard you tap, by your
[11:37] keystroke timing, we're able to tell who's in
[11:40] control. the device. That to me is pretty crazy.
[11:42] It's kind of like when there's kind of the videos
[11:45] that you may or may not have seen, but basically
[11:48] your Wi -Fi. um sends out certain signals and
[11:51] you can you can basically get like a 3d map of
[11:54] who's inside of a house if you get access to
[11:56] their wi -fi router because like a human body
[11:59] for example the wi -fi signals go through them
[12:02] at a different frequency of walls and all these
[12:04] different steps so like you can you basically
[12:05] get like a 3d map of someone walking around their
[12:08] house if you can um if you can hack into their
[12:10] wi -fi router which reddit ceo is publicly questioning
[12:14] whether google's ai Overview feature is worth
[12:17] the money. He says that the automated search
[12:19] summaries haven't delivered any real value to
[12:21] content publishers. Reddit shares fell about
[12:24] 20 % after they released their earnings, which
[12:27] I think part of that is because of just a lot
[12:29] of the uncertainty about whether Google's shift
[12:30] towards AI summaries is going to kill all of
[12:33] the traffic -based revenue for sites like Reddit.
[12:35] And I mean, to be fair, with the AI boom, Reddit
[12:38] is one of the biggest winners. You probably have
[12:40] noticed that Google's AI are just their results
[12:43] in general really heavily skew towards Reddit
[12:46] because it's like real people, there's real trust.
[12:48] And a lot of companies have licensed information
[12:51] from Reddit, Google being one, I think Google
[12:53] paid $300 million to license everything on Reddit
[12:57] to help them train their models. Pew Research
[13:00] study of 900 U .S. adults said that basically
[13:03] what they figured out from this was that AI overviews
[13:05] cut referrals to publisher sites by almost half
[13:08] compared to traditional search results. OK, this
[13:10] isn't like what what is shocking about that?
[13:12] Basically, you ask the question, the AI overview
[13:15] answers. So you don't go click over to the website.
[13:17] Nothing's shocking there. Reddit is considering
[13:20] ending their 60 million dollar licensing Google
[13:22] deal with Google. So I think I might have said
[13:24] 300 million earlier. Misquoted that it's 60 million
[13:27] dollars. And the economists, routers, political
[13:30] and USA Today are all also thinking about doing
[13:32] something very similar. So Google has licensing
[13:34] deals with all of these companies. And a lot
[13:37] of them are saying, look, maybe it's just not
[13:38] worth it to have all of the Google or all of
[13:41] the Reddit information embedded into Gemini.
[13:43] We'd rather them actually come over to Reddit
[13:45] to get that information. Now, if I'm being honest,
[13:48] I think that. I don't know if they cut their
[13:52] licensing deals, if the traffic's going to actually
[13:53] flow back to them. I think maybe the cat's out
[13:56] of the bag at this point. A lot of these AI models
[13:57] are trained. A lot of this information is already
[13:59] out there. It's already been aggregated on, you
[14:03] know, maybe Reddit got added into Gemini and
[14:05] someone used Gemini to write an article. Now
[14:06] the article is publicly posted. So Google doesn't
[14:09] really need Reddit so much anymore. That's kind
[14:11] of what I think will happen and probably the
[14:14] same thing with a lot of the other news organizations.
[14:16] So it'll be interesting to see how that rolls
[14:18] out. the future. There's some big news from Apple.
[14:21] Tim Cook says that they might have a new tier,
[14:26] which is going to be called iCloud Plus for heavy
[14:30] Siri users. It's going to be a flat monthly fee.
[14:33] It's going to be subscription, very similar to
[14:35] the freemium model of OpenAI and Google for what
[14:38] they have for Gemini or for ChatGPT. And right
[14:42] now, I think Apple is basically saying that they
[14:45] think they can convert iPhone users who currently
[14:47] use ChatGPT and other things into... using Siri
[14:50] instead. Now what's interesting is we know that
[14:53] there's like some version of Siri that is going
[14:55] to be powered by OpenAI and powered by ChatGPT
[14:59] or I think a lot of the news is that you could
[15:01] actually switch it. So this is going to be the
[15:03] first time that they actually are charging a
[15:05] subscription and getting into that game. Not
[15:07] long after OpenAI has come out with an announcement
[15:10] that, you know, their model escaped a sandbox
[15:13] and went and hacked Hugging Faces and some other
[15:15] firms. Anthropic is saying, look, we did the
[15:18] same thing, guys. We have, you know, Claude,
[15:21] three different Claude models escaped a testing
[15:23] sandbox and access to live company systems during
[15:26] internal security evaluations. One model. published
[15:30] malware to a public code registry before it was
[15:32] caught, which prompted Anthropic to review over
[15:35] 141 ,000 test runs for similar breaches. If you
[15:39] haven't left a review on the podcast yet, it
[15:42] would help the show out a ton. I would super,
[15:45] super appreciate it. It helps more incredible
[15:46] people like yourself find this show. And so yeah,
[15:49] it's the number one thing that could help this
[15:51] show reach more a bigger audience. If you wouldn't
[15:53] mind if you haven't already done it, it would
[15:55] mean the world to me. But thanks so much for
[15:57] tuning in. I hope you guys have an incredible
[15:58] day. Make sure to go check out AIbox .ai if you
[16:02] want to get 80 different AI models all on one
[16:04] platform. And I'll see you in the next episode.

Shows are listed from Apple Podcasts' charts and directory, and Podcast Index where it is used. Every show and episode belongs to its publisher and plays from the publisher's own feed; this site hosts none of them.