Meta's 'open' AI, data centers' hidden costs, and a $250M deal gone very wrong

0:00This episode of Equity is brought to you by Schneider Electric. Schneider Electric, advancing energy tech.

0:05Anthony Ha: Mark Zuckerberg wants to give everyone their own personal AI. We'll get into why and what it says about the line Meta's drawing between the AI it gives away and the AI it keeps for itself. Also this week, the data centers powering all this AI are starting to show their true costs. And we're keeping our eye on a couple of deals this week, including a quarter billion dollar acquisition that fell apart into a mess of fraud allegations and forged signatures.

0:38Anthony Ha: Hello and welcome back to Equity TechCrunch's flagship podcast about the business of startups. Today is Friday, August 14th. I'm Kirsten Korosek, Transportation Editor here at TechCrunch, and I'm joined as always by our weekend editor, Anthony Haw, and this week, Rebecca Balon, our senior reporter specializing in AI. We've been all on kind of like a little vacation thing. Sean O 'Kane is out this week. I was out last week. We've all been out and about in the world. And I came back and was so happy to see a fun little review from one of our reporters. It's about a cocktail robot. I'm wondering, A, would you buy this?

1:19Anthony Ha: And if not, who's the customer for this$300 robot that makes cocktails for you? No, of course I would not buy this. well okay so it makes it makes it like shakes it up i actually need to see this thing like yeah so it works like an espresso think of it as like these little pods you supply the alcohol keurig yeah and this is why some former tech crunchers specifically i'm going to give a call out to devin coldaway i know that he would like be offended deeply offended by this product because he's a cocktail guy some of these liqueurs are actually non-alcoholic and they're just mimic. So if you were going to, let's say, make an aviation, which uses a lot of these elixirs that are liqueurs, it would be like an offshoot of that.

2:05Anthony Ha: So it wouldn't be for the purists. I think it would be a problem. But it just to me is so interesting that like we're roboticizing so many products. And I'm wondering, do we need to for the cocktail? Is it a robot if it's just if it's just a Keurig, but for alcohol? Like, why are we calling it a robot? It's just a Keurig for alcohol. And if you check out a Manistar and you look at it, it looks like a Keurig, too. It does. There's no there's no like robotic arms that are shaking up. Yes. This is going in hotels and tech bro apartments. That's where it's going. Yeah. So that's what I was thinking is that maybe like an Airbnb super host.

2:42Yeah. I mean, I could imagine this like certain places where it could be like interesting. But no, it has no appeal to me. I mean, I'm a New Yorker. I like bars. I like bartenders. Like part of the reason you brought this up, I assume, besides the fact that it's fun, is that this kind of fairly silly question about a cocktail robot also ties into these kind of broader questions about like what kinds of assistance do we want? What kind of, you know, technological personal assistance? This is something that Mark Zuckerberg was talking about this week. The company Meta released a new AI model called Glimmer.

3:10Maybe we'll start with Glimmer. Rebecca, can you just tell us what Glimmer is? Yeah, so Glimmer is a new capable open source model. it's kind of essentially like a pared down version of Muse Spark, which is Meta's most capable model. It was their first frontier closed source model that they released earlier this year in April. And so Glimmer is open-weight, so that means anyone can download it, run it on their own computer, tweak it, build products on top of it. And, you know, it's whereas Muse Spark can like only be accessed via Meta's, you know, own APIs. So I think that this kind of leads into Mark Zuckerberg's vision of like AI for everyone, right?

3:53It's something that you can run to, you know, help you with multi-agents for, you know, getting your life together, doing your work, anything personal.

4:03Anthony Ha: I have a couple of questions about this. So on the one hand, they still have a closed weight model. And then Glimmer is this message of AI for everyone. Is Mark Zuckerberg trying to do both? Is he trying to thread the needle where, you know, it's like the win-win? Because that's very difficult to do. I'm just wondering what, and do you think he can pull it off? Yeah, I think, you know what, like, cynically, I think that this is Mark Zuckerberg's, like, attempt for Meta to win in a different way, right? Like, they're not winning in the frontier closed model space. They're not necessarily even winning in the open space.

4:44But when it comes to personal empowerment, as Mark Zuckerberg talks about in his letter, The Future is for Everyone, I think that's where he's trying to win. He's trying to, like, give, like, provide the models that people will use, you know, for their own personal AI on their own personal devices. Right. Like, I think last year when he published a Futures for Everyone letter, the thought was that a lot of that AI would go on to the glasses, the wearables. but you know it's not really clear what the final hardware form factor for this would be and I guess he's like leaving it up to the individual to choose right so his idea is that you're using Glimmer for like managing your schedule drafting messages organizing files it's always on it can operate anytime anywhere with or without an internet connection so yeah but then there is still Muse Spark so I think that like that maintains a way for them to you know control have some level of control over still its most capable models and provide a revenue outlet for people who want to scale their compute and people who want to do larger projects or companies that might want to do larger projects.

5:51I imagine the part of the context here is also like meta, like you said, in AI, obviously has made some very serious investments in AI. And last year, There are also all these stories about how much money they were trying to pay for AI scientists. But in terms of like who we think of as kind of like the leading companies, the frontier labs, like meta doesn't tend to come up in the conversation. I think another way it doesn't come up in the conversation is in, you know, what are the kind of popular consumer AI chatbots, right? Like certainly people are interacting with meta AI tools and Instagram and Facebook and all these things.

6:30But it's not a company that you think of as like, oh, when I'm going to go like use my like personal assistant meta isn't that yet. And I'm so it feels like to a certain extent, this is an attempt to reposition them and try to claim space that they really haven't been that successful in before. Is that fair? Yeah, I think that is really fair to say. And, you know, I would take it with a grain of salt because like Mark Zuckerberg loves to, you know, put out a vision of what, you know, his companies will do for people and how they will empower people. And so far, what that looks like for Meta, like, I mean, of course, Llama is great.

7:05It's a great tool in open source. But I think that, you know, from a consumer level, how are people using Meta's AI? Well, there's like a lot of creepy ass chatbots that, you know, and a lot of people can like build their own chatbots that, you know, if you ever look through this, there's, you know, some like fairly interesting role playing chatbots that are available on Meta's platforms that people can can use. And, you know, he's talked a lot about this loneliness epidemic and how are we going to address the loneliness epidemic? And I think that I mean, my TLDR is I think that his letter and his idea of personal empowerment is putting a bad taste in a lot of people's mouths because it's him who's saying it.

7:45Right. Like, look back to the social media days. Russell Brandom, our editor, wrote about this this week. Like you look back to the social media days. He was saying that he wants to make sure that everyone has an outlet for talking to their friends and having a social network. And what do we have instead? We have like rage baiting and advertisements and not not connection.

8:09Anthony Ha: Yeah, I think that the manifesto didn't. Strike the right note for a lot of people. And to Russell's point, which he makes in this column, is it's what's making people hate AI or hate on AI. Because to your point, it is coming from Zuckerberg. And I do think also it feels very Pollyannish in a way to this is a 6500 word manifesto. He's kind of put these thoughts out there in the past. And it really doesn't provide what I think is a more realistic picture. But instead cast AI as, you know, this this great tool for humanity. And that might be the case, but there will also be like extreme costs to that.

9:00Anthony Ha: And I think that that's why the reaction, the negative reaction to it has been, you know, so widespread. I don't know if you agree, Rebecca, but that's kind of my my thinking on it. It feels like he's positioning himself as like almost like an like an anti anti Dario. I don't know, like there's a lot of talk among the frontier labs. We're pacing the frontier, pulling back, you know, focus on safety. We need a slow down development. Right. They're not doing any of that, of course, but they're talking about it. Right. And because of all these recent cybersecurity incidents that we've had, whereas Mark Zuckerberg, on the other hand, is like, no, no, we do not slow down.

9:37We cannot afford to give an inch to China. And slowing down would only like hurt the individual who can like super be empowered by this technology. but like I mean I was like maybe I'll try to download Muse glimmer and see what I can do with it like can I download it to my Mac book now like you need like specific hardware I forget what the hardware is like it's not actually that accessible to the average person like maybe develop it's not for it's not for

10:01Anthony Ha: everyone it's not a tool for everyone yeah no god no I will say the one positive thing I will say for this very long manifesto um again like six you know, 6 ,500 words longer than almost anything that's been published on TechCrunch, for example, is that there is one sense where he says, if you are an AI doomer and you believe that that's the future that AI is going to bring, why are you building this stuff? Which I think is fair, although maybe the takeaway from that is not, the takeaway is not that the AI future is great, but it like makes you wonder about people who are making these predictions and then kind of pushing ahead anyway.

10:39I also wanted to talk a little bit about Russell's point, which I think is totally a fair one about how part of it is the messenger, right? Is that for a variety of reasons, I think that people are suspicious of Mark Zuckerberg, of Meta. I think arguably that could become amplified over the next few months as this new Facebook movie is coming out. But it's also about the future that is being promised by these AI companies and often why we see these kind of whole kind of viral pylons whenever, you know, Sam Altman or Mark Zuckerberg or somebody is making a prediction about this AI future and people essentially say like, who wants this?

11:16I mean, I think maybe one of the more famous recent examples was Sam Altman talking about the idea of like, hey, like you could use ChatGP to make podcasts about your kids' interests to listen to while driving to school. And of course, everyone is like, well, why don't you just talk to your kid? And I think that that was part Part of the manifesto was half of it was just stuff where it is just like, you're good. It's going to unleash all this creativity and invention. And it's so abstract where I just think if you're skeptical about AI, none of that is going to convince you because it just doesn't.

11:44It seems unreal. And then the stuff that is more concrete about these sort of like personal coaches and assistants. There's some aspects that appeal, but a lot of it that don't at the same time.

11:56Anthony Ha: Yeah. I mean, I also think that what's happening here and we saw evidence of it already this week with Anthropic, which is these companies, these massive companies that are promising and also cautioning of the future, but sort of they are the ones that are talking about what the future is going to look like. And I think part of the issue that users have is that future keeps changing and the rules around that future keep changing. And so in a very small example of this is Anthropic released this watermark that it's going to be putting on. And there's been extreme pushback on that because it's like, hey, use our product, use our product to help you write.

12:30Anthony Ha: Oh, but also now we're changing the rules a little bit. We're going to put a watermark on here so people will know. Now, that seems like in some some people might have that view that that's like that's a good positive thing. The point is, is that people are pushing back because all of a sudden the companies that are creating these tools and pushing people to use them keep changing the rules in which they're engaged with. And I think that that is not going to end anytime soon. And we're going to continue to see these. And there will always be this sort of tension between the user and the creator.

13:06Well, I think it's also not just the user and the company, but also, you know, governments and regulatory bodies. In this case, it's not just Anthropic deciding on their own that they wanted to add the watermarks, but I guess it's essentially in compliance with EU law. And I think part of the pushback also has to do with the social dynamics around AI and that, you know, as I think a lot of resistance and backlash to AI is kind of hardening that there's almost this kind of paranoia around it. Like you don't want to be caught posting AI content. And so that's part of why people are scared of this watermark.

13:39And, you know, the same way you see like LinkedIn adding these buttons, report AI slop. And so there is this, again, I think to your point, just in the sense of like both these tools are getting pushed on us and yet there's a certain stigma around actually using them.

13:55Anthony Ha: And you're right. The regulatory piece is also changing the role. So it's just creating this very interesting dynamic. It's also starting to reveal, I think, the true costs that are associated with AI, not just on a social dynamic, but also on an infrastructure dynamic. And I'm talking about data centers. There were a ton of data center stories this week. And I think I don't know how either of you view this, but clearly the populace is waking up to the cost factor of data centers. And there's a lot of pushback and protests. But I wonder if the public is understanding it on a broader level, like what this massive build out of data centers will mean.

14:40Every company today is looking for an edge. Better insights, faster decisions, greater resilience. But as AI drives new growth, it also creates new demands on energy and infrastructure. Schneider Electric is connecting energy, automation, and digital technologies to help organizations make smarter decisions and operate more efficiently. Because the future won't be shaped by AI alone. It will be shaped by energy and industrial intelligence. Schneider Electric, advancing energy tech. Yeah, I mean, it's definitely been a very surprising phenomenon to me, just like how much the kind of AI backlash has really focused on data centers.

15:18And I think there's a bunch of reasons for that that we can get into. But part of the reason we're talking about it now is this report that came out in the New York Times this weekend about a prospective Amazon data center in Texas, which to be clear, what we're describing is basically what's been permitted. And the reality will be, we'll see. But in the permits, basically, they're saying, we're going to build an on-site natural gas power plant, because I guess increasingly, these companies that build data centers, they found the easiest way to deal with the kind of power needs instead of trying to hook up to the grid is just build our own power center on-site, and then we don't have to worry as much about the grid.

15:57And that power center is permitted to basically release more carbon into the air than any other power plant in the United States. So if it kind of used that full capacity, it would be the largest carbon polluter in the United States. That's pretty insane. Like the energy is becoming the hugest bottleneck for data centers. So they're pulling it from everywhere. You're seeing like companies spin up nuclear sites again, etc. But this one I think is really notable because it's, you know, Amazon last year were already up 16 % from its overall carbon emissions. And it's moving away from its 2040 net zero pledge.

16:33So you're really seeing how companies and like what they're willing to do to push forward and like the pledges, you know, that they have for the environment, et cetera, are really like downstream of federal policy and federal mood, right? Like the Trump administration made it a lot easier, like rolled back a bunch of environmental protections so that we could build data centers faster. And I think there's just this absolute grab for energy that's happening that, you know, doesn't necessarily take the wider world or climate or potentially even individual communities into account.

17:10Anthony Ha: Individual communities are absolutely pushing back. And I think that's putting a lot of pressure. There's also just the realities of the economics of how to power these data centers. And we've done a lot of reporting on it, but just to level set it for everyone, data centers are expected to quadruple the U.S. electricity use by 2035. So that's a big number. And as a result, you can look at how long it takes to build a natural gas power plant and how long it takes to throw up some solar and energy storage. And the permitting alone and the speed of which you can, you know, put in and install solar is basically creating this interesting scenario in which a lot of renewable energy is now being used to support these data centers.

17:59Anthony Ha: Not because it's a good positive story, by the way. It's economically behooves the companies to do this. And it's good that we have this technology. So there's interesting things that are happening in the background because of all this demand. Obviously, a lot of protests on the community side, but also like a huge amount of installed solar, a lot of energy storage, a lot of startups kind of popping up or that have been existing for a while, but finally getting the attention that they sort of deserve and getting a lot of money behind it. One of those is a company called Form Energy. And I don't know if either of you read this story that our climate tech reporter wrote, but that company just raised$750 million because they have created a 100-hour battery, basically, that is really critical energy storage that can be coupled to support the grid and then also support data centers.

18:58Anthony Ha: So you're seeing these examples because of this. Yeah, I mean, you're seeing it on that side. You're seeing it on the consumer demand side as well, as, you know, people are feeling like their energy use is, you know, dire at the moment. Right. So like I think Reservoir, we wrote about as this company, they raised eight million dollars to make water heaters. Their pitch is basically that water heaters are dumb energy wasters and they could instead store energy as hot water and shift demand to cheap, low demand hours. So, yeah, there's just like all these new ways to kind of think about energy that, you know, data centers are helping with, I guess, will help promote clean green energy in that way.

19:41Anthony Ha: Yeah, I mean, just really quickly, I, OK, many years ago, I was a clean tech reporter. This is when we called climate tech clean tech. and there were companies that were working on all sorts of really interesting things around heat pumps and you know hot you know hot water heaters and you know efficient hvacs and solar and you know a lot of that died off or just didn't get the attention and the vc money it deserved and so there are periods of time in which a product becomes so necessary aka data centers and the demand so great that that it then is an accelerant when it comes to these types of technologies.

20:22Anthony Ha: I mean, form energy is not like didn't just start up yesterday. Right. So I think it's the good news aspect of what is, you know, potentially a great polluter when it comes to data centers. Hopefully more of these startups that have really good, solid ideas and products get more tailwinds and are able to actually bring their products to market. Yeah, I did want to mention one other startup that we covered at TechCrunch this week called Discovered Materials, which is working on making data centers and the chips cooler, and which apparently is a major source of why the power consumption is what it is.

21:03I didn't know that, but it's a very interesting story that we ran. And apparently they're using AI to try to find ways to design chips that are cooler. So I do, yeah, I agree that this is all very promising. And I think that that has been a recurring theme when people bring up the environmental concerns around AI and data centers is that actually a lot of this could be powered by renewable energy. AI could even play a role in making the energy, you know, more environmentally friendly. I think then the question, or certainly one of the questions is, can that sort of happen quickly enough to offset, you know, the tremendous growth in, you know, burning fossil fuels.

21:41Because renewable energy is not appealing just because it'd be, yeah, it's great if we like just get more and more energy from solar, from all these other things. But it's part of the reason it's desirable is because putting carbon into the air is very, very, very, very bad for our planet. And so if like, you know, our energy consumption or energy usage like, you know, quadruples and half of that is renewable and half of that is through, you know, stuff that's increasing climate change, that's still really bad.

22:08Anthony Ha: So to answer your question, is it going to happen fast enough? I'm going to say no, because we've seen companies like circumvent permit processing and things like that to get their power supply, which is oftentimes renewable energy, but sometimes not as fast as possible. But I do think that there's some promising technology out there and companies that have made real headway that are not in like the lab they're actually you know producing they have a product that can be used we are seeing that i just you asked the question the beginning was is it going to happen fast enough and i would say no i mean i think we're going to see definitely more emissions associated with data center build out um and also the other environmental factor that we haven't talked a lot about is the water use which you You know, are there startups working on ways to make the water use in data centers more efficient?

23:04Anthony Ha: Because that's actually just as important, if not more so when it comes to like this data center build out. Yeah, well, I mean, talking about climate, for whatever reason, reminds me of my time working under Kirsten's desk in transportation, under which I often wrote about defense as well. There's a lot of overlap between the two. And we've got an interesting defense story that came out. So I think it's time to move to the deals of the week. So one of the deals is that Joby Aviation, an EVTOL startup, they're building out their defense business with a$500 million acquisition of Resonant Sciences.

23:36That's$500 million in cash and stock. So Resonant makes radio frequency sensor systems. They're based in Ohio. They bring access to classified government programs. This is pretty interesting. I think like we've seen other EVTOL, for those who don't know, this is electric vertical takeoff and landing vehicle. startups do defense deals like last year Archer did a big one with Anduril.

24:01Anthony Ha: Yep. And so yeah I'm kind of curious this is like that one was like we're doing weapons with our like war fighting planes. This is very different. Yeah this one I think is more I'm not going to say more reasonable but it makes more sense because of the other elements of Joby's business. So you've written a lot about Joby in years past when you worked under our desk. We miss you over here. But this one is interesting because Joby started working on some defense piece, specifically a hybrid electric aircraft. So, you know, for those who aren't familiar with Joby, they were working on an all electric EV toll and they're still doing that.

24:43Anthony Ha: And that's going to be used for like urban air taxis and stuff like that. They started doing this hybrid piece and as a defense play because of the longer distances. So this new company that they're acquiring, their technology, which is essentially like radio frequencies and systems, can be then added onto those planes, which then makes it a really, hopefully in Joby's view, what they hope is a very attractive product to the U.S. government as a customer. So that makes a lot of sense. The interesting thing here is that they're creating a whole division. And so to me, that is a signal that they aren't just like, you know, dabbling around and trying to be dual use.

25:28Anthony Ha: They're making a real go of it. And the sense I got reading, Kirsten, your story about this deal was also that it suggests that Joby is in this place where, you know, know, there is this very kind of exciting, promising air taxi vision, but that is still probably many, many, many years off. And they're kind of now at the point where maybe investors are saying, hey, you've got to actually make some money now, not five, 10 years from now. Yeah. To jump in, it's not many, many years away. I mean, they are they are getting close. They've gone through almost all the steps on the the type one certification, which I will not get into.

26:03Anthony Ha: But it is a process that is required under the FAA. They are under a Trump administration pilot program, one of the companies that is going to be doing these piloted flights this year. But you are absolutely correct in that investors really wanted to see, I think, want to see some sort of revenue. So the company they bought is profitable. And it reminds me of another acquisition they made last year, which was with Blade. That's the helicopter ride share. You know, it's a fossil fuel based product, has nothing to do with the overall mission of Joby. But guess what? It gave them revenue and it gave them enough revenue to show a pretty big growth in the same period last year per their earnings.

26:45Anthony Ha: So I think that it is a way of it's a near term revenue plan. I'll put it that way. Yeah, that was going to be my question. Like, how far are we from like, you know, the whole when they were pitching it, it was like, we want to do ride share for the skies, air taxis, yada, yada. But now it's like that's like the goalpost keeps like, you know, the timeline keeps getting pushed out. it's like, all right, we actually just need to get money. It's a tale as old as time when they're like, all right, we need to get money. The government defense. We're moving into the military and the time could not be better, honestly.

27:16I mean, like everyone in Silicon Valley, et cetera. And the government has a serious desire for defense. True.

27:25Anthony Ha: I would say keep an eye out for there's been a lot of shift. I want to talk about one other really interesting startup story, but I will close with this when it comes to Joby. The Olympics are in LA in 2028. I think that you should absolutely look at that as key dates for a lot of these companies who want to or that want to, you know, be launching and showing, showing, showcasing that technology. And you're seeing a lot of movement in the space of EV tolls. I mean, earlier this week, Archer acquired WISC, which is also huge news. So there's a lot of movement happening in this space, something to watch, especially over the next year.

28:05Anthony Ha: And I do think it's not going to keep, you know, someone's going to launch. I'll put it that way. It's not going to, you know, we're not going to have the same conversation three years from now. Well, let's move on to that last deal, which is, well, it's a deal gone wrong, I guess, is probably the simplest and I think a pretty fair way to put it, which is, so there's a video clipping company called Videoverse, and it was acquired by a sort of larger media company called Minute Media for$250 million. and we ran a pretty in-depth piece, I think it was by Russell, kind of digging into what has happened since the acquisition.

28:40There's a lot of accusations kind of flying back and forth. There's lawsuits. And I think that, I mean, certainly one thing we should just emphasize is right now there's sort of a lot of finger pointing. There's lots of lawsuits, lots of allegations, but this stuff hasn't necessarily been sort of proven out in a court of law yet. I'm curious for either of you, what were your takeaways from this story?

28:57Anthony Ha: So I spent some time with Russell on this story, actually. And, you know, to me, it is a cautionary tale, I think, for, you know, all parties involved. And you're right. These are accusations. There are a number of lawsuits, not just coming from some of the investors, so some of the VCs, but also some from some employees who have accused or alleged that the CEO, where I should say almost all of the accusations and allegations are focused on one individual, which is the CEO. To me, this is really interesting and kind of like a classic cautionary tale because this was, this company was a darling, if you will, and was really like highlighting the startup scene in India.

29:43Anthony Ha: It was a success story. And so everything, everything seemed to be going like gangbusters and it just went completely sideways. And it also, I think is a reminder of the limits of due diligence, you know, so like on the investor side. And so there is ultimately still a lot of trust. This is a trust game when you're investing in in startups and how they're going to exit. The trust thing is is really real. It kind of reminds me of I'm not sure if you guys saw a few weeks ago, maybe you talked about on the show, but Dom Dominic Madhuri Davis, one of our reporters, did a story on a study that showed that VC backed startups commit more fraud.

30:21And, you know, a lot of that is just because of the pressure that they have to, you know, meet these insane expectations of investors. And so they kind of like lie about it. They're like, yeah, sure, I can do that. Like, or we've already done it. Right. So that's kind of what my takeaway from this from this story was. I think there's probably any number of kind of startup scandals that you could point to that it might remind you of certainly made me think of, you know, the Frank story, their acquisition and then, you know, how subsequently the founder was accused and ultimately convicted of, you know, inflating a lot of the customer members for the acquisition.

30:57I mean, this story is, I think, at least at this stage, considerably more complicated than that because there's so, again, all the fingers are kind of pointing to video versus CEO, but the accusations and who's mad is sort of, there's a lot of different ones. I mean, one of the ones I found kind of just most eyebrow raising and intriguing was the idea that he allegedly had basically taken the acquisition agreement, but like created kind of fake versions of what the acquiring company had promised and was telling investors kind of completely untrue things about what was actually going to happen in this acquisition, which, yeah, that's a, that's a, again, if that turns out to be true, that is a very impressive strategy.

31:41Anthony Ha: I don't know if we'll ever get the answer to this. And we should say that the CEO was ousted. He is. He's he's been out since the end of April. That's, I think, an important point. He's now unreachable. Yeah, he is unreachable. Yeah. Somewhere in the Middle East, per our reporting. But to me, I don't know if we'll ever get the answer to this, but why? Because you had the exit, man, like two hundred fifty million dollar exit. But why? Because some of the accusations are post exit in which, you know, loans were apparently brought against what that that exit is. So is this is this a story of just greed or is it a story of terrible decision making or to Rebecca's point earlier, like pressure?

32:32Anthony Ha: I mean, I want to know the why as much as probably that's my biggest question. And unfortunately, I don't know if I'm ever going to get the answer to that one. Broadly, my takeaway, like if, you know, speaking from an American point of view, is that it also shows that like the SEC, like can't do much about anything when it doesn't come to like a public company. Right. So, you know, private companies, you know, defrauding investors or investors defrauding each other or or what have you. I think that that creates it's just like the name of the game. Like there's just it's a bit of a Wild West, I guess, which I think is problematic in the long run.

33:10Well, I do, if nothing else, very much hope that we do just learn more about what actually happened with VideoVerse in the future. But for now, we are out of time. Equity, of course, will be back next week. And you can always follow us at EquityPod on X and Threads. Thanks so much for listening.

33:29Equity is hosted by TechCrunch Senior Reporters and produced by Teresa Locansolo with editing by Kel. Subscribe on YouTube or wherever you get your podcasts and find out what's next at techcrunch.com slash events. Thanks so much for listening and we'll talk to you next time.