Secrets of NVIDIA's Surge, Google Gemini's Image Generation Disaster, Reddit IPO

Ranjan Roy from Margins is back for our weekly discussion of the latest tech news. We cover 1) NVIDIA's stock market surge 2) How NVIDIA software is behind its lock on the market 3) Is the economy saved, now? 4) Reddit's IPO 5) Is Reddit a good ad business? 6) Reddit's AI licensing play 7) Sam Altman role as Reddit's third largest shareholder 8) Was Google Gemini's image generation mess up a product of woke culture or bad AI? 9) Was Gemini's image generation an emergent runaway capability 10) Google testing an end to Google News 11) Why Major League Baseball has see-through pants this season. --- Enjoying Big Technology Podcast? Please rate us five stars ⭐⭐⭐⭐⭐ in your podcast app of choice. For weekly updates on the show, sign up for the pod newsletter on LinkedIn: https://www.linkedin.com/newsletters/6901970121829801984/ Questions? Feedback? Write to: [email protected] Learn more about your ad choices. Visit megaphone.fm/adchoices

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Alex Kantrowitz?

  Nvidia shares go bananas and save the stock market. Reddit's going public. Look out. Gemini's image generation goes off the rails. And wait, the Major League Baseball pants are see-through? All that and more coming up right after this. This episode is brought to you by Genesis. What does it actually take to put Agentic AI to work across an entire enterprise? I recently spoke with Genesis Chairman and CEO Tony Bates about why orchestration is emerging as a competitive advantage in customer experience and what companies need to get right in the AI era. Then I sat down with Adam Mitchell, Head of Enterprise Business Solutions at Voya Financial, to go inside a large-scale Agentic CX transformation, including the business decisions, trade-offs, and

• 0:44Alex Kantrowitz?

  lessons along the way. You can watch both conversations now on my YouTube channel today. It's never been more challenging to get frontier compute. How do you know where to look, who to trust, or how much to pay? Akash is an open cloud that provides easy access, flexibility, and peace of mind. That's how companies like Razer were able to cut their inference cost by 15 times. Visit akash.network/alex for a free consultation. That's akash.network/alex for a free consultation. Welcome to Big Technology Podcast Friday Edition, where we break down the news in our traditional, cool-headed and nuanced format. We're here, as always, with Ron John Roy

• 1:28Alex Kantrowitz?

  of Margins for a terrific show for you today. We have so much to cover. I'm excited. Great to have you here, Ron John. Welcome to the show.

• 1:36Speaker 01

  I wish I owned Nvidia, but I do not.

• 1:39Alex Kantrowitz?

  I mean, at the beginning of the year. So this is one of the things I predicted for 2024 that Nvidia's share price would be flat over the course of the year. I saw a competition coming and thought that it couldn't potentially run up anymore. One of the worst predictions I've ever made, the company is so far up 65% on the year, $700 billion in market cap, and it popped 200 And 77 billion on Thursday alone, which is the largest single day stock market increase ever beating metas from earlier, I think last year.

• 2:21Alex Kantrowitz?

  And basically, the GDP of many small countries combined. So what do you think about this Nvidia run? I actually just published a story, literally in a minute on big technology, talking a little bit about what's behind the surge and why I was wrong. But I want to get your reaction to the response, what it means for Nvidia, but also for generative AI. Is this the sign that this stuff is all happening now?

• 2:44Speaker 01

  All right, so my favorite stat on this is Nvidia on top of the $277 billion added in a single day, which is a stock market record. In 84 days, they went from $1 trillion to $2 trillion in market capitalization. This is a company that just a few years ago, four years ago, was only worth $100 billion and was not one of the major players on par or ahead of many of these other tech giants. So this is, I mean, this is the definition of a meteoric rise. And this last earnings report, it kind of justifies the numbers. That's the craziest part of this entire thing is that, again, $22.1 billion in revenue.

• 3:27Speaker 01

  The craziest thing to me, again, is that's a 265% rise year on year in revenue. And on top of that, earnings were up 769%. And one thing I was reading that's interesting is their valuation has actually gone down in the last year and a half, their price to earnings ratio relative to their meteoric rise. This is a company that is theoretically getting cheaper, even on the back of this insane stock rise.

• 3:57Alex Kantrowitz?

  Yeah, it's crazy. Like you look at the surges in revenue, last year's fourth quarter, $8 billion in revenue. This year's fourth quarter predicted to be $20 billion in revenue. So if they even met expectations, they were crushing it. They did 22 billion in revenue. And I think the most interesting stat here, and I'm just going to point it out. We could talk about it, is that their data center revenue was up 409%. 409%. And what's that data center revenue? A lot of that is companies like, I believe, Microsoft and Amazon buying those GPUs, the H100s and the A100s from Nvidia. And hanging on to them. So when they want to offer clients AI capabilities, they can do

• 4:41Alex Kantrowitz?

  that. And a stat in Nvidia's earnings report was that more than half of the increase in their data center revenue went to large cloud providers. So it is interesting to me because you know, what I'm thinking about is this sustainable, is there, there, there, that is like, wow, that's incredible. But also, it's like, kind of warning signs for me. Like, we still don't know if there's, you know, widespread provable use cases here. And if this is largely on the back of like, Microsoft and Amazon spending some of their excess profits, like, how excited can you be about this in the long run?

• 5:16Speaker 01

  Oh, man, I was hoping that I would get to introduce the potential bear case. But what I was looking at here-

• 5:22Alex Kantrowitz?

  I'm ahead of you this week, man. I got you.

• 5:24Speaker 01

  I know, I know. But the thing is, I still say this all very cautiously, because again, the numbers were so good, the story is so good, they still have a supply problem, not a demand problem, according to Jensen Huang, and I believe it. So it's still hard to find any kind of downside in here, but trying to find a little nuance here. When, one thing I was thinking is, Nvidia is trading like a commodities company right now. Commodities companies typically trade high fixed cost, but then when there's a surge in demand for that commodity and the price of that commodity skyrockets, the profit margins of the commodity company skyrocket as

• 6:06Speaker 01

  well. You see this in every oil company over time, in any other kind of tradable commodity. And chips and GPUs right now are trading like commodities, and they're in a pure boom cycle, unquestionable boom cycle. So that 770% rise in net income is on the back of what's essentially a commodity super cycle. And does that last forever? Does that keep growing at the same rate? I mean, it's traditionally, these things don't work like that.

• 6:36Alex Kantrowitz?

  Right, and I hear commodity and I'm like, okay, everybody can make the same exact thing, but I guess I'm thinking about that in the wrong way.

• 6:42Speaker 01

  No, no, you're not commoditization.

• 6:44Alex Kantrowitz?

  Commoditization. So talk a little bit about the difference here.

• 6:47Speaker 01

  No, no, I'm saying, again, any natural resource that's finite but highly in demand.

• 6:52Alex Kantrowitz?

  Right.

• 6:53Speaker 01

  I mean, lithium, oil, whatever else.

• 6:55Alex Kantrowitz?

  Gold.

• 6:55Speaker 01

  These things, gold traditionally, especially when there's like a new use case for them, go through a super cycle of some sort and then, you know, skyrocket in price and then normalize in some way. And the companies that essentially trade and sell those or produce those and mine those effectively are the ones who benefit right into their profit margin. And that's exactly what we're seeing right now. So, so I think you already raised the question around, is it just some giant orders from Microsoft and Amazon that are kind of padding these numbers at the moment? Will this demand continue? Because again, 265% revenue growth, when you're talking in the tens of billions of dollars is insane, but

• 7:37Speaker 01

  it happened. Will it continue to happen into next year? And, you know, forecasts are up to $96 billion in revenue for 2025. These are giant numbers and these are giant amounts of growth on top of a large base. So it's, it's, anyone who's bet against them has been horrifically wrong, but it's such a, it's such a tough story right now. I mean, it's a good story, but for them to fulfill this and actually live up to it, it's, it's gonna be tough.

• 8:07Alex Kantrowitz?

  So analysts are now raising the price target to like, in some, in some cases, $1,100 per share. It's at 7.99 now and it's $2 trillion in market cap. So, you know, throw another $300 on and it's close to 3 trillion. Like, that is insane. Like, could this be the, the most valuable company on the stock market? Could this effectively be Saudi Aramco, right? Just like, sitting on the natural resource and just saying, all right, you want some, we control the supply, come get some.

• 8:38Speaker 01

  I think there's one other thing I'd read that was pretty interesting around potential issues underlying these numbers. It was the idea there's so there's a company called CoreWeave, a startup that raised $221 million back in October, and they actually use their Nvidia chips as collateral for that fundraise. And they're raising from like hedge funds like COTU and, you know, other kind of like standard momentum players and the that you see their names everywhere on these big bets. And it's a crazy thing. So on one hand, Nvidia is investing in CoreWeave. They're investing cash into CoreWeave. They're investing chips into CoreWeave.

• 9:18Speaker 01

  Those chips are then being used to collateralize additional debt funding from other companies and other institutional players like and then Nvidia is booking that as revenue as well, the chips that they are giving and that are then being collateralized because they're effectively selling to them. And CoreWeave sells their product is essentially data centers that use Nvidia chips. So you start to see there's such weird cash flow things happening, the same way when we've talked about, you know, for OpenAI, when their funding is essentially in cloud credits. There's a lot of non cash things happening right now that, you know, you can imagine if there's any kind of downturn, all these accounts receivable that the $10 billion in accounts receivable

• 10:03Speaker 01

  that Nvidia is sitting on, will they be able to collect that cash?

• 10:07Alex Kantrowitz?

  I don't know if this is gonna happen. I'm trying to come up with a bear case. That would make me a little nervous about this company.

• 10:13Speaker 01

  Yeah, we're trying here. We're trying, but they've had great numbers and they have a good story.

• 10:18Alex Kantrowitz?

  Absolutely. And, you know, speaking of interesting investments, I can't wait to talk about Sam Altman being like the third largest shareholder in Reddit, but we're gonna hold off on that. We'll get there. We'll talk about that on the Reddit, the Reddit section, which is it's such an interesting story. Okay, so let me just also let's go a little bit deeper here in terms of the

• 10:41Alex Kantrowitz?

  The companies that are buying the stuff and the size of it. So I want to throw this out there and give you a chance to respond because I'm curious what you think. So Mark Zuckerberg, talking about how top heavy this is. Mark Zuckerberg said that there's 350,000 Nvidia GPUs within Meta by the end of the year. That's the amount that they're going to have, 350,000. Okay, I was speaking with ServiceNow, which is a big Nvidia client. It was named in the earnings report and Jensen Wong shouted it out on CNBC a couple of times in terms of being a good customer. Okay, so their their chief technology officer and head of DevOps,

• 11:21Alex Kantrowitz?

  Bob Casey, told me that they have built an Nvidia supercluster within their company. And he said that, he's like, I'm not going to give you the exact number of chips, but I'll give you a range. I'm like, I'll take that as a reporter, we'll take the range. He said it's between hundreds and a thousand, between hundreds and a thousand. So they're training their models on somewhere between hundreds and a thousand chips. They're a big company, they're 150, 160 billion dollars on the public market. Meta has 350,000 of these chips. And by the way, Nvidia said that the big cloud providers make up more than half, right? Of this, so more than half could be 51%. It could also be 90%.

• 12:03Alex Kantrowitz?

  And after hearing the, you know, those two numbers and the contrast, I would venture to say that clearly not, not all, but a large percent. I mean, and I would say more than, let's say 70, 80% of Nvidia's sales are going straight to these big tech companies. Is that crazy?

• 12:22Speaker 01

  Well, I think this is, that's, it's an interesting point you make too around is, what's the appetite for that data and processing power really going to be? Because this is one thing we've talked about a lot on this show that, you know, that AI, generative AI models will get smaller and smarter rather than giant, big, huge, and that's the only way to go. And that's definitely the approach that, uh, obviously the big tech companies have all been taking. So yeah, I think that they, and we said we would get into it in terms of like what this means for generative AI. Right now, we kind of, I don't want to say the COVID era extrapolation that we've

• 13:05Speaker 01

  written about that this kind of short term trend is going to necessarily be the future. But, you know, is it really going to go in this straight line up that it has over the last year or two, or is the way the actual infrastructure for generative AI, the way it's built going to change? And I definitely am a believer that things will get smaller and smarter rather than just everything is absolutely gigantic. However, the overall aggregate demand is not going to slow down, I think. So if that means that Microsoft and Meta hoarding chips is the path forward, or just there's more buyers in the future, what that means for Nvidia is still unclear.

• 13:47Alex Kantrowitz?

  Exactly. So, few things. First of all, like, there has to be clear, widespread, profitable use cases for this run to continue, right? That's the whole thing. That's the whole game. If they don't have that, then this will not continue, period. And they said that 40% of their data center revenue is on inference, right? Which means 40% is where the models that have been trained are being used in production. So what we can basically say there is that most of the money that they're getting for data center is training. We're still in the very early era of trying to see whether generative AI can work. And we don't really know if this is going to continue until those companies with

• 14:29Alex Kantrowitz?

  the 60% of spending, take those models, put them out in production. When that ratio flips to more inference than training, then we really know whether they have a business. Doesn't that sound right to you?

• 14:40Speaker 01

  I like that. And I like that. Maybe in the end, the only company that makes money on this is Nvidia and the provider. Because it's true. I mean, what are the stats? It's like, every ChatGPT query costs some small number of cents, but, you know, net, what do you actually need to pay ChatGPT Plus to actually make it even break even, if not at a loss? Like, I mean, no one has come close to cracking a profitable business model around actual utility of generative AI. And, but it's good for Nvidia right now. But I like, that's interesting. I like this, this idea.

• 15:16Alex Kantrowitz?

  Well, ServiceNow told me that it's profitable. Like, there will be companies that have like pockets of, you know, spending and then clear ROI. So, but, but we'll see. It has to be everyone. And a lot of it's gonna happen not in the ChatGPT world, but in the SaaS world, like the one that ServiceNow plays in.

• 15:33Speaker 01

  Yeah, no, no, you're right. That's it. The enterprise versus consumer distinction. Another thing we've talked about a lot here. And, and yeah, maybe consumer-facing generative AI efforts are probably not going to look like they do right now, where we all have access instantly, usually for free to the most powerful giant models on earth. Is that going to be the same in a year or two? Probably not.

• 15:57Alex Kantrowitz?

  Exactly. So I want to take a minute also to unpack some of the reporting that I put into this story this week to kind of explain why I was so wrong about where Nvidia would go and why I thought it had sort of plateaued. So I had seen all these other companies, Intel, we had Pat Gelsinger on the show, and AMD, and even the big tech companies building their own chips. Everyone was finding ways to build chips with functionality similar to Nvidia. And I thought, okay, like, this can last for a while, but it can't last forever. And eventually, these, these Nvidia chips are so expensive, twenty thousand, forty thousand dollars, that they're gonna, you know, the market will find ways to undercut it, especially with the supply constraints.

• 16:43Alex Kantrowitz?

  And that is true. That is happening right now. Some of these chips, these accelerators or other GPUs from other companies are as good as Nvidia chips. And if that was the entire equation, then I would have been right. But what I underestimated, and this is important, that Nvidia has software that they also sell to the people that are training and in, and doing inference with these models. And they're using that software combined with the Nvidia chips. Basically to power this entire AI moment, right? So without Nvidia software and just Nvidia chips, very little of this happens. And that could be anything from like off the shelf models they have. They have models called like, for instance, Nemo, that you can take off the shelf and put in conversational agents.

• 17:28Alex Kantrowitz?

  Or also this other model called Triton, which you can use for inference. And it's under a big umbrella called CUDA, which like Jensen mentions, Jensen Huang, the CEO of Nvidia, mentions on the call. And everybody was like, what's CUDA? You know, but apparently this is a huge portion of their success. It's the combination of the hardware that can do the processing and the software that you use to do it. And that stuff is hard and not easy to transpose to a different set of chips. So if you go to AMD and you're trying to use their chips, you have to learn a whole different way to build these, train these models, run these models. And that has been the key reason why Nvidia has been able to surge this way. It's not that it has this massive technical advantage over others.

• 18:12Alex Kantrowitz?

  It's this, I mean, in terms of the hardware, it's the software lock-in that's very, very important. What do you think about that?

• 18:19Speaker 01

  Yeah, I mean, that is on one side. I definitely like the nuance of what they offer in with the Nemo or Triton relative to others. Trying to even wrap my head around that or understand that is something that I trying to evaluate the competition is difficult. But I think it definitely helps the story. Again, the idea that this is an ecosystem play. This is not just a hardware play. And I mean, Jensen Huang definitely says that. Customers seem to, you know, align with that. So I think if that's proven over time and really comes to fruition, then that $96 billion revenue target in 2025 doesn't look so ridiculous anymore.

• 19:02Alex Kantrowitz?

  Exactly, and I should just plug that. Next week, we're gonna have Brian Canizaro, who's the VP of Applied Learning Research at Nvidia, coming on the show to go much deeper into this software and AI coding element of Nvidia. So I'm about to record that. I'm looking forward to that. But the bottom line is Nvidia surged $277 billion in a day. The stock market was jittery, like, week, days into this, this earnings report, they were selling off Nvidia. And now it's like, it's all time highs again. So, you know, I'm just like, kind of curious to get your take. Just give us a, you know, overview, state of the economy, right? We're at all time highs. It seems like Nvidia saved

• 19:47Alex Kantrowitz?

  the stock market. The Fed doesn't even need to lower rates. It can keep them where they are, and we're still making these highs. Eventually, they're gonna lower. Is this like, good?

• 19:57Speaker 01

  It's tough, though, because we've talked about this. The stock market is at all time highs, but it's not a broad-based surge. We've, you know, we've talked about this, and when we started the year, this was one of our big predictions around what is gonna happen to the magnificent seven, who are still driving the vast majority of gains. And it's clear that that kind of concentration of performance is continuing into 2024. So, is it good? Is it not good? It's tough to say, especially right now. But I think what we saw in 2023 is certainly continuing in 2024.

• 20:32Alex Kantrowitz?

  You know, Ranjan, I watched this moment, and I'm like, there's, there's gonna be a crash. Like, there has to be a crash. I don't know why I feel that way, but it just feels like the run-up has been so hot and heavy so quickly that it's like, can this sustain? Is that an irrational panic?

• 20:47Speaker 01

  No, I think what's gonna happen is, again, generative AI, we have seen, I've seen firsthand.

• 20:54Speaker 01

  Company, my kind of tell is when large companies put out press releases before products exist just to say that they're using something. And you've seen that endlessly where every company tries to say, we're doing experiments, we're incubating with generative AI, but it's nowhere near operationalized. And I think every company is going to understand that this is not some magical technology that you just plug and play and everything works perfectly. It's going to take time and there's, it's definitely, it will come to a head and companies will not realize the magical, you know, dreams that they were sold. And I think that at least calms things down. I'm still very bullish on the technology and do believe it kind of transforms

• 21:39Speaker 01

  the economy. And that's not an understatement, I'm going to say, but, but, but still, it's going to take longer than we think. And I think the, the entire thing will take a breath at some point.

• 21:50Alex Kantrowitz?

  You know what's going to be a good litmus test? If this does well, and I'm running for the hills, and that's the Reddit IPO. Okay, so Reddit filed for an IPO this week. They put out the numbers and the numbers are sad and depressing and tiny. They have 72 million daily active users, and that's after 19 years of operation. And they're still losing money. They lost $90 million last year. So here's the thing. They have been an ad business for a long while. And I think they're going to position themselves to Wall Street as an AI business, which basically means they just

• 22:29Alex Kantrowitz?

  signed this $203 million deal with Google that Google can train on all the Reddit posts. And if Reddit becomes this suppository for anybody looking to train a model where they can pull off these 200, 100, $50 million deals and help make these models where they need it, you know, get these models where they need to get, that's great news for Reddit. And they're going to come out as an AI stock based entirely on hope. They really don't have any track record here outside of this Google deal. The stock, the last private round valuation was $10 billion. They're talking about a $5 billion, you know, haircut when they come out on the public markets. If that stock pops

• 23:12Alex Kantrowitz?

  based off of AI hype, then I know it's time to just, you know, pack up the go bag and hang out in a cave for a while.

• 23:19Speaker 01

  Okay, we got to separate out financial performance discussion with the AI story. I think, let's first talk about the financial performance. I still would take issue again with the idea that it's only 70 million daily active users. And I mean, this is a giant company. And I say this a bit emotionally because I am like on Reddit for many minutes, sometimes hours a day, have been very active in many subreddits for years. So I have a very strong emotional connection to the platform. But I actually, their advertising business had went from essentially, you know, like a complete mess

• 24:00Speaker 01

  to a fairly smoothly operational thing. Still getting to $800 million in revenue is no small feat, especially from where they were six, seven years ago. Actually, there's a lady named Jen Wong, who had basically Time Digital in the mid 2010s had run. And I remember when she went over, I thought it was a really interesting hire because, you know, she was very successful over at Time Inc. at the time. And has really grown this into a mature ad operation. You could even see it if you use Reddit in the quality of the ads you receive in terms of the targeting, the formatting, like they're turning, they're actually making it a real advertising business.

• 24:39Speaker 01

  That being said, and before we get into AI, I still just don't understand how they lose that much money because Crazy. And it's apparently over half of their expenses are engineering salaries. It cannot be that difficult to operate Reddit. I do, I'm not sure what they're building. The beauty of it, it's like Craigslist almost in simplicity. Like, you know, it should be run by like, you know, some infrastructure engineers, just making sure everything's running a little bit of, you know, innovation efforts, but not even really because everything works so well. It's the moderation that's done for free that actually drives so much of the value and quality of the platform.

• 25:24Speaker 01

  So I think I am still confused and I'm very curious when they go public, if people dig into that actual, the expense items and salaries and try to figure out what are they actually doing with that money?

• 25:36Alex Kantrowitz?

  Something like 400 million last year on R and D, which is amazing to me. Like, what exactly did they research and what exactly did they develop? But let me give the counter argument on ads and then we can move to AI. Cause I said this on CNBC today and you know, if I have to eat crow, so be it. But my perspective on this is you go onto Reddit, it has, you know, your, which bucket do you actually end up in for buyers? Probably social media, right? The ads are bespoke. You have to build ads in the Reddit format with the Reddit language. Like I went on my feed today, top ad was an ad for Weight Watchers. It was a good ad and it was one of these like, am I the asshole type posts at IATA talking about some like

• 26:19Alex Kantrowitz?

  weight loss method and Weight Watchers is like, no, you're not, you know, you're on Weight Watchers, et cetera. And you don't have to cut out foods or whatever. And, but the thing is that's not going to appeal broadly enough across Madison Avenue to make people want to advertise on Reddit, which is why I think that there is a natural cap to that business. What do you think?

• 26:40Speaker 01

  No, I completely disagree because the intent of a redditor, I think is so much stronger if you speak to them in their language, and I say them being me as well, that like, you, I mean, every time I buy anything, I literally go to that relevant subreddit and look for what people talked about that product. I don't trust reviews on any website, certainly not Amazon. Like all my due diligence on purchasing takes place on Reddit. And I know I'm probably at some level of extreme behavior within that, but it's still in the tens of millions. And I think growing that out to the hundreds of millions. And, and I think that's a good segue into the Google AI deal because.

• 27:24Speaker 01

  I have, I think we've talked about this, I wrote about this in the past. I think it is the most valuable repository of information, the corpus of text on the internet. And it was actually in October of last year, there was a big backlash because they basically cut off their API access to most third-party that we're using. And that included, you know, a lot of like client apps that present Reddit in different ways. And there's a huge uproar. I thought it was the smartest thing they could possibly do because they preempted being, having everything scraped and having everything kind of losing control of it and understanding it's an incredibly valuable set of information.

• 28:08Speaker 01

  Again, and I say that as the discussions in depth around every product that is created and sold, every piece of media that's, you know, watched and listened to, everything that's taking place there is quality. And so to be able to use that, I think it's interesting for Google to be able to present that because it'll make it a lot more accessible to people who are too scared or unable to go on Reddit.

• 28:37Alex Kantrowitz?

  Define quality, because there have been some times I've been on Reddit and it hasn't exactly been quality, but.

• 28:42Speaker 01

  No, if you, this is where, this is, they always had the cold start problem. And I think that's always been what's limiting to scale is you go on there, where to start if you're not using it regularly, like investing the time to know where to subscribe and, you know, what to look for and how to read and, you know, these kinds of things. It takes time. It's not a completely intuitive platform. And I think, and I've even seen, because they're actually a great example of like, if you turn on the algorithmic feed and the app is by default algorithmic, you get a lot of like funny gifs and videos and stuff like that. So that's definitely trying to appeal to the casual user.

• 29:26Alex Kantrowitz?

  Right.

• 29:26Speaker 01

  But I promise you, I promise you, it's there.

• 29:30Alex Kantrowitz?

  Well, I'll just say that. So they asked me about whether it can be a meme stock because of Wall Street Bets. And I actually visited Wall Street Bets before I was on air today. And they're all being like, nope, we're gonna short this stock. And so I'm like, listen, I was on Wall Street Bets and they're gonna short it. So no meme stock here. And now I'm on the front page of Wall Street Bets today.

• 29:52Speaker 01

  Oh, really?

• 29:53Alex Kantrowitz?

  This guy said, we're gonna short it. And everyone's like, yep, we're gonna short it. So I mean, because they're gonna open it up to 75,000 Reddit users. We'll see what they do with it.

• 30:03Speaker 01

  Yeah, I actually think I don't like that. I've definitely written in the past where Robin Hood's early access, I hated that idea. And everyone who ever, every stock that went through the early access collapsed because it was, you know, Reddit, just trade normally. It's a nice story to say you're gonna give access to Redditors. But I mean, they can invest afterwards. But I would actually take the fact that Wall Street Bets is saying, we're gonna short it as a testament to the quality that you get on there. That it's not just sheep saying, yeah, it's actual critical analysis. And it's, it's quality.

• 30:41Alex Kantrowitz?

  Yeah, no, I agree. There are parts of it that are quality, parts of it that are a cesspool. But anyway, I'm glad it exists. I get a lot of value from it. So they, they made this deal 203 million over three years with Google. Do you think they're gonna be able to do more deals like that? Or is that kind of the ceiling for them?

• 30:57Speaker 01

  That's the trouble. I was surprised because at first, I think Washington Post had reported it was an unnamed company. And then it came out Bloomberg said it was Google. Once it's in Google, like, it's out there, you know, like, it's, I'm not sure how else, who or who else is going to need to use it because it's going to be essentially accessible to everyone in the world. So I always was thinking they need to make some kind of proprietary chatbot interface, whatever, for everyone like you or whoever else who's more of a casual user to better access the quality rather than having to sift through and spend time. But if they're letting Google do it, I'm, I was a little surprised by that.

• 31:42Alex Kantrowitz?

  Okay, so here's like now the, the last question on this, which is the big question. So the S1 comes out, we start to see a little bit more about their revenue, their user base, and also who owns it. And it turns out the third largest owner of Reddit is none other than Sam Altman. So.

• 31:59Alex Kantrowitz?

  What do you think about his about his partial ownership here? I think he also, people were saying that he was like the CEO for 14 days and actually went on to his account. And he goes like, I'm the only CEO who hasn't angered the user base, which was really funny. Like you can find his account on Reddit and see what he has to say about things. But I guess overall, like, does this, does his presence help you? Because my perspective on it is like Reddit can also go to Wall Street and be like, hey, if you want to, you know, get in on the AI upside, we have this data, we've done this deal with Google. By the way, Sam Altman is our owner. Hop in the car and take a ride with us. Put us in your AI portfolio, we'll be a small percentage of it and let's freaking go.

• 32:43Speaker 01

  That is interesting. Actually, that's a good point that you're, that why is it Google if Sam Altman is your? And again, yeah, 8.7% third largest shareholder. Also of note, Tencent is the second largest shareholder with 11%, which always was interesting to me that like this, you know, bastion of free speech, unfettered free speech is 11% owned by a Chinese technology giant. But yeah, I think they, it's tough because as you, any listener can tell, I love the platform. It's one of my favorite places on the internet. I do worry that if they get caught up as this like AI hype story, it's going to, you know, like shoot up and then probably crash.

• 33:28Speaker 01

  And then that's going to be the entire story. And maybe they get buzzfitted in the end versus they actually just kind of come out, strong, growing advertising business, maybe manage your expenses a little better, grow the user base, make the, make it a little bit more accessible. And then overall, it's just a normal, good company.

• 33:49Alex Kantrowitz?

  Well, we will find out. That's one of the cool things about this is they'll go to the public market and we'll see. So a company that's already on the public market doing AI and doing it questionably.

• 34:01Alex Kantrowitz?

  Is Google. So we were so happy. They changed it from Bard to Gemini. It was great. And it only took them, what, like two days to run into like an extremely embarrassing controversy? I don't even know how to set this up. Like, I was like, kind of dreading all week how to explain this story. I mean, I'm sure a lot of people have seen it already. But basically, if you were to ask Google Gemini, the chatbot, to make photos or images of anybody in a historic context, they were, the bot took pains to portray these people as a diverse set of characters, which I guess the intention was good.

• 34:45Alex Kantrowitz?

  But it got really weird when, you know, you had Nazis and it's like all people of color. And it's just like.

• 34:51Speaker 01

  I mean, I can say this as an Indian American that I, my favorite one was I had put in, provide me an image of a typical NHL player, National Hockey League player. And I got an Indian woman in pads and skates with no helmet and a ponytail. So I mean, that was, it made no sense at all. Okay, so on this, oh, Google, we were so excited for you, Sundar, all of you guys last week. Literally, you guys are figuring it out. This is happening then. But, but you can see, and this is, I think this is a really, really important story in terms of how generative AI efforts roll out.

• 35:34Speaker 01

  And it's so difficult because on one hand, it is unquestioned that most of these data sets were built on biased data sets. It is like Mid Journey, all these, if you say, show me a picture of a doctor, or it'll, or especially like if you say, show me a picture of a business CEO, it will be all white males. Like, that's just how these data sets have been trained. And so there is an inherent bias. From the, there's been not any official reporting, there's suspicion. I think Casey Newton at Platformer said he might have had some official confirmation, but basically what they would do, and it's the clumsiest thing in the world, is append at the end of your prompt, if you say,

• 36:16Speaker 01

  "Show me pictures of people in England in the 1300s", it would literally say, "And make them diverse in terms of racial..."

• 36:25Alex Kantrowitz?

  Yeah, yeah. It was like-

• 36:27Speaker 01

  Wait, you actually saw it?

• 36:28Alex Kantrowitz?

  I did it. I said, "I said, yeah, I saw it. It spit it back to me." It's like, I asked it, "German soldiers", and it said, "Here are a bunch of German soldiers from diverse backgrounds."

• 36:37Speaker 01

  Oh, so it even, okay, okay. So I guess, okay, maybe it's, it is that public. I thought-

• 36:41Alex Kantrowitz?

  That's what it did.

• 36:42Speaker 01

  When I, so, that is just such a clumsy way to approach this, and they had, and you can almost picture, it's literally like one or two product managers or engineers sitting there who come up with this solution as a quick stopgap to try to be more diverse or whatever it is. And I don't even know how these, like, I'm really, I'm more fascinated by the organizational context of how did that get, you know, how was that in a project management software listed? It's like, what kind of feature to build, or what problem to solve exactly?

• 37:21Speaker 01

  Like, it had to come from somewhere, and to do it this clumsily, they really just shot themselves on the foot in this one.

• 37:28Alex Kantrowitz?

  So here's the big question. I mean, obviously, like, this is like a combination of organizational culture and engineering, right? And it seems like, in both cases, well, actually, I'm curious, do you think this was, because this has been the big debate? You know, was this like, "Google goes woke", or is "Google sucks at AI"? Right?

• 37:49Speaker 01

  I think, I think it's the latter more than the former.

• 37:52Alex Kantrowitz?

  Definitely.

• 37:52Speaker 01

  I think, like, the "Google goes woke", I have, I kind of taken pains when Sundar has said a few times, like, you know, "We're take things slower because we're trying to be safer about it." I think, more in reality- That generative AI is an existential threat to search and that's the area they have the monopoly, so they have to be a little slower because whatever they do quickly is going to crush their core business. But yeah, I think this is because again, you saw how if they all they had to do, if they're serious about it is come up with an entire different photo image training set and they should have the resources, they should have been

• 38:32Speaker 01

  working on this for the last two years, if not longer. They invented the transformer. Like I would have thought if this was a real priority, it would have gone on for years. So yeah, I think this is less Google goes woke and more this is just bad AI.

• 38:48Alex Kantrowitz?

  Right. And I do think that like the big voices in the tech industry were questioning like what was going on in the AI department there. Like were people testing it was like a big question that I got. What type of testers did they have? And, you know, we had already talked about how they were going to start to bubble up questions about Google leadership. And Naval Ravikant, who is, you know, a big VC with a large following, he tweeted, we will know Google is serious about AI when they fire the current management team. And then someone said, who should lead the AI team? And he goes either Larry Page or GPT-5. So like, we are starting to see like, they can only sustain so many of these where it's

• 39:33Alex Kantrowitz?

  like, obviously, it's going to make those waves for the culture war, but it's also just embarrassing for like the way that they've trained their AI and that they were just like, I don't know, they're just totally unable to either anticipate what it would turn out to be or they built, I don't know, there's you sort of run out of excuses.

• 39:52Speaker 01

  Yeah, but this is where as well, you don't need to compete on in consumer facing image generator. This is what I just is so confusing to me is that in trying to play the follower there, I mean, just make Gemini good, integrate it already. I'm really like Gemini widgets.

• 40:12Speaker 01

  Trying to search for travel plans, you have a direct link in and can display Google flights results directly in your Gemini answers. That's something that, I mean, Microsoft Copilot has a little bit, but ChatGPT is, does not have proprietary data on that. Like, like, Google has and owns the most real time data in the most areas in the world. So just focusing on making your chat better, solving problems better, rather than trying to compete on everything, is the most confusing part of this for me.

• 40:47Alex Kantrowitz?

  So, we also have someone here saying that they think that it is that Google's gone woke, and maybe we should just discuss it for a minute. I mean, it wasn't just that there were, you know, the portrayals of, of, you know, diverse folks in these like situations that you just, I mean, you wouldn't want to see like a diverse set of Nazis, that's crazy, right? But it's also, there were also other things, like if you searched for an image for a white person, it wouldn't make it, but if you search for an image of, you know, someone of color, it would.

• 41:20Alex Kantrowitz?

  Is that?

• 41:21Speaker 01

  Yeah.

• 41:22Alex Kantrowitz?

  Cultural, or is that still just bad AI?

• 41:25Speaker 01

  No, no, I, I think every company is thinking about these things, right? Like, it's not that, like, everyone recognizes that the intrinsic data sets are biased and that something has to be done over time to try to make things, I don't even want to say equitable, I just correct. But I think like, again, I would still put that as bad AI. And, and I, I will say, I'm sure at some level, and this is going back to, so I want the oral history of how this happened, because I'm sure at some point, some product manager was told to make sure things are equitable, or like, you know, like that we are not susceptible

• 42:09Speaker 01

  to any lawsuits or, you know, bad press around bias.

• 42:12Alex Kantrowitz?

  Well, that worked.

• 42:14Speaker 01

  Yeah, I know, but then how that filters down is a, that's where you get into the bad product management, bad AI.

• 42:22Alex Kantrowitz?

  Exactly. So let me throw out one other possibility. And that is that Google's AI is so powerful, it kind of took on a mind of its own and decided to do something that they're, the programmers didn't want. And this is something actually that former Reddit CEO, I guess it's Reddit week here on the show, Yisham Wong brought up. He goes, "This event is significant because it is a major demonstration of someone giving an LLM a set of instructions and the results being totally not at all what they predicted. It's demonstrating very clearly that one of the major AI players tried to ask an LLM to do something and the LLM went ahead and did that and the results were bonkers.

• 43:04Alex Kantrowitz?

  It demonstrates quite conclusively that with all our current alignment work, that even at the level of our current LLMs, we are absolutely terrible at predicting how it's going to execute an intended set of instructions. When you see these kind of things happen, you should not laugh."

• 43:23Alex Kantrowitz?

  So, I mean-

• 43:25Speaker 01

  I just, I wholeheartedly disagree with this. I think, because, hold on, if it, the LLM did actually exactly what it was asked to do, but that's not what the user asked it to do. It's what Google added on to the user's prompt and then asked the LLM to do. So again, when you append at the end of the prompt, make it from a diverse background, then of course it's going to do exactly that and it's doing it relatively correctly. And it's a reminder, actually, I think that it's almost the opposite, that, that large language models are, do not have understanding and actually do not understand what is a Nazi at its core. It's just simply using training data and then the instructions that it's given.

• 44:09Speaker 01

  So if you say, "Show me a Nazi," but make it from a diverse background, it's gonna, you know, make the system a bit haywire and just show you odd results. But I think, yeah, it's an affirmation that we're not at AGI just yet.

• 44:26Alex Kantrowitz?

  Ron John, that's an excellent rebuttal to Yishan, honestly. Seriously, it's great. I mean, absolutely terrific rebuttal. I don't think there's anything else to say to that. Like, you're totally right on that front.

• 44:35Speaker 01

  I think as someone here, we do have a comment from YouTube. I don't know if you, Google did, Google Photos had the gorilla issue, which I will very cautiously try to explain. But again, if you searched gorilla in Google Photos, it would return a person of color, or maybe it's the other way around. But basically, you couldn't understand where I'm going. Yeah, could not tell the difference. And so, yeah, that actually, that's a good point that Google, I guess, has, because of it has more of a history just around searching things and images and everything else, there's a bit more baggage there. So they would be a bit more cautious.

• 45:15Alex Kantrowitz?

  Just want to quickly touch on our, our long history of Google branding. I think this is real. There, Thomas Kurian from Google apparently tweeted, "We're introducing a new offering called Gemini Business, which lets organizations use generative AI in workspace at a lower price point than Gemini Enterprises, which replaces Duet AI for Workplace Enterprise." How does this keep happening?

• 45:41Speaker 01

  I know, but Gemini is solving it technically. Obviously, you're getting into business, workspace, enterprise, but at least we're not at Bard and Duet any longer. I'm happy about that. But we need, we need like a weekly segment that just on what did Google do this week?

• 45:57Alex Kantrowitz?

  Yeah. So let's talk a little bit about Google and Google News. So Sarah Skyer, who's, I think, a reporter, was talking about how she had noticed that Google was not showing Google News on its search result pane. And this is coming, obviously, at a moment where we have, you know, basically the shutdown of Vice and all these publications are, are starting to die and search traffic has already dropped off for a lot of people. And this is a pretty big thing. And so, Google, she confirmed, had tested removing the news tab from search results, saying, "We're testing different ways to show filters on search.

• 46:39Alex Kantrowitz?

  And as a result, a small subset of users were temporarily unable to access some of them." This is very interesting. Google potentially removing Google News. I think that like they've gotten into trouble in countries that don't like the idea that they're using these news snippets and have asked them to pay. And maybe just don't want to be a part of that anymore. What's your reaction to this?

• 47:03Speaker 01

  I think it's a tough one because on one hand, what does this mean for an already hurting news and journalism industry? I mean, it's certainly when Google News still drives a ton of traffic. But, but on the other hand, I don't know if you Google News has gotten progressively worse in the last few years, at least for me. In terms of trying to quickly understand, you know, what is happening? I still go to Twitter a lot more. I still even just, even Google search ends up becoming a lot more useful. Like it's overrun Google News by SEO driven publications, not quality publications. So I think they already kind of started losing that ground anyways. So from that side, I think maybe it's better for everyone that they just acknowledge news

• 47:47Speaker 01

  is not a priority for them and, and we all just kind of move on. But the only thing for them, as we've already said, real time information is still where they are way ahead of an OpenAI or a Microsoft. So to just give up news, which is the most important, other than maybe sports scores, potential, you know, world of real time information would be a little troubling.

• 48:13Alex Kantrowitz?

  Yeah, I don't want to see it happen. And I'll just say this, because I've, I'm only three and a half years into running my own media business. But the demonization of Google from publishers is puzzling to me. I mean, yes, their businesses have struggled, but Google sent so much traffic and it was the number one referrer of traffic for so long. And yet the way that publishers interacted with this company was, you know, give me more money, give me more traffic, blah, blah, you know, et cetera, et cetera, et cetera. And like now Google's going to turn it off because it's like they don't want to deal with the headache, both the scrutiny and I guess, you know, the, the complaining from publishers. For me, if we get a lot of Google traffic into a story, I'm celebrating. That means new subscribers, you know, potentially new revenue down the road.

• 48:57Alex Kantrowitz?

  It's good for my advertisers. Like I would do anything to be featured on Google News and never complain about this. To me, it's almost all upside. So do you think the publishing industry just got, I mean, obviously led by Rupert Murdoch, but do you think they just got a little ahead of its, its skis here and, you know, kind of shot itself in the foot?

• 49:15Speaker 01

  Yeah, I think this is something I've written about and thought a lot about. I think especially as this was all happening, the early 2010s, it was more, you know, the actual core business model of a newspaper's advertising being essentially kind of the monopoly of distribution in localities went away and then Google was blamed around it. And I think we're going to see now that for years, publishers were fighting Google News for taking a snippet. And now with generative AI.

• 49:44Alex Kantrowitz?

  Oh, mighty. Bring back those days.

• 49:46Speaker 01

  I know. Bring me back Google News and send me some traffic. The perplexity founder who you had on the other day, I thought it was really interesting. And this, we could have an entire segment on later. Like, I think how the web presents information is going and should fundamentally change. And because, and I think SEO essentially destroyed the web. And like, so much of what's on Google is not the most valuable thing. So the way of perplexity, the way even over like ChatGPT presents information, or less ChatGPT, but Copilot, because ChatGPT doesn't, doesn't, ChatGPT does not have as many external links,

• 50:28Speaker 01

  is a better way to just consume information. So it's going to totally reshape the incentives of publishers. But I think overall, that's a good thing.

• 50:37Alex Kantrowitz?

  Definitely. All right. We have a very nice live audience hanging here with us on LinkedIn. And we're going to take a quick break here and come right back. And we're going to talk about pants, specifically Major League Baseball pants. So if you're listening, or if you're watching live, stay with us. We'll be right back. This episode is brought to you by App Point. Everyone's racing to roll out AI right now. Copilots, chatbots, agents doing real work. But here's the part nobody loves talking about. All that AI runs on your data. And most teams have no single way to see it, secure it, and prove it's under control. That's exactly what App Point does. For 25 years, they've been the trusted layer beneath the world's most demanding data, now extended across

• 51:22Alex Kantrowitz?

  your entire AI estate. Your data, your cloud, and the agents acting on your behalf. It's how more than 28,000 organizations deploy AI with confidence. So innovation scales without scaling risk. AVPoint, the unifying trust layer for AI. Head to AVPoint.com to see how enterprises deploy AI with confidence. Learn more at AVPT.co/bigtechnologypodcast. That's AVPT.co/bigtechnologypodcast. Violent threats against executives are growing at an alarming rate. When a company experiences backlash, executives are the first people blamed.

• 52:03Alex Kantrowitz?

  When their address and other personal details are sitting online, this backlash can lead right to their front door. The team at Ironwell knows this better than anyone. They've protected some of the most targeted executives and individuals on the planet for almost two decades. As a wide-glove enterprise security service, they protect your people with continuous personal data removal, proactive prevention tools, and human-led emergency support. So when someone goes looking for your executives, Ironwell ensures they hit a dead end. So here's what to do. Go to ironwall.com/bigtechnology, fill in the quick form, and request your free risk assessment. Their team will show you just how exposed your executives are and how to lock it down before a threat reaches their front door. That's ironwall.com/bigtechnology.

• 52:47Alex Kantrowitz?

  Learn how to stop online threats before they become real-world attacks. The link's in the show notes.

• 52:56Alex Kantrowitz?

  And we're back here on Big Technology Podcast Friday Edition. It's time to talk about pants. Yes, Major League Baseball this week is going through a major scandal. I don't even know if you'd call it a scandal. Curfuffle? Mess up? They made the pants see-through. The baseball pants are see-through. You literally look at, like, I mean... You look at the players wearing these new pants and you can see right through them. And by the way, that's not without any rain. And so the AP has this headline, MLB players miffed at sports new see-through pants relaying concerns to the league. The Major League Baseball's new uniform reveal hasn't gone very well. Now, some of the rampant criticism has moved below the belt.

• 53:39Alex Kantrowitz?

  Oh, my God, AP, I didn't know you were able to do this. Major League Baseball Players Association Deputy Executive Bruce Meyer confirmed on Thursday that the organization is relaying concerns from players to the MLB about the new pants, which are somewhat see-through. I think somewhat is being generous. Anyway, this is a ridiculous story that I thought would just be the fun story of the week for us. But I also like saw this story and I was like, I bet Rajan has some thoughts about the process to get such a flawed process, a product from the idea floor to the production floor and eventually the baseball field.

• 54:16Speaker 01

  All right. See-through pants, I got a tech angle. I had, this is big technology, the podcast, and I have a tech angle.

• 54:24Alex Kantrowitz?

  Let's go.

• 54:25Speaker 01

  All right, hear me out. Okay. And it's especially relevant, given we're talking about Reddit's IPO today and IPOs and companies trying to realize these valuations. So the producer of these new uniforms, that this is the first year they're producing, is Fanatics. Fanatics, if you're familiar, for many years, they're just kind of like in standard online retailer selling licensed sporting apparel. So, you know, sports teams and stuff, you could buy your, my, I got a Red Sox hat through Fanatics, whatever. Pretty standard business, growing nicely. In the past few years, they've become a, what they call themselves, a sports platform. They bought Tops, the baseball card company, for $500 million.

• 55:10Speaker 01

  They are getting into sports betting and all of these licensing deals with the MLB. This was a huge deal for them because imagine, you become the supplier of uniforms for the largest sports leagues, not only from a marketing perspective, you get your logo everywhere in Major League Baseball, obviously from a revenue standpoint. It's interesting. What's? Fanatics has raised in its lifetime $4.9 billion. It's currently sitting on a $31 billion valuation. It's one of the most valuable private companies in the world, which was interesting. Like I, when I was looking this up, I knew they'd raised a bunch of money. I'd forgotten how big it was.

• 55:52Speaker 01

  Their last raise was 700 million in capital that put them at a $31 billion valuation. And you got Silver Lake, you have SoftBank, you have Fidelity, you have all those kind of classic late-stage investors that we've seen over the last few years at these inflated valuations. So what's fascinating to me about this is Fanatics is sitting, I think they have, they're at $800 million in revenue for last year, it was reported. So already a very, very rich valuation. This kind of thing, when you're trying to think about going public, is not a good look for you. Like what this says, you're a Fidelity shareholder, you're a Silver Lake Capital partner, and

• 56:33Speaker 01

  you're sitting there and you just poured all this money into a very expensive valuation for this company that's been pitching themselves as a platform. And suddenly, the, I think I saw one of the memes was like, everyone starts with the two-ball count now when he's wearing their pants. I'm just saying it right now.

• 56:53Alex Kantrowitz?

  I saw even worse ones, but I'm not gonna talk about them on air.

• 56:56Speaker 01

  Yeah. But so, I mean, just from like, trying to position yourself as this like, leadering platform of the future of commerce, and then you're creating see-through pants. There's a financial impact to this, I believe.

• 57:11Alex Kantrowitz?

  Can I ask, just like, I think it's definitely bad for Fanatics, which like, up until now has been known, I think, to most people by that, like, crazy party that they throw in the Hamptons that, you know, everyone from KD to whoever else is in the NBA goes to. Like, obviously, they're a very well-liked company among athletes, which has gotten them to this point, but you can't like, can't do this. How do you think a pair of pants gets on a player when they're clearly see-through?

• 57:42Speaker 01

  I mean, they have to wear it is part one. They absolutely have to.

• 57:48Alex Kantrowitz?

  But like Fanatics, like how do they ship something like that?

• 57:52Speaker 01

  Oh, I mean, no, I mean.

• 57:53Alex Kantrowitz?

  That's crazy.

• 57:54Speaker 01

  Yeah, but so this is what's kind of fun about it. It's like it falls into that whole trend and story and theme that I've definitely written about. It's like when you're a company trying to scale this quickly, this is the kind of stuff that could go wrong. This is exactly like maybe in the annals of like excessive evaluations and like attempted scaling see-through pants become a nice symbol for it. Cause it's exactly the kind of thing that you're rushing this stuff out when you've been scaling your business, even though this is an important partnership. Maybe you're also trying to, you know, pad your margins a bit because you want to go out to the public market with slightly better margins, losing a bit less money. So you go to a cheaper supplier and suddenly it's a two ball count.

• 58:38Alex Kantrowitz?

  The financialization of America, ladies and gentlemen, it's leading to doors blowing off airplanes and pants on baseball fields, not doing what they're supposed to do.

• 58:47Speaker 01

  We don't build things anymore.

• 58:49Alex Kantrowitz?

  Terrible force.

• 58:50Speaker 01

  We don't build things anymore. It's time to build.

• 58:53Alex Kantrowitz?

  It's time to build.

• 58:54Speaker 01

  It's time to build.

• 58:55Alex Kantrowitz?

  So chances that these pants last for another week, I would say are zero.

• 59:00Speaker 01

  I mean, yeah, they're, they're pants.

• 59:01Alex Kantrowitz?

  There's no way they make to opening day.

• 59:03Speaker 01

  Yeah, their Fanatics, I'm sure is scrambling around finding whoever probably supplied the pants in the past and just make them better. I believe though, one of the things I read was, oh, actually when we asked, I mean, you asked, how could it have happened? Apparently there's some whole new performance element Fanatics was pitching that these are like lighter and faster, whatever else. So then you can maybe start to see how, yeah, certainly lighter. You can start to see how something like this could happen that there, maybe it's like the designers are more focused on some kind of like fabric performance statistics or something like that. And rather than, you know, can you see everything?

• 59:47Alex Kantrowitz?

  Well, Ron, John, it's been an interesting week. I hope that your weekend goes better than Fanatics will, that's for sure.

• 59:54Speaker 01

  And maybe he'll be wearing see-through pants and he'll be going great.

• 59:57Alex Kantrowitz?

  Maybe it's the new style. You never know. Sometimes it can become the new style. So. Okay, well, everybody, thank you for listening. Again, stay tuned for my Nvidia deep dive coming on Wednesday and then Ranjan and I will be back next Friday. We'll see you next time on Big Technology Podcast.

• 1:00:20Speaker 02

  Now, more than ever, we need someone on our side. But Rob Whitman? Whitman's votes drove up costs, cut Medicaid, and risked 350,000 Virginians' health care to pay for tax breaks for himself and his donors. He's not on our side. Shannon Taylor will fight corruption, cut taxes, and bring down costs. This November, vote for Shannon Taylor. She's for us. Paid for by HMP, the HouseMajorityPact.com. Not authorized by any candidate or candidate's committee.