Chinese AI Just Made History

0:01Chinese AI just made history. You're listening to Motley Fool Hidden Gems Investing.

0:10Welcome to Motley Fool Hidden Gems Investing. I'm your host today, Jon Quast, and I am joined by Foolish contributors Matt Frankel and Rachel Warren. Today on the show, we have multiple topics. We're going to be talking about that lead, Chinese AI. We're also going to take a question from the mailbag regarding digital advertising. But first, we want to talk about what happened yesterday. Millions of soccer fans around the world watched Spain defeat Argentina in the World Cup final. And I think that we were all just moved throughout the tournament as many people from around the world visited North America and shared their experiences online.

0:48That was so much fun. But one of the things that maybe we didn't hear about was how much the World Cup has really propelled adoption for the predictions markets. Kalshi specifically said that they got 3 million new users during the World Cup. So clearly that is pushing this whole space forward. And Rachel, I want you to talk about where we're at competitively in the prediction market space because you have companies such as DraftKings and FanDuel out there, but you also have Meta looking to get in on this space. So what can you tell us about how this market is growing and how these companies want to profit from it?

1:29The trading activity that we saw around the World Cup final with Spain's victory over Argentina, that was one of many examples we've seen that sort of are serving as proof of concept for sports event contracts, which if you're not familiar, these essentially treat match outcomes like peer-to-peer financial derivatives rather than traditional sport wagers. So Kalshi, which you mentioned, John, they cleared about$1.9 billion in trading volume on the final match alone. So to understand exactly how this works and where it differs from, say, traditional gambling. So traditional gambling tends to involve an individual wagering directly against a bookmaker, for example, who profits from their losses.

2:07Event contracts like these operate as an exchange where peers trade financial derivatives against each other, and then the platform collects a flat transaction fee. So CalSHI, for example. Now, because these contracts are legally classified as commodities, they actually fall under the jurisdiction of the Commodity Futures Trading Commission, or the CFTC, instead of state gaming boards. And that's a very important distinction because it essentially allows these prediction markets to bypass the state-by-state licensing laws and heavy gaming taxes that the traditional sportsbooks have been forced to navigate.

2:41Obviously, there are some vulnerabilities in these business models. You tend to see trading volumes and even liquidity plummet once some of these cultural events wind down. But you're seeing a lot of the big tech and legacy sportsbook players deploy their own opposing strategies to try to capture and retain the user engagement that they're seeing these platforms like Calci and others capitalize on. So talked about recently how meta platforms, they're entering the space with their internal application that they codenamed Arena. This is essentially an AI-driven, non-monetary framework. And because of that, they're able to bypass really strict financial compliance rules and capture engagement data from their billions of users and avoid a lot of the regulatory friction.

3:24You've got the traditional players like DraftKings and FanDuel, which you also mentioned, John. They're dealing with severe margin compression right now. they're launching their own low fee event contract products to try to really protect those embedded customer bases from churning to those lower cost financial platforms like the of the world. So we're seeing this shift towards event contracts, if you will. That's the term very much this asset light exchange model. I mean, this is a market that's expected to approach a trillion dollars by the end of the decade. So there's a lot happening in this space and a lot to watch, whether or not you participate in it.

4:01Matt, I wanted to bring this topic to the table today because of a Kalshi study that came out fairly recently, and it really bothered me personally. So according to the study that Kalshi released, 89 % of people say that buying stocks or mutual funds isn't gambling. Okay, that's fine. But the majority of the people in the study also felt like predicting on the outcome of events, like what we're talking about, such as Spain versus Argentina, predicting the outcome. Most people also view that as not gambling. And that's such an interesting thing. So for most people out there, according to this study, they would view it fundamentally the same, investing$100 in the stock market and betting $100 on the outcome of an event, such as Spain winning.

4:50And what I want to talk about here with you is what do you think about that? Is it fundamentally the same thing or is it different? Because I think for some people, the idea is I have to research the two teams that are involved in the game. So there's an element that I don't know the future, but I've researched to make an educated opinion about the outcome of the event. So for some people, that's no different than investing in a stock.

5:14Jon Quast:I understand why people might feel that way. And I understand the appeal of the projection markets, especially after watching that game. I mean, 115 minutes with no scoring. You need a way to make it interesting. I understand why people feel that way. But at the same time, the key difference is whether the underlying asset you're talking about is expected to compound in value over time, rather than just kind of settle in a binary zero sum matter. Buying a share of a business, it gives you a claim that hopefully, not always, but hopefully, will grow earnings. They'll reinvest capital. They'll create value for you over time.

5:47Jon Quast:Even if you're wrong about the next quarter, your investment goes on. On the other hand, the stock market, it's a positive sum game. So economic growth, productivity, innovation, reinvestment, they can all make everybody richer over time and over the past they have. On the other hand, a prediction market, it's a peer-to-peer market. It's a zero-sum game. For every dollar someone won betting on Spain, someone, or excuse me, predicting on Spain, someone lost it on Argentina. no new value was created there if anything value was lost because you know robin hood and calci take their cut money just kind of moved around sideways so the general rule here is if you can lose 100 of your money because of a single expected event that you know an event that's going to happen one way or the other you're speculating not investing regardless of what robin hood or calci might call it and that's true in the stock market as well if you buy like a an out-of-the-money call option.

6:41Jon Quast:It's a binary event. It's the same idea here. So yes, they are securities in that sense, but they're dependent on one binary outcome event. And that's the really big difference between buying a stock and predicting on the outcome of an event. Yeah. And to anyone listening, there is some entertainment value perhaps in the prediction markets. And certainly we wouldn't want to tell anyone out there that you should definitely avoid it at all costs. Maybe there's a case where you can use it responsibly. But I think for me, it's really important to remember that we are talking about two fundamentally different things and to keep those separate in our minds.

7:18In one category, we're investing for the future. In another category, we're maybe playing around with a little bit of money. That could be okay, but they are different things. So keep that in mind. When we come up after the break, we're going to talk about how China is disrupting the AI market. You're listening to Motley Fool, hidden gems investing.

7:37Jon Quast:Support for the show comes from Fundrise. For the past 70 years, there's been a room in finance most people couldn't enter. A room where you could have invested in some of the biggest names in tech, companies like Airbnb and Uber before their multi-billion dollar IPOs. I'm talking about venture capital. Fundrise recently took a sledgehammer to those closed doors by launching a venture capital product that's available to anyone. Their mission is to give everyone the chance to invest in the best tech and AI companies before they go public. You can visit Fundrise.com slash Fool to check out Fundrise's venture portfolio and get in early today.

8:12Jon Quast:All investments involve risk, including the potential loss of principal. Past performance is not indicative of future results. This is a paid advertisement. Welcome back to Motley Fool Hidden Gems Investing. So in the last week, something massive has happened in the AI world. According to Arena.ai, a model called Kimi K3 has jumped to the top spot among AI models. So it's ahead of Anthropics Fable 5. It's ahead of OpenAI's GPT 5.6. Rumors are that it is three times cheaper to run than other models. And that claim alone right there is absolutely astounding. Rachel, here's my question. Does this change the game in AI?

8:59Because from my perspective, if businesses can use a cheaper and better model out of China, then of course they were going to adopt it. And if they adopt this model, then that means that they're moving away from other models. And maybe that impacts the economics of some of the top AI providers in the game right now. I'm not going to downplay the fact that we're seeing immense technological and AI advances coming out of China. We've seen a lot of really impressive models rolled out in recent months. But I do think it's an oversimplification to look at the launch of Kimi K3 and say that that might defeat the likes of OpenAI and the Silicon Valley-backed players.

9:39And there's a few reasons for that. I think a lot of the focus has been on the fact that the blueprint behind Kimi K3 is essentially free, but there's also very much the physical the laws of economics and computing power. Hardware is still the key bottleneck here, right? Not the software. And KimiK3 is an absolute data monster with 2.8 trillion parameters. I mean, the fact that Moonshot AI, which launched this model, they had to freeze new user signups just 48 hours after launching because their servers literally hit a physical limit. I think it continues to prove that the computing power behind all of these models that we're seeing continues to be a finite, scarce resource.

10:16So even when that software blueprint, if you will, is free, running it safely at scale can be really difficult. It can be really expensive. And I think we're still seeing a lot of CTOs are still going to want to pay a reliable, secure subscription fee to the likes of Anthropic or others to handle that key infrastructure challenge. So yes, we're seeing that a lot of the raw AI intelligence is becoming a cheap commodity. I think that that will be increasingly so in the years ahead, but the cloud infrastructure required to run it, not so. I think that that could change through the years, but a lot of the business models, I think, for these closed providers like OpenAI, like Anthropic are safe because they're really selling the stable ecosystems that make that software usable for big businesses.

10:59And that's not something that's going to change anytime soon. That being said, I think there are a wide range of useful models out there. I think democratizing the space is important, but I do think it's important to understand that this is not going to just disrupt the dynamic of OpenAI and Anthropic overnight.

11:13Jon Quast:One thing that I would add is that having a capable AI model is one thing. Having a cheaper AI model is one thing. But having enterprises trusting in your product is another thing altogether. Cheaper inference for tasks, it doesn't automatically hurt companies like OpenAI and Anthropic. The highest value cases for AI need more than just raw text generation. The selling points are how these ecosystems are safety tested. They're reliable. They have other key features that enterprise clients want, like how Rachel described selling the ecosystem. I mean, the bottom line, I don't think this changes the game for OpenAI, Anthropic, and all the other ones.

11:51Jon Quast:I mean, having a capable but capacity constrained rival, it really kind of underscores just how the hyperscalers that have these large pipelines of compute that keep growing still have the clear advantage in this space. Yeah. Yeah. And that's an interesting point to bring up here. Both of you have alluded to it already, but let's just make it explicit. Kimi actually had to pause new subscribers. So it had basically when the news came out that it was now the top model and some of the people started coming out saying, Hey, look at what we're doing and look at what it's costing us compared to the other models.

12:25There was such a surge in subscriber demand that the GPUs from Kimi could not keep up. And so they actually had to say, listen, we can't actually even take new subscribers right now. We're going to have to hit the pause button. We're going to invest in compute so that we can meet all this demand that we're seeing. That's a really interesting thing to think about. But what I got to thinking about in this was NVIDIA. And I know we're the hidden gems team. And I know that NVIDIA is either the largest or second largest stock in the world, depending on the moment, but it trades at just 22 times forward earnings.

12:59That's actually kind of cheap. and just last week it started shipping h200 chips to china which isn't really in the calculus right now now you have a chinese model saying hey we're going to actually need to invest more in compute nvidia just now starting to ship to china can we shift here to nvidia stock for a second and say is there a case that nvidia stock is actually a good buy right now yeah i think that nvidia looks like a really strong buy right now for a variety of reasons. I mean, going back, this bottleneck that is obviously affecting players across the industry, but forced KBK3 to freeze subscriptions.

13:36I mean, that key pain point is where we're seeing that global backlog of guaranteed revenue for NVIDIA come from. I mean, companies trading at just about 22 times forward earnings last I checked. I would say personally, I think the market's priced in a lot of the AI fatigue, the geopolitical risk. I do think it's a really healthy entry point into a company that's still really healthfully growing its data center revenue. It's interesting, that newly approved shipment of H200 chips to China, I do think that there's a nice regional tailwind there. I mean, NVIDIA's broader growth engine is still the insatiable demand from the Western cloud giants that are racing to host these huge multi-trillion parameter models.

14:15So I think if anything, we're seeing that NVIDIA is continuing to dominate the market as this toll booth for the entire AI industry. I don't think that's going to change anytime soon. And it is operating off an incredibly robust financial foundation, really profitable cash producing business, which I think also very much lends itself to being a good buy for long-term shareholders.

14:36Jon Quast:I agree with Rachel that NVIDIA by most valuation metrics looks very cheap considering you said 22 times forward earnings, that growth rate that it keeps posting and the H200s to China. It's a real growth lever and could be a serious near-term tailwind for it. But I don't own NVIDIA in my portfolio, not directly anyway. I have plenty of exposure through ETFs. And there are a few reasons for it. First, things like comparing that PE of 22 to the massive growth rate assumes that the growth is going to continue. That's what the whole basis for comparison is. It's not a realistic growth rate to maintain forever for any company, not just one of the largest in the world.

15:11Jon Quast:Second, NVIDIA has a lot of customer concentration. They sell to hundreds of thousands of customers, but a lot of their revenue comes from the big hyperscalers. The hyperscalers themselves are starting to make in-house chips a lot of times with the stated goal of reducing dependence on Nvidia. So long-term, who knows what that's going to mean. And then the China approval, it's a policy decision. And as we've seen many times, policy decisions can be reversed and put back on and reversed and put back on several times. So I'm not saying that Nvidia is not a great investment or a great business. I couldn't fault anybody for buying Nvidia right now, but it's not a risk-free investment and it's important to put the attractive valuation into context before you buy it's hard to argue with that matt thank you for always reminding us to think soberly about the stocks that we invest in so i'll definitely keep that in mind but i am eyeing nvidia right here myself personally after the break we're going to take a question from our mailbag about digital advertising you're listening to motley fool hidden gems investing trading at schwab is now powered by ameritrade bringing you an expanding library of education with even more ways to sharpen your trading skills access new online courses insightful webcasts articles engaging videos and more all curated just for traders plus guided learning paths with content designed to fit your unique interests no sifting to find exactly what you need so you can spend your time learning to trade brilliantly.

16:39Jon Quast:Learn more at schwab.com slash trading. Whether you're trying to analyze market or business trends or dig through hundreds of pages of quarterly reports, the sheer volume of data can feel overwhelming. Lately, I've been using Claude's deep research feature as my ultimate thinking partner. I hooked it up to my Google Workspace tools and it was able to run a reliable, comprehensive analysis across dozens of different financial sources in minutes, Helping me spot long-term connections that on my own, it may have taken me hours or even days to see. Claude is the AI for minds that don't stop at good enough.

17:12Jon Quast:It's the collaborator that actually understands your entire workflow and thinks with you. Whether you're debugging code at midnight or strategizing your next business move, Claude extends your thinking to tackle the problems that matter. For problems worth solving, get started with Claude at Claude.ai slash fool. That's Claude.ai slash fool. And check out Claude Pro, which includes access to all of the features mentioned in today's episode. Claude.ai slash fool. Welcome back to Motley Fool Hidden Gems Investing. One quick note, we want to make you part of the conversation. So if you have a stock or investing question for anyone on this show, you can send those in to podcast at fool.com.

17:52Our preference is that you keep them short, keep them foolish. And remember that we can't give personalized investing advice. So generalized questions are better. but if you have a question for us send those in at podcast at fool.com podcast at fool.com this question today we did like and it goes like this some of the largest companies alphabet and meta make money mostly from advertising has that come from them taking market share of the advertising industry from newspapers and television or has the total amount of advertising as a share of the U.S. economy, grown a lot in recent years. Rachel, let's start with you here because this is a really interesting question.

18:34Alphabet and Meta are indeed advertising powerhouses combined. You're looking at a$6 trillion market cap. So just how big are the digital advertising businesses for these two massive companies? It's really interesting because obviously this is what we know Alphabet and Meta for, but I don't think we often talk a lot about how that came to be. And these are companies that combined generated hundreds of billions of dollars in revenue from ads in 2025 alone. Alphabet and Meta control about half of the entire global advertising market. For Alphabet, advertising accounts for about 70 % of its total business, give or take in a particular year.

19:13Meta relies on advertising for about 98 % of its entire revenue. Now, what's interesting is you think about where all of this business came from. And, you know, you go back a few decades. it was sort of a mix of a few different factors. It'd be very much, of course, drawing off a growth from legacy media, but also very much expanding the total economic pie, if you will. I mean, if you look back over the last couple of decades, we have seen digital platforms carve out a lot of the local newspapers and television by offering these very highly precise data-driven targeting solutions that traditional media just couldn't match.

19:48and the total amount of advertising as a share of the economy has also grown significantly because platforms like Alphabet and Meta also kind of invented a brand new marketplace. I mean, they lowered the financial barrier to entry. So Alphabet and Meta, yes, they obviously work with these huge brands, but they also allowed millions of small and medium-sized enterprises who could have never afforded a multi-million dollar TV commercial, for example, or a major print campaign. They allowed those players to buy these highly targeted hyper-local ads. And that's also been really, really critical to the growth of those businesses over the last few decades.

20:23Well, certainly the targeting capabilities of digital advertising is it changed the game in advertising. It's no longer just a billboard on the highway that you have to count on whoever drives past it and looking at it and making a decision off of that. Now we can actually target online with intent all of that. It certainly changed the game. But Matt, my question here for you is, did this actually increase the pie or is it just that the pie shifted to digital channels?

20:50Jon Quast:Well, first of all, don't count out billboard advertising. One of the companies that offers it, Outfront Media, has been doing great lately. And I've said before, it's the one type of advertising you can't click away or turn the page from. You're literally forced to look at it. So it has some advantages. But to answer your question, it's both. The pie got bigger. Global ad spending just as a percentage of GDP has risen significantly recently. The reason is a lot of what Rachel talked about because digital ads are generally more valuable to advertisers. They target better and they're just more efficient.

21:18Jon Quast:They took a lot of market share as TV and print ads have kind of been declining for the past two decades or so, and understandably so. Alphabet and Meta didn't just win the old pie. They baked some new pie as well, and they become the pie's primary baker, to kind of use your pie analogy there. Okay, so here's a question for each of you. I want you both to weigh in here. based on the fact that you're saying matt that the pie actually did get a little bit of little bit bigger here i mean it did shift but the pie also did get bigger does the digital advertising pie keep getting bigger from here and if it does does that benefit these top two players or is ai going to come in here now many of these ai companies are looking to get into the advertising game is ai going to come in here and shift the whole digital advertising market and shift who the winners are?

22:09Rachel, you're up first. I think it's a combination of things. I absolutely think the pie will keep getting bigger, but I think that you're still going to see the top players dominate. And AI is fundamentally rewriting how ad money is spent. It's shifting a lot of the winning growth to who controls the backend consumer data. Now, of course, that means that we're seeing the likes of Meta and Alphabet succeed immensely because Because they have these huge capital reserves required to build the AI automated ad infrastructure that these businesses rely on to survive, but also they have all of the data to fuel that growth.

22:44So there's been a lot of fear. There had been these companies would see their business models completely crushed. But in fact, the AI driven targeting tools have actually triggered an advertising boom. Now, one thing I'll note, I mean, AI is changing the landscape in terms of how search and discovery works. We're in a time where conversational AI search tools are compressing the traditional process of scanning links. We're seeing generic web traffic shrinking. We're seeing a real relocation of ad dollars towards those more high intent channels with direct transactional data. Again, you go back to the big players that we're talking about here.

23:16So I do think you'll see new AI startups will capture a piece of the conversational search market. But I think that we're still going to see, at least for the medium term, the alphabets and metas of the world remain the dominant winners because their ecosystems and platforms are very well insulated against AI disruption and also closest to where consumers are making buying decisions and leaving that data that grows that AI flywheel in the first place.

23:41Jon Quast:On the near and medium term, I agree with most of what Rachel just said, but I'd push back a little bit on that duopoly framing. There are just a few things to add here. So number one, don't count out Amazon. Amazon's ad business, it's still not the biggest part of their business, obviously, because they have AWS, they have their e-commerce revenue. It's rapidly becoming really a third major player in ads, and its ads sit a lot closer to the actual purchases you make than either Google or Facebook ads do. Not only that, many traditional retailers that have big e-commerce presence like Walmart, like Target, have also been kind of quietly growing their own sponsored ad revenue.