The Week: China Is Undercutting America’s AI Boom

0:00Support for the show comes from MongoDB. AI-assisted and agentic coding is helping you build faster than ever. But if your data layer is a bottleneck, what's the point? Instead of wrestling with rigid schemas or translating data formats, MongoDB's native data model mirrors the language LLMs already speak. It ships at the speed of AI, is ACID compliant, and scales to handle massive Fortune 500 workloads. Ask any developer. It's a great database. Start building at MagoDB.com slash AI.

0:58Switch. Uprunt payment of$45 for three months,$90 for six months, or$180 for 12-month plan required. $15 per month equivalent. Taxes and fees extra. Initial plan term only. Greater than 50 gigabytes may slow when network is busy. See terms. This episode is brought to you by Google Chrome. You think you know a browser, but Gemini and Chrome? That's new. It can help you with practically anything on the web, like restoring a vintage motorcycle from a 50-page restoration block, or finally break down that long article you've had open for weeks. Gemini and Chrome is here for it. Ready to make anything online make sense?

1:26There's no place like Chrome. Check responses set up required compatibility and availability varies 18+. Welcome to the week from Prof G Media, where we break down what mattered and what it all means. I'm George Han, and it's Friday, July 24th. Today, China is challenging America's AI dominance with cheaper models and a better global reputation. Then, the debt piling up beneath the AI boom. And finally, how the casino economy is turning young men into monks. Let's get into it. Last Thursday, a Chinese startup, Moonshot AI, released a model named Kimi K3, the largest open-weight model ever built.

2:17On some benchmarks, it outperforms the best models from OpenAI and Anthropic. It also runs at about a third of Anthropik's price. On Monday's Prof G. Markets episode, Scott argued that Kimi is part of a much bigger threat to America's AI industry. Well, I think the biggest story that has, that it's got almost no coverage is that, is what I think is, I would refer to affectionately as, is AI dumping from China. and that is free Chinese models went from less than a third of all traffic in late 2025 to about two-thirds recently and it's much less expensive. I mean essentially what's happening in the air market has condensed what's happened to industrial economies in Europe but instead of it taking two decades it's taken two months and that is they steal the IP on a cell phone tower from Siemens and then they manufacture it with cheaper labor and they sell it back to telcos for 40 cents on the dollar.

3:18And there are a lot of regional LLMs in China, and they have access to cheaper power, cheaper chips, less power consumptive, and they're also subsidized by their own local governments. And I believe, and I said this before, that Xi has a vested interest that the ultimate Ohio-class submarine to be flipped against us is to engage in AI dumping. And I think it's happening. And you hear stories of VCs encouraging the portfolio companies to use these open-weight Chinese models. But you're seeing the best business model in history is IP theft. A close second in terms of a pricing mechanism is 80 percent of the market leader for half the price.

3:58And DeepSeek and these Chinese open-way models claim they're 90 percent of the market leader for a third of the price. I think this is going to be the next geopolitical pawn once Trump actually, if Trump actually figures out what's going on here quickly enough. Ed's read is that an AI price war has already begun and that America's leading labs are fighting it with money they don't actually have. I think the most significant and the most important is the price war that appears to be breaking out right now. And that is the pricing of these AI models is clearly becoming front of mind for enterprises.

4:36And it is very quickly becoming a race to the bottom because you have Meta that is coming out with their model, which is dramatically cheaper. And they have literally stated that it is their mission to initiate a price war because they know that they can play that game. They have the cash. OpenAI and Anthropic can only play that game for so long because where are they getting their money from? They're not getting it from profits. They're getting it from venture capital. And so at some point, if the investor spigot turns off, they can no longer play that game. Gemini is also pricing their models lower.

5:09But the elephant in the room, and I'm 100 % in agreement with you on this, is clearly the Chinese models. It is clearly DeepSeek, Kimi, all of these Chinese models, which are so dramatically cheaper than the US models that it's hard to actually fathom. Just to go through the data here, the price per million output tokens for OpenAI's model, GPT 5.6, is$45. For Claude Fable 5, it's$50. The equivalent price for DeepSeq's model is 87 cents. So it is 99 % lower than the American alternatives. And obviously now we're seeing that China is stealing market share. I don't know. I mean, we see all these benchmarks of like, which model is the best.

6:03And there's all these debates like, you know, people kind of generally assume that OpenAI and Anthropic have the best, best in class, world class models. I'm not going to chime in on that. But what I do know is that people are using the Chinese models because the Chinese models can get the job done fast enough, well enough, and they are so much cheaper that it would be ridiculous not to turn to them, which is why you're seeing the increase in market share. It's also why you're seeing a lot of these blue-chip American companies that are making that switch. That's the supply side of the story. But on Tuesday, Ed asked Charlie O 'Neill, co-head of model training at the AI infrastructure company Base10, whether this is really a fight between two countries.

6:48O 'Neill has spent his career working with both open and closed models. He thinks nationality may be a distraction. Yeah, I think the big story here is not necessarily Chinese models versus American models. I think the big story here is open source versus closed source. So obviously the story we've been told for the last several years is that closed source is going to continue to dominate. The American frontier closed source labs are going to continue to pull ahead and open source will never catch up to that. And I think what we're seeing with Kimi, with other Chinese models like GLM, GLM caused a very, very big wave.

7:23It may not have done the rounds in the same way that Kimmy did, but it was certainly a great model. And even releases like Inkling from Thinking Machines, which is an American company, what we're seeing is that basically the recipe to build these things, there's no secret sauce. The big labs, they don't have anything that the open source labs don't have. And open source is going to continue to improve the capabilities and intelligence of the models they release as we scale up the size of these models and the amount of data and compute that goes into them. And so, yes, from one kind of aspect, it's concerning that this is like a Chinese model that is leading the charge with this sort of like open source versus closed source debate.

7:58But I think there's really promising signs for the open source ecosystem in general. And I think a lot of people are starting to realize that that's potentially a better world to end up in compared to where you have maybe a duopoly with open-air and anthropic having these models that pull away from everyone else. And they dictate all the terms of access and control that intelligence. The models are just one front. Reputation is another. And China is gaining ground there, too. Pew found 46 percent of people worldwide now view China more favorably than the United States. Just three years ago, that split ran decisively in America's favor.

8:37And it isn't only Europe. Here's China Decode's Alice Han. It's pretty shocking. You look to the U.S.'s closest neighbors, Canada and Mexico. So Mexico, there's a 19 percentage point swing in favor of China, where 59 percent of people have a more favorable opinion of China versus the U.S. In Canada, there's an 11 percentage point swing in favor of China. I mean, it seemed like only yesterday where we had scandals involving the two Michaels and Canada, you know, putting on sanctions on China, et cetera, et cetera. It's astounding to me that the two closest neighbors and most of Latin America have a more favorable view of China versus the U.S.

9:19The divide shows up inside the AI story, too. Alice discussed it with guest Selena Xu, former China reporter for Bloomberg. 84 percent. This is a shocking number. 84 % of Chinese surveyed respondents said that they were more excited than worried about AI versus in the U.S. 10 % said that they were more excited rather than worried, meaning that a higher number potentially could be worried in the U.S. about AI. But it speaks to how there is quite a dichotomy and sentiment between Americans and Chinese. Can you explain that? I think this statistic took a lot of people by surprise. There were a few other surveys by like Edelman and others that people started paying attention to.

10:00But I think as negativity in the U.S. towards AI grew to quite a feverish pitch this year, people are suddenly like, oh, my God, why is China so much more positive? And obviously, you know, the surveys are not fully representative of sentiment on the ground, which is often not monolithic. But being in China a few times this past few months, I think the main thing, the main attitude difference that I sense between, you know, Chinese and Americans is that in China, people are very pragmatic. They don't see AI as this, you know, in general, this machine god. Talk of AGI or artificial general intelligence is pretty much absent from most convos, other than if you're talking to a very specific minority of westernized, you know, policy elites.

10:47And for the average Chinese person, AI is either a tool where maybe they have to learn because they are mandated to by their bosses and their jobs. all these very doomsday, what people now call doom trolling in the U.S., of like, AI will make us extinct. AI machine god's going to come. It will be like the Terminator. It's terrible. There's none of that, even with all the humanoid craze in China, which actually is another structural advantage on hardware side for the U.S.-China AI race for China. Cheaper models and a public that is far more eager to use them. For America, that's a difficult combination to compete with.

11:29We'll be right back after the break.

11:36Support for the show comes from Jevity. Most people who get blood work done hit the same wall. Results come back, you get a nod, and everything looks fine. And then you go about your day. But what happens when something does come up, and you're still left to your own devices? Your blood work flags something, but now you need to figure out what to do about it. Jevity was made to close that gap. It's the first all-in-one Precision Health membership. As a member, you get access to a dedicated three-person care team, including a functional health expert and longevity clinician. They review your blood work and build a plan personalized to you.

12:07Then, whatever you need, such as supplements or prescriptions, are shipped to your door every month. And their full body testing goes beyond blood work. For example, you can get gut microbiome analysis, genetic testing, cancer screening, and more. It's testing plus clinical support plus intervention all in one membership. Here's some data. 93 % of members see meaningful improvements from their first to second round of blood work. Right now, listeners can get 20 % off membership at gogevity.com slash profg. That's G-O-G-E-V-I-T-I dot com slash profg for 20 % off your membership.

12:47Support for the show comes from Rippling. These days, you can chat with AI about almost any business problem. But with Rippling AI, you can actually solve them. That's because Rippling AI is built on your live global workforce data, giving you full visibility into your business and the ability to take action with AI that executes across departments. Say you want to retain top talent. Just ask Rippling AI, who are my top performers this year? And you'll instantly receive a workforce report flagging potential at-risk employees with supporting data like comp ratios, recent performance reviews, and engagement metrics.

13:19But it doesn't stop there. Rippling AI can turn these insights into a proposed reiteration strategy. including a recommended 10 % spot bonus for top performers. All you have to do is tap confirm, and the spot bonus is added to the next month's pay run. So don't settle for AI that's all talk. Head to rippling.ai slash propg and get AI that turns insight into action. That's R-I-P-P-L-I-N-G dot A-I slash propg. Sign up for exclusive access today.

14:14Thank you. like gas or groceries, all while building credit through regular everyday spending. You can also grow your money faster with a savings rate that is nine times the national average. And if you're ever in a pinch, SpotMe lets you overdraft up to$200 fee-free. Join the millions who are already banking fee-free with America's number one choice for banking. Head to chime.com slash profg. That's chime.com slash profg. Sign up now. It only takes a few minutes. Chime is a fintech, not a bank. Banking services and Chime card provided by Chime's bank partners. Qualifying direct deposits require terms and limits apply.

14:50Go to chime.com slash disclosures for details.

15:03Welcome back. For months, the debate over AI has gone back and forth. Is it a bubble? Is it not a bubble? And would we even recognize the warning signs before it burst? This week brought one of the clearest signs yet that the market's AI trance may be starting to break. Oracle is down 35 % this year, and its credit rating now sits just one notch above junk. On Wednesday, the state of Wisconsin demanded the company post$7 billion in collateral before it builds a data center there. Here's Ed on why a utility regulator suddenly wants its money up front. Oracle borrowed$43 billion to build data centers over the past year.

15:51That's against its$67 billion in revenue. But revenue doesn't pay debt down. Free cash flow does. And Oracle's free cash flow is negative. It burned roughly$24 billion over the same period. So taking on record debt while running negative cash flows, well, that is a great recipe for a default. which means that every time anyone agrees to loan Oracle money, they are now taking on an increased level of risk, which means they must now charge Oracle higher interest rates to compensate for that risk. The result? Higher borrowing costs and a$7 billion collateral bill, which will cost Oracle more than$100 million a year.

16:35We have said it before, we will say it again, bubbles aren't built with equity, they are built with debt. And increasingly, the AI build-out is becoming reliant on debt. And speaking of risk, writer Derek Thompson joined Scott Thursday to talk about what he calls the anti-social century and about an essay he wrote on young men called Monks in the Casino. Scott asked whether people have a fixed appetite for risk and whether we've simply changed where that appetite gets spent. I think of it as almost a clean transference of the risk impulse outside of one's home or room toward the risk impulse being entirely cultivated and spent, so to speak, inside of one's room.

17:22And I mean, I'm obsessed with this phenomenon, and there's a bunch of different places to take it. I remember at the end of that essay about monks in the casino, I'm really initiated in this inversion of risk, where I think I was talking to someone about this phenomenon, and they said, you know, the sociologist, the early 20th century sociologist, Max Weber, had this idea that it was Christianity that gave birth to capitalism, and Christian asceticism in particular, this idea that, you know, Puritans wanted to save money that could then be pooled. And then that pooled saved money could be used to be invested in other enterprises.

18:00And that was the foundation of capitalism. But in a way, what we're seeing now is an entire inversion of that principle, because it is capitalism that is giving birth to a kind of wretched asceticism. That is to say, like, the casino economy is turning our young people into monks. And like that is an unbelievably strange and fascinating phenomenon to me. And again, it's not just fascinating. I think it is fundamentally sad. It is fundamentally sad even if you like gambling, even if you think the calci is like the most interesting and wonderful invention in the last 20 years. It's sad to me because the research that we have says so strongly that we tend to overvalue and overdemand solitude when we are given solitude as an opportunity.

18:47But what do we lose when isolation begins to feel safer than participation? Life isn't just Netflix and work. Life is often tragedy. Life is often losing a job, losing a parent, dealing with a mental health crisis. And in those moments, not having a social group to fall back on, that is the real risk. It's interesting. You mentioned two things in your windup. You mentioned socialization and you mentioned vaccines. And there's a way in which those ideas connect in a weird way. You can almost think of socialization as a kind of vaccine against a certain kind of isolating misery. Like, why do you get a vaccine?

19:30Do you get a vaccine in order to help you when you're at your healthiest? No, you get a vaccine so that in the worst case, you don't get COVID, measles, rubella, et cetera. So while I think friendship is delightful in its own intrinsic sense, like I love getting drinks with my friends, I love playing board games with my friends, I love hanging out with my wife, it's also the case that maybe the ultimate payoff of social connection isn't just sharing a martini on a Tuesday night. It's that moment when your parents die, when your sister won't talk to you, when your boss is being terrible to you, when you've lost your job.

20:10These kind of, you can almost call them like moments of social illness. When what you need is the equivalent of a social vaccine to inoculate you against the inevitable misery that every single person is going to feel in this world, in this life. So that to me is the real cost of not exercising the friendship muscle. It's like being with friends is awesome. But when you really, really need people who understand and love you, like when you're in those deep, dark moments, that's where the biggest payoff of friendship can come. And if you haven't been investing in these kind of relationships, then you are entirely on your own at the very moment that you need to be surrounded by love.

20:55Friendship isn't just a source of happiness It's an investment in the people who will be there When happiness becomes harder to find That's The Week I'm George Han We'll see you next Friday

21:35www.adventory.com platform. No messy integrations, no bouncing between tabs, and best of all, no spreadsheets. Stop managing software and start managing your business with one unified system. Try for free today at odoo.com slash vox. That's odoo.com slash vox.