TURMOIL IN AI STOCKS | STOCK MARKET TALK 2026-07-16 20:56

Hosted by @Stocks On Spaces · 2026-07-12 · Tags: KIMI

TLDR

The panel debated whether China’s low-cost Kimi K3 model undermines the valuations and capital-spending assumptions supporting US AI companies and stocks. Participants saw meaningful downside risk to the current AI infrastructure trade, though Speaker 2 argued that cheap intelligence could accelerate a new generation of AI beneficiaries after a possible correction. The conversation later shifted to air fryers, meal preparation, nutrition, workouts, and an upcoming in-person event.

StockTalk argued that digital AI models cannot be blocked from US markets as easily as physical Chinese products.
Kimi K3 was described as competitive with leading US models despite its developer reportedly raising funds at only a $35 billion valuation.
The comparison with trillion-dollar US frontier labs raised concerns about their spending, economics, and valuations.
StockTalk said the rapid Chinese advance materially changes the investment calculus and should not be dismissed like DeepSeek.
The panel acknowledged that Kimi K3’s capabilities, development costs, and possible government subsidies still require confirmation.
Speaker 3 predicted that cheaper intelligence and greater supply would reduce demand and push relevant stock prices lower.
Speaker 2 suggested the data-center and AI CapEx trade could unwind, potentially producing a major market correction.
Speaker 2 remained constructive on longer-term opportunities for companies that successfully apply cheaper intelligence.
The discussion briefly noted war-related uncertainty before moving into consumer appliances, meal services, nutrition, and fitness.
Speaker 2 promoted an upcoming August 3 in-person event in New York City.

Speakers

StockTalk — Led the market discussion by arguing that low-cost Chinese AI models pose a serious competitive and valuation threat to US frontier labs and the broader AI CapEx trade. Later discussed air fryers, Cook Unity, nutrition, cooking, and workouts.
Speaker 2 — Questioned whether Kimi K3’s competitiveness had been proven, explored how an AI infrastructure correction might unfold, and argued that cheap intelligence could create a new class of winners. Also steered the lifestyle tangent and closed by promoting the upcoming event.
Speaker 3 — Offered a supply-and-demand interpretation in which cheaper intelligence lowers demand and stock prices, while also contributing brief jokes and comments about meal pricing.
Speaker 4 — Joined the later discussion about meal preparation, macro tracking, protein, fiber, calorie goals, and workout scheduling.

Notable quotes

“You can't ban a model the same way you can ban a Chinese car being physically sold in the United States.” — StockTalk
“A model was just built that is beating Anthropic's best model on media generation and front-end generation in general.” — StockTalk
“Anthropic is worth a trillion dollars, supposedly.” — StockTalk
“It changes the calculus.” — StockTalk
“cheaper costs, as in supply, lower demand, stock prices go down.” — Speaker 3
“The CapEx trade is the AI trade.” — StockTalk
“I think you just find what the next version of the AI trade is, and you find all those companies that figure out how to use cheap intelligence, and the promise of everything that AI was offered comes quicker.” — Speaker 2
“I think it's foolish to have an answer right now, honestly.” — Speaker 2

Transcript

StockTalk: Chinese dumping of materials or Chinese dumping of solar panels. We try to put protocols in place to prevent the Chinese from competing in our markets because we don't want our companies to die. The issue with AI is that it's not a physical product, right? It's difficult to like ban the usage of Kimi K3 across the US and allied nations. It's going to be pretty impossible to do that. You know, unless you want to introduce some sort of like controls that they have in communist nations where, you know, we start banning internet sites and banning digital access and things like that, which I don't want us to go to that point. But who knows? Maybe at some juncture we will because. You can't ban a model the same way you can ban a Chinese car being physically sold in the United States. It's just not the same thing. So, this is problematic. You know, I know I'm always the bull guy up here that's like trying to come up with, you know, counteract negative arguments, but this is a concern. because this time they're dumping product on us that is digital, and it is built at a fraction of the price that our models are, and it's competitive with our models. And that's a problem. I mean, Grok 4.5 or whatever was a great model. It's the best model that XAI's ever released. I used it for a little bit. It's not as good as Fable 5, it's not as good as GPT Soul, but it's a good model. Elon's selling it at, you know, a third, tokens at a third the price or a fourth the price of what Anthropic is, right? So it's good. It's a good model. But how much do you think it costs XAI to build that? You think a training run cost? Billions, right? How much do you think it cost the Chinese to make Kimi K3? I mean, the company that built it is raising at a 35 billion total valuation. Think about that for perspective for a second. A model was just built that is beating Anthropic's best model on media generation and front-end generation in general. The company that built it is raising money at a $35 billion valuation. Anthropic is worth a trillion dollars, supposedly. Is that not a problem? I like, am I the only one that sees that as a problem? That's a ******* problem. So I don't know. I mean, I'm going to think about this a lot this weekend, but I still haven't figured out my calculus on this, but I was already knowing that this was a possibility, the quantitization of intelligence. I did not expect the Chinese to release this kind of a model as quickly. So I don't know. It changes the calculus. It changes the calculus. It does. You'd have to be coping or LARPing to not think that this does, to think that this doesn't matter. You know, you'd just be like, you're just whistling past the graveyard if you're just ignoring this. This is not Deepseek. Deepseek was not competitive with US models at all.

Speaker 2: Well, how do we know today it is competitive with US models and not just some sort of breakthrough that...

StockTalk: We don't know for sure, but I mean, you can go through Twitter today and there's a lot of posts about Kimi K3 and people are already using it. You don't need a lot of time to benchmark a model.

Speaker 2: I guess we got a lot to think about this weekend. We haven't talked once about the war that's going on actively again right now, which who knows what ends up happening this time next week.

StockTalk: Remember the mark? Remember I said earlier, the markets can only really, truly worry about one thing at a time?

Speaker 3: Yeah, casino's closed tomorrow, by the way. I mean, the cost is closed tomorrow.

StockTalk: Oh, I was like, I thought you meant our casino.

Speaker 2: Yeah, I thought we were done. I was like, what?

StockTalk: Our market's getting our market's getting pretty close. Our market's getting pretty close to looking like Cosby, brother. Yeah, I don't know. It's a problem. It is a problem.

Speaker 2: This this this has been a market of those problems working themselves out and and continuously raging on.

StockTalk: Yeah but I mean okay like look you're not an idiot let's let's play or what is logical back up here logical is back up here am I stupid let's play no no I'm just saying in general I mean that we I am asking I'm asking the whole panel this Sam logical you fast H Market news everyone okay let's play brainstorm here what is the solution to this Any ideas?

Speaker 3: I mean, the solution to.

StockTalk: The cost of intelligence problem. The fact that our frontier labs are worth a trillion dollars making the same models as a $35 billion company across the pond. How do we solve that? I don't know. I don't have an answer. That's why I'm asking you guys. I don't know.

Speaker 3: Yeah, it's a supply and demand, so. cheaper costs, as in supply, lower demand, stock prices go down. That's the only solution.

Speaker 2: But here, so those companies aren't public, but StockTalk, you're saying that the companies who are spending aggressively on data centers, you don't need to spend as much. So really, all these CapEx trades kind of unwind a little bit.

StockTalk: The CapEx trade is the AI trade. Same thing.

Speaker 2: But I just don't see that. It's been the AI trade that's been happening right now. I get it. I get it. When we talk about the market and the AI trade that's been happening, but to think that's actually what the AI trade is going to end up being, I mean, you're in the picks and shovels stage, and there might be a crash between that and getting the gold and actually kind of having a stable gold mine enough. I get it. But I don't know. I think you just find what the next version of the AI trade is, and you find all those companies that figure out how to use cheap intelligence, and the promise of everything that AI was offered comes quicker. And yeah, that probably is a ****** crash. And who knows, maybe it ends up taking a while to come back from, but the Magic 7 probably come back stronger than ever. Nvidium, I don't know that environment. But Yeah, I know. I know you're talking about like, hey, what does the next couple of months look like after that environment? And I'm saying more like, I don't know, but. Yeah, I'm not. I'm not some.

StockTalk: I'm not some Oracle. I'm just talking to you guys about what I'm seeing today. Like, I'm just speaking plainly as it is. Like, today there is a $35 billion Chinese company that released a model that is competitive with Claude and OpenAI who are trillion dollar companies.

Speaker 2: I think it's foolish to have an answer right now, honestly.

StockTalk: Yeah, I don't have an answer.

Speaker 2: Yeah, consistently China has proven that.

StockTalk: They're full of ****.

Speaker 2: Yeah, give or take. I mean, yes, like just all the things end up coming up or it gets taken in a different way. And I'm sure there might be some copium or whatever, but like.

StockTalk: Yeah, who knows? Maybe the Chinese government subsidied this for $200 billion, but then why is the company raising funds at 35 billion?

Speaker 2: We shall see. We shall see. I imagine by the time we're talking on Monday, there is more just news confirmation, just different weird headlines coming out in this direction. So I have a feeling the next 72 hours is going to be chock full of the most random news stories that you could imagine. Did you know that Ninja or Shark Ninja? Wall Street engine is not hidden here. They released a microwave that's also an air fryer. That's innovation.

Speaker 3: He's here. He's listening.

Speaker 2: Oh, I don't see him. Oh, no, I do. Yeah, he posted it too. I posted when the Ninja Creamy came out a while ago, and it did really well. So for some reason, people really like seeing the Ninja product. Some people thought it was ads. I would take a Ninja air fryer thingy, but stock talk, is that a good invention? What an invention? I know Josh loves these questions. Shark Ninja released a... Microwave that air fryers the food at the same time.

StockTalk: Well, what's the point of that? Just air fry it.

Speaker 2: I think that's a fair point. I don't know, maybe it's, I think it's quicker.

StockTalk: Oh, like it like heats it up and air fries it at the same time.

Speaker 2: Yeah, so it's like similar times as the microwave, but then it comes out crispy is I think the concept.

StockTalk: Interesting. I mean, I have a Prevel air fryer, which is like the best in class, and I mean, I think it is the best air fryer.

Speaker 2: If you have a girl coming over and you need it and she's expecting a meal to be cooked.

StockTalk: I'm air frying anything in her meal.

Speaker 2: I'm not saying, what are you gonna be air fried? What are you making? Or, you know, are you just, is she-?

StockTalk: Now we're just done, now we're a chef show or what? I mean, I like to cook, but I mean, I can make anything. I'm a pretty good cook.

Speaker 2: You're hiring a chef? Is that more impressive?

StockTalk: I used to have a chef, but that would do meal prep for me, but I just do cook unity now. It's easier.

Speaker 2: What's Cook Unity?

StockTalk: It's a meal prep service that uses real chefs. I think it's pretty great. I think the food is fantastic.

Speaker 4: What's it called?

StockTalk: Cook Unity.

Speaker 4: Okay, I'll check it out.

StockTalk: It's like it has real chefs, like actual, like celebrity chefs that design the meals. And I mean, not everything is meal preppy. Like some of the macros are on some of the meals are *** but they have like, good 1 to 10 meals, 1 to 10 people that don't know, 1 to 10 protein to calorie ratio stuff. That's good. Like I'm eating one right now and it's pretty fantastic. But yeah, the food like tastes really good and is healthy and they're like complete meals and they list all the macros on their site for every meal and they rotate every week and it's pretty affordable compared to other meal prep services. Like, I mean, maybe it's not as cheap as some other ones, but it's like, 10 to 12 bucks per meal and they're really good.

Speaker 3: It's actually not that bad price, to be honest.

StockTalk: It's not, bro. And they send it to you fresh every week. It's good. I mean, I think it's every most of the meals have like a sauce component. They have a ton of meat. They're like a minimum of six ounces of meat in every portion. They're all under like. A thousand calories for each meal. Most of them, like the one I'm eating right now is 490 calories, 43 grams of protein, 18 grams of fat, 40 carbs, pretty good macros. And it's delicious. Like way better. Like I've tried Factor and I've tried like the other meal prep services. They've all been **** in my opinion, except Cook Unity, I think is excellent. I do like 10 meals a week and I eat out on the weekends. So yeah, I was just ordering too much Uber Eats. So I decided to do, I only got this like a month and a half, two months ago, but I've been ordering Uber Eats like crazy. So I was like, you know what? I'm trying to be a little healthier. So I got this, but yeah, it's good. I think it's good. I vouched for 'em. I'm not sponsored by them for anything, obviously.

Speaker 3: I was gonna say we're in the film.

StockTalk: No, no, no, no, yeah. I'm not sponsored by it. But hey, if logical, if you do want to sign up, actually, I think I do have an affiliate link through my membership with it.

Speaker 4: But my wife and I have been pretty good on the meal prepping stuff. I've been counting macros very well.

StockTalk: I just hate meal prepping. Like I like cooking. I hate meal prepping.

Speaker 4: No, no, I agree. What I do is I meal prep very little bits, like not like my entire week's worth of meals. Like I'll try to do like, I don't know, like two pounds of chicken and like some lentils and sweet potatoes, and I'll just have those in there for whenever I don't feel like cooking, like in the middle of a workday or something. But most of my meals are fresh or like, you know, I'll even for my late, like an early lunch, I'll have like eggs, avocado, sourdough, toast or something, and that's obviously fresh. So then like maybe later I'll have like an early dinner and do like chicken and sweet potato or something like that, which will be pre-cooked. But, you know, just depends or like, yeah, just depends. Or then I'll cook something fresh, whatever. But, you know, I'm pretty good at that. Like macro counting and like, you know, Greek yogurt bowls with raspberries and stuff or, you know, protein shakes with frozen blueberries, blah, blah, blah. You get the idea. But yeah, I mean, as long as you're trying to do like, I mean, I try to hit like 160 grams of protein a day and minimum, which is kind of tough. Try to get close to like 30 grams of fiber. Like those are the two main things and like try to stay below 2000 calories. That's yeah, it's very, very good.

Speaker 2: Sorry. All right. I got a call here in a couple of minutes. I appreciate you all for hanging out with us today. Should definitely make sure you are following the speakers. They're all fantastic. They will improve your experience on this app. Again, our first in-person event is like 2 1/2 weeks away, August 3rd, New York City. Talked about it a bunch in here, but if you guys are already coming, we're sending out some emails and stuff like that. What do you want in your name badges? We're doing breakout Q&As where you can pick who you want to be able to ask questions to. Stock Talk will be there. So sending those out. If you guys haven't received those emails, just send the stock market news accounts that Evan account up here at DM and I will make sure I get back to you on that. Getting very excited. I'm looking forward to it. Yeah, it's crazy. That's literally like 2 weeks away. 2 1/2 weeks away a wild. Is there anything you gents want to leave the people with? I got a 530 cut off so you guys got a 2 minutes here if you want to say anything. Any more food recommendations? Stock talk. What's the workout today?

StockTalk: Chest. Every time you ask me, it's chess. Literally, every time I think the last five times you've asked me, it's.

Speaker 2: Only worked out chess.

StockTalk: No, I mean, you just don't ask me every day.

Speaker 2: No, I'm just joking.

StockTalk: It's usually like end of week when you ask me, so that's usually when my chess is.

Speaker 4: It's so backwards because everyone starts with chess on like Monday or Tuesday.

StockTalk: I don't like doing that. I like finishing the week with it.

Speaker 4: It's just tough because I feel like it affects like shoulder workouts after and.

StockTalk: I mean, yeah, but it's like, you're in a perpetual cycle anyway, right? It's like, it's not like I was born on Monday. Yeah, my first workout ever was Monday.

Speaker 4: Yeah, time is not real, so that makes sense.

Speaker 2: Nixon five.

StockTalk: The one thing you were right about this year?

Speaker 2: At one point in my life, it was bound to happen. All right, I appreciate you all. I hope you have a good one. I will catch you all later. Like I said, in-person event, I'm very much excited for this. You're getting close. For those people who already signed up, like I said, we got stuff going on, so make sure you send me a DM. I want to make sure that you guys get all the extra stuff. If you have no idea what I'm talking about, I guess you could send me a DM too and I could send you the details, but I'm sure you've heard us talk about it. Have a great one, everyone. We'll catch you all later. Peace.