AI Stocks Are Crashing… So Why Is Demand SOLD OUT Until 2029?

Vincent (0:00)

We're just 5% into American households paying for LLMs, right? Not even talking about using the Atlantic solutions that are out there.

So no matter how fast we build the data centers, we will always be too slow. This will ultimately prevent this kind of burst of a bubble if there's any.

SPEAKER_2 (0:22)

What's up, everybody? It's LGD Uset here, and welcome to Milk Road Stocks, the daily market show that gets bored the second the market starts chopping, and somehow always comes back just in time by the local top. Today is October 14th, 2026, recording on the 13th. What happens when Elon Musk replies to your tweet? Apparently, memory stocks sell off. Melvin posted that the drop in memory names made zero sense after reports that a Chinese company figured out a way denser way to make DRAM. Elon replied to his tweet and disagreed, pointing to how fast China scaled up batteries and solar. Melvin doubled down and he thinks his post might have played a role in the sell-off.

Who's right? On today's episode, our top AI analysts break down whether China is a real threat to the memory trade or a problem that's still years away. They also ask whether Anthropping and OpenAI can actually hit Dylan Patel's call of $600-700 billion in revenue by the end of next year. That's a 4-5X jump. Plus, they don't agree on whether this is the dip to buy or whether you wait for midterms to clear. Then they head into after hours, their 15-minute portfolio strategy session to show exactly how they're investing around all this. You could check out that extra session for just a dollar in Milk Road PRO at the link below. Otherwise, enjoy the episode.

Kyle Reidhead (1:29)

Gentlemen, we're back for another weekly roll up. I feel like this week we have a lot less news. The last couple of weeks, I've been like, man, this is crazy, jam-packed week. This week, maybe not as much, I would say.

Let's go around the horn. We'll talk about everyone's thoughts on the markets, and then we'll jump into it. Vincent, you're shaking your head. You think there's a lot going on? What's on your mind?

Vincent (1:54)

Well, just OpenAI apparently missing $20 billion in revenue, Meta launching is one of its best models, Lumentum CEO coming out saying we're sold out until 2029, memory news, blah, blah, blah. So I actually think there is. But yeah, a typical week in AI, I would say. If I were to summarize it, there were a lot of bullish sides or news from the models, because A, I think the OpenAI news with those 20 billion missing is complete fad. It's just a miscounting issue of the Financial Times, actually. Then you have Metamuse.

You all said Dylan Patel coming out saying he sees 600-700 billion in revenue by end of 2027 from Frontier AI. So OpenAI.

Kyle Reidhead (2:42)

I don't think it was revenue. I think it was AR, wasn't it?

Vincent (2:44)

Yes, sorry.

That would be crazy. We can talk about that. So I think the underlying news were really bullish, even though the market performance the last couple of days was not so good.

Kyle Reidhead (2:58)

Yeah. Nolan?

Melvin (3:00)

Yeah. I have a long take on this, what's happening in the market. But I think there's a lot of shit that's been happening in the market. I genuinely don't know where to even start because, as Vincent mentioned, we had the Financial Times report last week that created the FUD and started with the sell-off. Then yesterday, we had another sell-off, which was this time in memory, after there's new surroundings, China's CXMT and their new DRAM technology, that absolutely crushed our markets, all the high beta stocks, along with Korean markets, especially like later yesterday.

Then that spilled over to today as well into the US markets and all the high beta stocks is once again getting crushed. Funny enough, I think I played a role in that sell-off yesterday, which I'll explain why. There's a lot to unpack there as well. On top of that, yesterday we had-

Kyle Reidhead (4:03)

Hold on. What Melvin's talking about is he did this whole long tweet. It's a really good tweet about the sell-off in memory stocks, and Elon Musk replied to his tweet. We'll bring it up and show it in a little bit. Elon disagreed with Melvin.

Melvin doubled down and was like, no, no, Elon's wrong, which is hilarious.

Vincent (4:22)

No, no, that's not what happened.

Kyle Reidhead (4:24)

I think the fact that Elon disagreed was one thing that led the market to go, oh, you know what, maybe there is a supply issue here with memory. But anyway, finish up, Melvin. We'll get to memory a little bit in the show.