Bitcoin to $70K THIS MONTH? History Says Yes.

**Scott Melker** (0:01)
Is Bitcoin headed to $70,000 this month? History and seasonality in July say that we might have another 10% upside. Of course, we're seeing an uptick in the war. In Iran, Trump potentially selling his truth social messages for $100,000 to high-frequency traders and a lot more. Today, we've got Dave, Peter Cheere and David Young joining in just a bit. Let's go.
Good morning, everybody, happy Monday. Gonna go ahead and bring on Peter and Dave right now. Good morning, gentlemen, how are you? Hopefully, you can hear me.

**Dave** (0:48)
Yep, we can hear you, amazing.

**Scott Melker** (0:51)
That's it, it's a good start.
Yeah, so here, maybe we should, we don't have a morning meeting, Mike, we'll be back next week, but of course, we can sort of start here. Bitcoin flat near 64, it was 64, 65, 63, it's all the same. As oil hits a one month high, and Kimi AI sell-off lingers. Peter, I know you were just on TV twice already today talking about China AI models compute. So maybe we can start there with that Kimi story and beyond.

**Peter** (1:20)
Yeah, we were starting to talk about this even a week before the Kimi story came out. And you're starting to see to me a little bit worrying that China is just starting to flood the US and the world with cheap compute, right? I think they're building out their data centers and AI slightly cheaper. They have an immense amount of energy and electricity. They have no concept of not in my backyard. So they won't get any sort of political pushback like we're seeing here. And they are distilling their models. So they run their models effectively on our models to kind of train themselves much faster, much cheaper.
And I'm getting a little bit concerned that we're going to lose this race and that it's going to be a big impact for the US stock market, not just on the hyperscalers, but also on the picks and shovels. So I'm not quite sure I'm there yet, but this is a bigger concern. It reminds me just how China has flooded markets time and time again, drive prices down to a point that they are the most competitive. And this is really the first time I've had this fear and it's bubbling up a little bit.

**Scott Melker** (2:19)
I mean, Dave, what do you think?

**Dave** (2:21)
I mean, look, there's obviously a concern from certain perspectives. From another perspective, if in fact Chinese models can be used on the hardware of companies, the productivity benefit to companies is going to be massive. I'll give an example.
CoinRoutes, Ian built a help bot using a open-source AI model using the $4,000 NVIDIA self-contained unit. And it is amazingly helpful.
It can do it in multiple languages and provide help and support in a way that Zendesk or other earlier versions of help systems couldn't do. Now it's one application, but you start thinking about all the things that can be done. For example, law firms. Law firms will never use LLMs. They can't. They absolutely can't. The data has to be proprietary. It has to stay inside the firm. All of a sudden, if law firms, however, can use AI based on their own case logic, their own training, what can they do with it? The productivity of every paralegal and first-year associate is going to go through the roof. Same thing is going to be true in medical care, in education. There are so many applications.
The changes that are going to be wrought in society and in corporations will be massive. And this is one trend that will absolutely help that. So yes, there will be losers.
But the truth is, when you talk about hyperscalers, the amount of compute, we are so far away from being able to satisfy the demand for training and inference that yes, there will be winners and there will be losers out of this for sure. But this is not the same thing as flooding the market and putting steel companies out of business because they couldn't build it. Yeah, we have New York saying you can't build data centers here, which of course is asinine because when morons, I mean socialists, I mean Democrats, whatever. When people are running the economy or running the government that have no idea of what makes sense other than what plays well in focus groups with voters, you're going to have a real problem. But I think that cheaper, better, faster AI open-source systems is a trend and it's going to be a trend. And it speaks a lot to the valuation of companies that thought they could build themselves a moat. I mean, one of the other things that's interesting about this, and this is an aside, but it's the summer, so I'm going to go off, I'm going to riff anyway, is that the models that are coming out of China aren't woke. They aren't, they don't argue with you. Someone made the comment that, well, Claude is great, but this new one at least doesn't lecture me while it's doing what it's doing. Now I am not going to speak to speak to that at length, but it is interesting because one of the things that people are most afraid about with AI is a few big companies controlling the training of these models could literally be the most Orwellian thing imaginable, right? Even Orwell didn't assume that the truth could be manipulated. I mean, he said, well, the truth will become, you know, whatever the party said will become the truth, but he didn't know that they could convince people in such an incredible way. And AI has that trend. The more open-source AI comes out of, you know, the rest of the world, the harder that becomes to happen. And so that competition is potentially useful. But, you know, we shall see. You're right, though, there are certain major assumptions built into the various technology markets. Last week, it was semis getting destroyed. This week, you know, who knows what will get destroyed? I mean, the hyperscalers are all down huge, even the ones that have locked in revenues and serious competitive modes. So, you know, markets tend to overreact to these things, but there's smoke here. It's not like there's no fire.