Sunday, January 26th, 2025
Laramie, Wyoming
By Dan Denning
[Editorial Note: In yesterday’s Private Briefing, hosted by Tom Dyson, we forgot to note you can learn more about Chris Weber’s work, including his latest call on silver, at his website here. There is no financial arrangement between BPR and Chris. We love his work and value his investment insights.
Hold the phone!
Drop the AI!
Is it possible that an obscure Chinese company which no one had ever heard of at the begging of the week is about to wipe out $20 trillion in market value of US chip makers, AI companies, and the companies that have invested heavily in both technologies?
Were the Magnificent Seven t-boned by a Chinese Mac truck out of nowhere on the Information Superhighway?
One question at a time in this week’s review. First off, yes, it IS possible. The nature of technological change is for the new to displace the old.
The term ‘creative destruction’ was first used by Austrian economist Joseph Schumpeter in 1942. He described it as, the ‘process of industrial mutation that incessantly revolutionizes the economic structure from within, incessantly destroying the old one, incessantly creating a new one.’
It’s hard to stay on top for long, especially in technology, where the rate of change and innovation is always faster. And even though Schumpeter was talking about ‘industrial’ mutation, his observation applies to the ‘information’ economy as well.
Chip makers, data centers, software, hardware, energy—ALL of it is subject to being blindsided by game-changing innovation, no matter how deep and wide the competitive moat appears to be for any given company. Just ask Cisco Systems…
Cisco built networking equipment in the late 1990s and early 2000s. That equipment was vital to the infrastructure of the Internet. Whatever the Internet was going to be or become, it would need Cisco’s hardware to get there.
The result, for shareholders, was that Cisco, for a brief time, became the most valuable company in the world. It peaked, then crashed, giving up almost all of the gains it racked up between 1998 and March of 2000.
Nvidia is today’s Cisco, at least in terms of share-price performance and producing the technology/hardware critical to THIS information revolution (its chips). But last week a Chinese firm named DeepSeek released not one but TWO new AI models that have disrupted the entire AI industry.
The models, called DeepSeek V3 and DeepSeek 1, are supposed to be better and cheaper than competitors like Claude 3.5, GPT 4o, and Open AI’s OI. Importantly, DeepSeek is also ‘Open Source’, That means you don’t have to pay the company a license fee to use it and the source code is, well, free.
But what do ‘better’ and ‘cheaper’ mean?
What does this mean to company’s like Nvidia whose share price is based on the belief that the world need more of its chips? And what does it mean to the companies that have spent billions of dollars on Nvidia’s chips (Amazon, Microsoft, Meta, Google etc.)?
Have those companies overspent on chips they didn’t need?
And are they now overvalued based on business models that have been severely undercut by China’s new, emerging, open source start up?
That’s the $18 trillion question!
When Chat GPT was first introduced to the world in early October of 2022, the total market cap of US stocks was $36.46 trillion. Today, it’s $52.88. That’s an increase of about $16.5 trillion in market value which you can chalk up to the AI hype, and lately, the hype around data centers AND the energy they will need (nuclear, natural gas, or anything that generates reliable electricity).
Is ALL of that at risk now that this little Chinese company proves you don’t need as many chips to build an AI that is better and faster than the models American companies have spent billions developing? That’s what we’re about to find out!
Paying subscribers can expect to hear more on this later in the week. The technical questions are fascinating but beyond the scope of this free weekly note. We’ll also look at whether this brings us suddenly closer to Artificial General Intelligence (AGI). The relevant investment questions are simple but urgent and important:
Will demand for Nvidia’s chips now be much LOWER than forecast based on DeepSeek’s breakthrough?
Will the margins and models of AI companies compress or collapse based on DeepSeek’s competitive (or superior) performance at much lower prices?
Is DeepSeek really a small, low-cost start up that started off as a quant trading shop and pivoted to AI due to Chinese restrictions on trading and American restrictions on chip exports? Could it be an CCP/PLA operation to subvert a key American industry?
How long will it take the stock market to make sense of this and correctly reprice the AI companies (and the whole AI Information Complex) for the new reality?
What is that price?
The Mag Seven stocks have a combined market value of just under $18 trillion as of Friday’s close. That’s bigger than the combined GDPs of France, the UK, Italy, Spain, Germany, Belgium, and the Netherlands.
But if the huge tech tailwind powering US mega-cap tech stocks higher turns into a Chinese headwind, it will slow down the entire market, not just the biggest companies. That’s what’s at stake here.
Stay tuned for more later this week.
Until then,
Dan
P.S. In the Terminator movie franchise, Sarah Conner (mother to future commander of the human resistance John Conner) packs up her red 1983 Jeep CJ-7 Renegade and heads to Mexico to prepare for the coming war between humanity and the artificial Super Intelligence known as Sky Net. Why Mexico?
I asked my friend Jenn Stevens, the publisher of International Living this question recently. I asked HER because readers were asking ME if Mexico might be better retirement destination for Americans than say, Montana or Mississipppi. Is it?
Jenn said interested American retirees should check out the following free articles on IL’s website. This will give you an overview to the whole idea of ‘retiring overseas,’ introducing to other regions you might consider (including ‘up-and-coming’ ones that are still off the beaten path, and specific insights about Mexico).
It’s true that a lot of native Mexicans may, themselves, be returning to Mexico now and the next few months. But it’s a big country. And there is plenty of room and opportunity for Americans too.
BPR readers may know that our founder, Bill Bonner, founded International Living in 1978. Inflation, rising crime, and a crashing economy were all reasons Americans back then may have wanted a second home (or a retirement home) overseas. Is history repeating?
I’ll ask that to Bill directly when I see him in Baltimore the first week of February. I’ll also be meeting privately with some key people from International Living. Hopefully they’ll agree to talk to me on camera for a Private Briefing. Stay tuned.
Learn more about all the work being done at International Living here.
The Buenos Aires of the North
Editor’s Note: Below is a note from Diego Samper, the Bonner Private Wine Partnership’s intrepid wine explorer. Diego’s travels between Paris and Buenos Aires give him a unique perspective on Argenti…
Oh, the drama!
I went shopping and forgot to mention something. So I've added it to my original line of thinking and edited the entire post.
A little-known Chinese startup shows up, waves its DeepSeek wand, and suddenly, the entire AI-industrial complex starts looking over its shoulder. Nvidia, OpenAI, and the rest of the Magnificent Seven are suddenly cast as vulnerable giants, and the stock market is teetering on the edge of a nervous breakdown. It’s all very exciting, but before we go full end-of-the-world-as-we-know-it, let’s take a step back.
Paul Zane Pilzer’s 1990 book "Unlimited Wealth" offers us a handy lens for moments like this. Pilzer believed that technology doesn’t just shift the game—it expands the board. He famously said, “Technology increases the efficiency of how we use resources, creating wealth faster than it consumes it.” If DeepSeek’s new AI models truly are groundbreaking—better, cheaper, and open-source—then what we’re seeing isn’t destruction but an acceleration of wealth creation.
Now, here’s the part where it gets really interesting: If DeepSeek’s coding is as efficient as it’s rumored to be, it won’t just threaten the current AI leaders—it will empower them. Why? Because open-source breakthroughs don’t belong to one company; they belong to everyone. Imagine Nvidia, OpenAI, or Google adopting this technology for their own massive systems.
If 10,000 GPUs can now do the work of 100,000+, as suggested, then consider the implications for the “big boys” who already have 100,000 GPUs or more. Suddenly, their computational power doesn’t just double—it multiplies exponentially. They’re no longer constrained by the inefficiencies of the past, and the scale of what they can achieve becomes almost unfathomable.
This isn’t just about making existing systems cheaper or faster. It’s about unleashing a whole new level of AI capabilities. With this kind of efficiency, those same companies could tackle projects and problems previously deemed too resource-intensive or expensive. Think quantum-level breakthroughs in medicine, energy, and who-knows-what else. As Pilzer might say, “When resources are measured in terms of technology, they become essentially unlimited.”
So yes, the immediate market reaction might be a little jittery—change always rattles people. But let’s not mistake this moment for the end of the AI revolution. If anything, it’s a catalyst. Nvidia and its peers aren’t going to fold—they’re going to adapt, integrate, and ultimately thrive. And the rest of us? We get to benefit from the innovations that emerge, both from DeepSeek and from the giants scrambling to stay ahead.
In short, this isn’t the sky falling—it’s the horizon expanding.
Yours in optimism (with a side of curiosity),
Luke
P.S. Imagine what happens when everyone can do more with less. It’s not just a win for the big players—it’s a win for innovation itself. Let’s not panic; let’s marvel.
AI, AGI, whatever….all part of what has to play out, be in place, for the emergence of the Anti-Christ.