“DeepSeek-R1 is AI’s Sputnik moment”
– Marc Andreessen, co-founder of Netscape
arvy’s teaser: DeepSeek has emerged as AI’s Sputnik moment. Born in China, this disruptive LLM rivals America’s best at a fraction of the cost. Big tech trembles, innovation accelerates, and market uncertainty deepens.
Sputnik moment.
The term refers to a period of public fear and anxiety in Western nations.
This fear was sparked by the Soviet Union’s launch of Sputnik 1 in 1957, the world’s first artificial satellite. The successful launch of Sputnik, powered by the R-7 rocket, also revealed that the Soviet Union had developed the capability to strike US territory with nuclear-armed intercontinental missiles. This realization sent shockwaves through the West and marked a turning point in the Cold War. It led to the establishment of NASA and intensified the Space Race between the USA and the USSR.
Today, whenever the US faces a reality check or a moment of awakening caused by a competitor from outside its borders, it is often described as a “Sputnik moment.”
And now, the US has just experienced another one!
It is called DeepSeek-R1.
Chart 1: Top Charts, US App Store over the Weekend, 25/26th of January 2025

Source: US App Store
Enter the dragon
Alright then, let’s get a coffee first and see what all the fuss is about.
In short, a hedge fund in China has launched its most novel and proprietary form of ChatGPT – DeepSeek-R1. And apparently it has done so in a very short time and at very little cost (10% vs. other well-known Large Language Models, LLM), and most importantly, it works almost as powerfully as anything coming out of the US.
This means that Chinese AI is now so close in quality to its American competitors that the head of OpenAI (ChatGPT), Sam Altman, felt obliged to explain the small gap.
“It’s (relatively) easy to copy something that you know works. It’s extremely difficult to do something new, risky and difficult if you don’t know if it will work.”
But why is the market reacting so strongly to this development?
Especially with the hyped AI stocks?
Because the dragon has entered.
Despite the US government’s efforts to slow down China’s AI industry, Chinese companies have reduced the technological lead of their American counterparts to just a few weeks.
China’s AI industry initially appeared to be second-rate. This may be partly since it had to contend with American sanctions. In 2022, America banned the export of advanced chips – remember, these models train with Graphic Processor Units, GPUs – to China. Nvidia, the leading chip manufacturer of GPUs, had to develop special downgrades of its products for the Chinese market.
Here’s where the problem comes in. They somehow managed to succeed anyway! And they offer it for a fraction of the price that others charge.
ChatGPT starts at $20/month, while DeepSeek is much more affordable.
Starting at just $0.50/month.
Chart 2: DeepSeek-R1 upsets AI Market with low prices

Source: Statista, DocsBot
The Market Hates Uncertainty
The news about LLM DeepSeek is causing uncertainty.
Why?
Let me give you an example and exaggerate a little.
It’s like you buy the most expensive house in the neighborhood for $1 million and a guy named “DeepSeek” builds a similar house next door for $100,000 the next month. And as if that wouldn’t be all, your annoying neighbor releases the entire protocol – called open source – of how he did it to the public for free. That means anyone can use the code and build whatever they want on top of it.
How would you feel after all this?
The emergence of something even remotely possible like this could therefore now pose a significant threat to leading AI companies such as Meta Platforms, Microsoft, Amazon and Alphabet as the open-source approach erodes competitive advantage.
In addition, the efficiency gains could have a dramatic impact on hardware companies such as Nvidia (chart 3) as demand for chips falls as fewer resources are needed to train these models.
Of course, it’s still early days, but given the market concentration – it involves the largest companies in the world, driven by the same AI narrative, high expectations and expensive valuations, this creates uncertainty.
And the market hates uncertainty.
And that’s not all.
Chart 3: Nvidia’s $560bn DeepSeek rout is largest in market history

Source: Bloomberg
The only constant is change
It is more evident than ever.
While the giants, led by Magnificent 7 (Apple, Microsoft, Alphabet, Amazon, Meta, NVIDIA and Tesla), have enjoyed significant tailwinds, the rise of DeepSeek in the blink of an eye has shown that things can look very different overnight.
In the fast-moving technology industry, where innovation can come from anyone with coding skills and little capital, the risk of disruption is high. Even the biggest players are not safe and live by the mantra: “It’s always Day-1”.
So not even the most advanced players can rest on their laurels.
And history has shown that they never could.
Below are the world’s ten largest stocks by market capitalization at the start of each decade (chart 4). 1980/1990/2000/2010 and now 2022. The market always tells a story or follows a narrative. This is also the case with the AI hype now.
We are therefore convinced that pure indexing of the portfolio will lead to mediocre returns and that after a flood of money over many years and historically high concentration, dispersion will increase, and the alpha opportunity will strengthen. This is also reflected in the best periods for stock picking in the 1970s and after the Dotcom crash.
After all, the only constant is change.
Chart 4: The world’s ten largest stocks by market capitalization at the start of each decade

Source: GaveKal, arvy
Mit deinen Freunden teilen. Deine Unterstützung hilft uns unendlich.
Newsletter Disclosures