Today is the first time in a long time that the term ‘Gray Rhino’ has been mentioned. It isn’t quite a Black Swan, but it might become one. A Gray Rhino is what is referred to as a highly probable but neglected threat. All this sounds very similar to what has happened to U.S. tech stocks today. Today we look at how DeepSeek sparked a tech selloff in the U.S.
The point here isn’t to go into detail about who DeepSeek is and how long the company has been on the market. What you should know is that DeepSeek is not a new player to the artificial intelligence scene. Its latest product is version 3, not version 1. DeepSeek’s latest innovation, R1, is the third iteration of its AI model.
On the other hand, the important thing to also consider here is when version 3 was released by DeepSeek. Late last year TechCrunch shared details about the latest AI model from DeepSeek. The latest model wasn’t properly released until the 20th of January. The same day that Donald Trump was inaugurated for his second term. Crucially, DeepSeek is a Chinese company. Today, one commentator referred to the company as a “disruptor disrupting the disruptors of AI”. Maybe it is.
What is DeepSeek?
The fuss centres on one of the most powerful ‘open source’ AI models to date. DeepSeek is reported to be better in most areas than its nearest competitors Claude or ChatGPT. Importantly, it costs dramatically less than the billions that OpenAI has spent on its large language models (LLMs). Incidentally the figure is over $7 billion. In the meantime, DeepSeek has spent less than $6 million to develop its open-source artificial intelligence model.
DeepSeek is based in Hangzhou, China and was founded in July 2023 by Liang Wenfeng. It is reported that like Sam Altman of OpenAI, Liang wants to build artificial general intelligence (AGI).
The launch of R1 by DeepSeek last week triggered a huge selloff on Nasdaq today, with Nvidia bearing the brunt of the hit, losing over 17% of its market value in one day. Which is ironic. Because you may ask, surely DeepSeek uses Nvidia’s hardware? And. Amazingly. It does. Perhaps not the latest version. But a very good alternative.
Why DeepSeek is a Wake-Up Call for Silicon Valley
Just last Friday, Meta announced that it would increase its allocation for capital expenditures to $65 billion. At the same time Wall Street only expected the tech giant to allocate only $51.3 billion. Clearly, Meta is aware of DeepSeek, and this didn’t impact on its decision. Much of this allocation is purported to be towards artificial intelligence investment.
The problem is that AI stocks have been driving the latest gains on the S&P500. And, this incident has shown how sensitive Wall Street and, by default, Silicon Valley is to China’s AI ambitions.
In a recent story from the MIT Technology Review, it was suggested that “DeepSeek’s success is even more remarkable given the constraints facing Chinese AI companies in the form of increasing U.S. export controls on cutting-edge chips. But early evidence shows that these measures are not working as intended. Rather than weakening China’s AI capabilities, the sanctions appear to be driving startups like DeepSeek to innovate in ways that prioritize efficiency, resource-pooling, and collaboration.”
This naturally means that the U.S. needs to re-evaluate its AI spend. Will it really need so many data centers? Will the cost of energy drop as the cost of creating models and the related energy needs reduce? This concern has led to a plunge in the price of energy stocks in the U.S. today as well.
Who is Using DeepSeek?
Whilst most large enterprise businesses the ilk of Shell or J.P. Morgan would never dream of using DeepSeek’s AI capabilities. There are some companies that would. Today two small tech stocks began surging after they announced they would use DeepSeek’s LLM.
The companies are very small, neither of them has a market value above $120 million. The point is that they have embraced DeepSeek, and they have seen the benefits. Both in the lower cost, and the better access through an open-source platform. This raises the question: should companies be concerned about collaborating with China?
Can the U.S. Avoid Working with China on Artificial Intelligence Innovation?
The simple answer to this question is probably not. There are several compelling arguments for this. One of the most prominent is how China already has a $1 trillion head start over the U.S. due to a trade surplus.
The other is how today Chinese models have an advantage according to the MIT Technology Review. The Chinese “are able to achieve near equivalent results [to U.S. companies like OpenAI] while using only a small fraction of the compute resources available to the leading Western labs.”
If the AI race boils down to efficiency, then the CCP may have pulled a ‘Trump’ card on Donald Trump. Let’s see how he responds. And how Wall Street will react.
It is important for all of us to appreciate that we need ethics to govern AI. DeepSeek is capturing data every day. It is in the interests of all of us that collaboration between the U.S. and China is in place. Artificial intelligence is not something that you can just have a trade war over.
Author: Andy Samu
See Also:
Should Texas Require New Data Centers to Build Their Own Power Plants? | Disruption Banking
Can Nvidia Keep Innovating? | Disruption Banking
How IBM was using GenAI before it was cool with Michael Conway | Disruption Banking
Market Commentary:
Rahul Bhushan, Managing Director of ARK Invest Europe shared with DisruptionBanking today:
“DeepSeek’s recent breakthroughs serve as a pivotal reminder that the AI opportunity is expanding far beyond the narrow focus on semiconductors and infrastructure. For over a year, we’ve been emphasizing to investors that concentrating too heavily on GPUs risks missing the transformative opportunities emerging in software, platforms, and open-source innovation.
“DeepSeek’s V3 model, which matches the performance of GPT-4 using just 5% of the GPU compute, and its R-1 model, delivered at 1/13th of the cost of GPT o1, underscore an important truth: AI’s future is not just about throwing more GPUs at the problem. These advancements demonstrate how necessity is driving invention, with resource constraints fostering breakthrough efficiencies that are redefining what’s possible in AI development.
“Moreover, the fact that DeepSeek’s innovations are open source cannot be overstated. This move opens the door to widespread adoption and decentralization, a trend that could democratize AI access and accelerate progress far beyond traditional players in the West. It also hints at China’s growing strategic ingenuity in shaping the AI landscape under constrained circumstances. We strongly urge investors to re-evaluate their AI funds and positions. Focusing solely on semiconductors risks being materially underexposed to where the real opportunities are emerging: scalable, efficient AI solutions and the open-source ecosystems enabling them. The paradigm is shifting—AI portfolios need to shift with it.”
