My RISE marketing community is always a few steps ahead of the curve. A month ago, I posted a comment questioning why nobody was paying attention to the startup DeepSeek platform coming out of China.
Well, the world finally caught up, and we woke up Monday morning to a whole new AI world. Some pundits claimed the AI world was upended overnight. One week ago, I read a Goldman Sachs analysis of five possible threats to Nvidia — a Chinese upstart was not one of them! As of this moment, Nvidia has shed $600 billion in market value, the biggest stock slide in history, all because of DeepSeek.
While this development shook the AI bluebloods to their core, I think there are upsides to this calamity. Let’s dig a little deeper today.
DeepSeek: What happened?
If you’re new to this news story, let’s have a quick review.
Last week, a 22-page paper was released by a Chinese A.I. start-up called DeepSeek. It took a few days for researchers to digest the paper’s claims, and the implications of what it described. The company had created a new A.I. model called DeepSeek-R1, built with a modest number of second-rate A.I. chips that matched the performance of leading American A.I. models at a fraction of the cost.
DeepSeek said it had done this by using clever engineering to substitute for raw compute power. Many experts thought China was in a distant second place in the global A.I. race so this came as a shock to the geopolitical community as well.
Because this is China, nobody really knows what to believe or what numbers have been fudged. But as analysts looked under the covers of the open source software, they concluded that the company was on to a radically more efficient process to build and train AI models. Mark Andreesen, a significant investor and Silicon Valley insider, declared the platform “amazing.”
By the end of last week, the DeepSeek mobile app hit the number one spot on Apple’s App Store, and it tipped off a full-blown panic among the tech elite.
Companies like Microsoft, Meta, and Google have already spent tens of billions of dollars building out the infrastructure they thought was needed to build and run next-generation A.I. models. DeepSeek appears to have spent a small fraction of that building R1. We don’t know the exact cost, but it is in the millions, not billions.
DeepSeek’s breakthrough on cost challenges the “bigger is better” narrative that has driven the A.I. arms race in recent years by showing that relatively small models, when trained properly, can match or exceed the performance of much bigger models.
This stock market meltdown hit me personally as I own Nvidia stock and since I was in California when the markets opened, the bloodbath was already well under way by the time I woke up!
Still, I think there are several positives here. I’ll put the emotions of my stock portfolio aside for a moment and look at a glass half-full …
1. We know about it
Most analysts thought China was far behind the U.S. on AI development, at least in part due to the restrictions the U.S. placed on advanced chip sales to China.
But now we know that scrappy Chinese researchers found a way around these restrictions and propelled AI development ahead with innovative programming techniques.
Here is the first positive about this development — we know about it. I’m scratching my head over why China would allow that research paper into the world. This paper revealed key details about how DeepSeek re-imagined AI development.
It seems like cracking the code on a new approach to AI training would be a closely-guarded state secret. But here we are. I consider this revelation a small miracle. Was publishing the paper a mistake? Part of a Chinese plan? Who knows, but it is a long-term positive that the world has this insight now.
We are still very, very early in the AI Era. The fact that this bombshell hit early in the game is good news.
2. Heightened competition
Google CEO Sundar Pichai often says that AI is the most profound technology ever — more important than fire, electricity, or the internet. It will have an immeasurable impact on the human race.
I don’t know about you, but it makes me anxious to have the destiny of the human race in the hands of a few Silicon Valley boys with their toys. American ingenuity will adapt and adopt the Chinese processes quickly. If we can build these models for millions instead of billions, the competitive landscape suddenly opens up dramatically.
The companies with the most money and resources will still dominate, but this new approach allows more startups to participate in the race. By demonstrating that expertise and efficiency can rival brute computational power, DeepSeek has opened the door for smaller players to compete effectively.
3. Land use and energy consumption
A week before the DeepSeek revelation, Meta announced a $60 billion data center in Louisiana that would be nearly as big as Manhattan. When I read that, I wondered if that much energy and land use would be the best decision for our planet.
My guess is that Mark Zuckerberg is re-thinking that announcement. DeepSeek’s approach emphasizes smarter training techniques over raw computational power. This shift could inspire other companies to adopt more sustainable and cost-effective methods for AI development, reducing environmental impact and resource consumption.
When a Chinese company reveals that it can beat your technology at one-tenth of the cost, you’ve got to slam on the brakes. In the same week, OpenAI, announced a $500 billion joint venture with Oracle and SoftBank. Really?
I’m sure all the top companies have organized emergency squads to re-evaulate fundamental strategies and the billions they’ve committed to AI training and new data centers.
I don’t think it will be easy to unwind many of these financial commitments, but we will see smaller-scale data centers as America adjusts to the new framework. Smaller footprints. Less energy consumption. I see that as a positive!
4. Faster and broader innovation
Building effective AI models for dramatically less money opens the door for faster and broader development. Companies can focus on research instead of billions in fundraising.
Many analysts have lamented the sluggish AI efforts in the EU. Could an efficient, low-cost process lower entry barriers and quickly raise development outside Silicon Valley?
Having broader, more global competition will also open the dialogue about the ethical considerations of AI development. Small businesses, entrepreneurs, and even governments could develop and deploy AI-driven products or services that directly challenge incumbents in the space. Increased competition could ensure that the technology serves a broader and more diverse set of stakeholders.
5. About Nvidia
OK. It’s a shock to see the tech stock prices collapse. But I have held steady in my stock ownership.
I have faith in American tech innovation and specifically in Nvidia, which has great leadership and a much broader strategy than just advanced chips. (This is NOT investment advice. My own view.)
I think DeepSeek will soon be banned in the U.S. If the government thought TikTok was a problem, pouring every type of data into a Chinese AI platform is a security concern of enormous magnitude.
So the timing of this was a gift — a fantastic wakeup call for entrepreneurs everywhere.
The current AI race is driven by more than economics. It is propelled by pride and ego. Musk, Altman, and the other tech titans will not let this development stop them. It will inspire them and create a safer, greener, and more competitive landscape at the same time.
Need a keynote speaker? Mark Schaefer is the most trusted voice in marketing. Your conference guests will buzz about his insights long after your event! Mark is the author of some of the world’s bestselling marketing books, a college educator, and an advisor to many of the world’s largest brands. Contact Mark to have him bring a fun, meaningful, and memorable presentation to your company event or conference.
Follow Mark on Twitter, LinkedIn, YouTube, and Instagram
Illustration courtesy MidJourney.
