In the audio version in this post and add some commentary.
Picture this: It's the 1980s, and giants like Digital Equipment Corporation (DEC) and Wang Laboratories rule the computing world with their massive, expensive machines. These companies couldn't imagine a future where small, personal computers would matter. Their machines cost hundreds of thousands of dollars, and surely, these toy-like PCs could never compete.
They were wrong. Catastrophically wrong.
One by one, these titans fell. DEC, once a colossus of computing, was sold to Compaq. In 1992, Wang Laboratories went bankrupt. Prime Computer stopped selling new computers that same year. Data General was swallowed by EMC in 1999. Even IBM, who helped create the PC revolution, stumbled badly as they underestimated how quickly things would change.
Today, in 2025, we're watching this same story unfold in artificial intelligence. The disruption has a name: DeepSeek.
The DeepSeek earthquake
Last week, a Chinese company called DeepSeek released something that has Silicon Valley in a panic. They demonstrated an AI model that matches or exceeds the capabilities of OpenAI's latest technology - but at a fraction of the cost. We're not talking about small savings here. DeepSeek claims they built their system for $5.6 million, compared to the half-billion dollars or more that U.S. companies spend on similar systems.
Just like the PC makers of the 1980s who figured out how to build powerful computers cheaply, DeepSeek reimagined how AI systems could work. They use clever mathematical tricks to do more with less - like using 8-bit numbers instead of 32-bit numbers, which dramatically reduces the computing power needed.
The results are stunning. Users around the world are downloading DeepSeek's models and running them on personal computers. Some are processing hundreds of thousands of AI queries for mere pennies. It's like watching the personal computer revolution happen all over again, but at AI speed.
Market tremors and the NVIDIA question
When markets opened after DeepSeek's announcement, NVIDIA - the company that makes the expensive chips used for AI - lost nearly a trillion dollars in market value. Some people ask: "Why does this matter? It's just stock prices going up and down."
But this misses the bigger picture. The trillion-dollar drop reflects a fundamental shift in how we think about AI's future. Just as the PC revolution showed we didn't need million-dollar mainframes to do powerful computing, DeepSeek is showing we might not need massive arrays of expensive chips to do powerful AI.
This isn't just about NVIDIA. The entire AI industry has been building on the assumption that more expensive hardware equals better AI. DeepSeek just proved that assumption wrong.
The pattern of creative destruction
Here's where we see a deeper pattern emerge. Every major technological revolution follows this path: First, something is expensive and exclusive. Then, someone figures out how to make it cheaper and more accessible. The established players panic, claiming the cheaper version can't possibly be as good. But if it is good enough, it changes everything.
We saw it with mainframes giving way to PCs. We saw it with expensive software being replaced by apps. Now we're seeing it with AI.
But here's the twist that many are missing: When technology gets cheaper, we don't use less of it - we use more. Economists call this Jevons Paradox. When personal computers got cheap, we didn't buy fewer computers - suddenly everyone needed one. When cloud storage got cheap, we didn't store less data - we started storing everything.
Tomorrow's AI landscape
This brings us to what happens next. Just as the PC revolution didn't kill computing - it exploded it into something far bigger - DeepSeek's breakthrough won't kill AI. Instead, it will likely transform AI from something that only big tech companies can afford into something that becomes part of everyday life.
Microsoft's CEO Satya Nadella captured this perfectly when he tweeted about Jevons Paradox in response to DeepSeek. As AI becomes more efficient and accessible, its use will skyrocket. The pie isn't shrinking - it's growing dramatically.
This could be an opportunity for the tech giants who can adapt. Amazon's cloud services could offer cheaper AI to millions of customers. Apple's devices could run powerful AI locally. Meta could embed AI throughout its services at a fraction of the current cost.
But for those who can't adapt - who cling to the old, expensive way of doing things - well, just ask the executives of DEC how that worked out in the 1990s.
The lesson is clear: in technology, the future belongs not to those who build the most expensive systems, but to those who figure out how to make powerful technology accessible to everyone. Deep Seek just showed us that future is coming faster than anyone expected.
And just like the PC revolution before it, this will create both winners and losers. The winners will be those who embrace the change and figure out how to use cheaper, more accessible AI to solve real problems. The losers will be those who refuse to believe the world is changing until it's too late.
History doesn't repeat, but it rhymes. And right now, it's rhyming pretty loudly.
It's like when a new student joins your class who's really good at something. They might seem like competition at first, but often they end up raising everyone's game and making the whole class better. That's how progress often works - through challenge and response, push and pull, each side making the other stronger.
What fascinates me most about this story is how it shows that big breakthroughs often come not from doing things bigger and more expensively, but from finding clever new ways to balance different approaches. It's kind of like finding out you don't need an expensive gym membership to get fit - sometimes a creative approach with simpler tools can work just as well or better.
The Relational Paradigm: challenges and opportunities
The DeepSeek breakthrough isn't just a story of disruption—it's a perfect example of how major changes create both challenges AND opportunities. Those who focus solely on the potential downfall of established players like NVIDIA or the threat to American tech dominance are missing half the picture. This is where the relational paradigm comes into play, teaching us that we need to examine both sides and how they interact to truly grasp the situation.
Think back to high school once again. You had the popular kids and the quiet kids. Neither group was inherently "better"—they balanced each other out, each contributing to the school community in their unique ways. The same principle applies here with American and Chinese AI developments. They're not locked in a zero-sum game, but rather pushing each other to improve in different ways.
This story is particularly fascinating because it highlights how quick people are to jump to extremes. Some cry, "This changes everything!" while others lament, "This ruins everything!" The reality, as is often the case, lies somewhere in the middle. It involves both positive and negative aspects that need to be understood in relation to each other.
Yes, DeepSeek's breakthrough might disrupt the current AI industry giants. But it also opens up new possibilities for AI integration in everyday life, potentially spurring innovation we can't yet imagine. It might challenge American tech dominance, but it could also lead to more robust international collaboration and competition, ultimately benefiting global AI development.
The key is to resist the urge to see this development in black and white terms. Instead, we need to embrace the complexity, understanding that the true impact of Deep Seek's innovation will be a tapestry of interrelated effects, some challenging, some opportune, all part of the ever-evolving landscape of technological progress.
As we move forward, the winners in this new AI paradigm won't just be those who can build the cheapest or most powerful systems. They'll be the ones who can navigate this complex relational landscape, understanding how different factors interact and finding opportunities in the balance between disruption and continuity, challenge and opportunity, East and West.
tp
Addendum: An understandable perspective on what is DeepSeek and what it means and doesn’t mean - well, so far …
Share this post