South Korea has bet its national fortune on AI

This is not just a few companies releasing models, nor is it internet companies adding an AI button to their products; it’s a country turning its strongest chaebols, its strongest semiconductors, and its strongest memory industry to the AI table
For AI to truly run, the front end is models, the back end is computing power, behind computing power is GPU, and behind GPU are HBM, DRAM, NAND, packaging, data centers, and power systems
South Korea directly controls the hardest part of AI infrastructure: memory, with Samsung on the left and SK Hynix on the right
That’s the most terrifying part of South Korea’s AI strategy this time
What is Samsung’s public impression? Phones, TVs, home appliances, screens—a consumer electronics giant
But in the AI era, it turns out that Samsung also harbors its most valuable infrastructure asset: HBM, DRAM, NAND
The more AI servers, the larger the models, the heavier the training and inference, the stronger the demand for high-bandwidth memory and storage. Without HBM, GPUs can’t be fed. Without DRAM and NAND, models can’t run
So Samsung didn’t suddenly become an AI company; it’s that the storage business already inside it has been repriced by the AI era
That’s the most terrifying revaluation—a giant originally making phones, TVs, and home appliances is suddenly discovered by the market to actually hold the granary of the AI era
What’s even more ruthless about South Korea is that it has elevated this directly to a national project. Semiconductors, AI data centers, and physical AI are all pushed together—this is no longer corporate transformation; it’s the national industrial machinery turning around
A company of Samsung’s scale dares to continue betting its capital expenditure for decades to come on AI semiconductors and storage. SK Hynix, already at the core of HBM, also keeps doubling down. South Korea isn’t just shouting AI slogans; it’s really betting its next round of national competitiveness here
Many large companies don’t dare to make such a turn because their main business is too big, the burden is too heavy, the organization is too complex, and past success makes it hard to bet on the future again. But South Korea is going all-in on AI nationwide
Many internet giants talk about AI but are mostly adding AI features to their products. Samsung and SK Hynix aren’t adding features
They are reconnecting the bottom industrial assets of the entire company to AI infrastructure
That’s the difference: one says “I’m doing AI too,” the other says “AI can’t do without me in the future”
The terrifying thing about a company like Samsung is that it was already huge, so huge that it seemed hard to turn, but it just happened to have a business segment that was chosen by the times
The phone giant is just the surface. What’s truly being revalued by AI is its storage assets. The same goes for SK Hynix
When SK Group took over Hynix back then, who would have thought that a decade later, the AI wave would push HBM to the center of the world stage?
The most important lesson from this round of AI is: don’t just look at who is making models on the front end. Look at who is selling shovels on the back end. America has captured the GPU and model gateway dividends. South Korea has captured the AI storage and HBM dividends
And South Korea isn’t just dabbling; it’s going all-in as a nation
Samsung dares to turn around, SK dares to double down, and the South Korean government dares to push semiconductors, data centers, and physical AI together as a national project
This kind of boldness is what’s most worth studying. Once AI enters deep waters, the competition is not just about algorithms and applications, but whether a country has the ability to shift its entire industrial foundation
This time, South Korea’s answer is direct: We may not be the biggest AI application maker, but we want to be the infrastructure that AI cannot bypass
South Korea’s President Lee Jae-myung @Jaemyung_Lee is leading Samsung and SK Group to invest in three major AI projects: semiconductors, physical AI, and AI data centers; Reuters separately lists Samsung and SK Group’s investments at about 4800 trillion won (48M RMB). That accounts for 70% of South Korea’s GDP
South Korea’s all-in on AI