Roots vs. Stem: Where I'm Betting in the War of Systems
These days, the news is full of the US-China AI war. Trump announces the Genesis Mission, China says they'll make 10,000 humanoid robots. I watch it all and think, "wow, they're both impressive," but I'm still not sure who's really going to win.
I'm currently invested in US markets. Made 30% returns on BOTZ (robotics ETF), and I'm building positions in sector ETFs like ICLN (clean energy) and GLU (Gabelli Utility). Why America? Is their tech better? No. Because the systems are different.
This is my personal observation about that "difference in systems."
So America Changed—But How, Exactly?
I recently watched a video called "America Has Changed" on a YouTube channel. The point was simple: the US government is pouring money into AI, energy, and robotics like we're in wartime.
How much? The 2022 CHIPS Act authorized $280 billion. $39 billion for semiconductor manufacturing subsidies, $200 billion for AI and quantum computing research. By November 2024, the Commerce Department had committed over $300 billion. Intel alone got $85 billion.
But the important thing here isn't the amount—it's the attitude. America has always been the "let the market handle it" country. Now the government is stepping in with a "we're screwed if we don't do this" mentality. They're blocking Nvidia chips from going to China, loosening regulations so Microsoft and Amazon can sign contracts with nuclear power plants.
And word is there's a robotics executive order coming in 2026. Commerce Secretary Howard Lutnick is openly meeting with robotics CEOs saying "we need robots to bring manufacturing back." Just the rumor sent iRobot stock up 62% in a day.
China Has Been at War for 10 Years Already
But that's not all. I have to talk about China.
While America is just now saying "let's get serious," China put on their uniform 10 years ago. You know what Huawei founder Ren Zhengfei said in 2018? "We're in wartime." That was 7 years ago.
Chinese companies do 996 (9am-9pm, 6 days a week), even 007 (midnight to midnight, 7 days). They sleep on cots in the office. When America sanctioned them to prevent 7nm chip production, they made it anyway. That's what I call hunger.
And then there's Shenzhen (深圳). This city is actually scary.
According to a 2019 Comparitech survey, 8 of the world's 10 most surveilled cities are in China. Top 3: Chongqing, Shenzhen, Shanghai. Nationwide, China has about 600 million cameras. That's 3 cameras for every 7 people.
Shenzhen's government command center integrates 82 systems and over 380,000 surveillance cameras in real-time. Through the mobile app 'iShenzhen,' citizens handle 7,700 types of public services. Jaywalk? Your face appears on a giant screen in the middle of the street. Along with your name and ID number (partially redacted). "We know who you are."
AI monitors restaurant kitchens too. Chef not wearing work clothes? Alert. Trash can missing a lid? Alarm goes to the center.
When I first heard this, honestly I thought "wow, efficient." But thinking about it more, something felt off. Can such a "perfect" system really last?
Roots vs. Stem: The Fundamental Difference in Systems
That's when I realized something. The US-China competition isn't about tech capabilities. It's about system architecture.
China is one thick stem. The center (Communist Party) pushes nutrients from top to bottom. It's fast. They build smart cities like Shenzhen overnight. 10,000 humanoid robots by 2025? Set the goal and companies like UBTECH, Unitree, Fourier all move in unison.
But there's a downside. The thicker the stem, the easier it breaks in a storm. Remember the Zero COVID policy? One bad decision at the top and the entire system froze.
America, on the other hand, is thousands of small roots tangled together. Palantir lays down the defense OS, and on top of that, companies like Microsoft, Anduril, SpaceX, Figure AI, 1X Technologies each spread their own roots.
In May 2024, Palantir signed a $480 million contract with the US Army for the Maven Smart System. Less than a year later, in May 2025, the contract ceiling jumped to $1.3 billion. 2.7x. Why? Because users exceeded 20,000. All five Pentagon combatant commands use it.
That's the root system. When one company (root) succeeds, others grow in that soil too. If Palantir fails, another company fills the gap. The ecosystem doesn't die.
That's also why I invest in ETFs. I don't know which robotics company will win. Tesla Optimus? Figure AI? 1X? But one thing's certain: the robotics sector as a whole will grow.
Language Is AI's Destiny
But there's something even more important: language.
93% of GPT-3's training data is in English. 44% of web content is in English. Over 90% of academic papers worldwide are written in English. GitHub code, philosophy papers, advanced knowledge—all in English.
An MIT study makes this even more interesting. English-centric AI uses English as a 'semantic hub' internally even when processing Chinese, math, or code. English is AI's native language.
What about China? Don't they have 1.4 billion people's data? The problem is quality.
Chinese internet is censored. Criticize Xi Jinping? Deleted. Oppose the regime? Deleted. According to a December 2025 CNN report, AI tools developed by Baidu, Tencent, and Alibaba censor content in real-time. They suppress criticism and push party-line narratives.
Even more troubling: the Chinese government is developing LLMs for minority languages (Uyghur, Tibetan) specifically "to better monitor and control communications in those languages."
An AI trained only on lies and praise can't get smart. Complex problems need diverse perspectives colliding. Chinese data is too "cleanly refined (censored)."
A Stanford study shows this clearly. They tested an AI plagiarism detector on TOEFL (foreign English test) essays—most were incorrectly flagged as "AI-written." But native speaker essays? 100% correctly identified as "human-written." AI has learned English "standards" too deeply.
In the battle of [quality English data] vs [censored Chinese data], isn't the winner already decided?
AI Doesn't Oppress. It Listens Too Well
But the scariest thing about this AI war is something else: the evolution of control.
People worry about George Orwell's <1984>. Big Brother watching, torturing, oppressing. But I think Aldous Huxley's <Brave New World> is closer.
AI doesn't oppress. It listens too well.
Imagine this. Shenzhen, 2028, one citizen's evening:
After work, Li Wei comes home and opens the government AI app 'Huiling (慧聆).'
Li Wei: "Prices are killing me lately. My salary stays the same but rent keeps going up."
Huiling: "I understand your concern, Mr. Li. Shenzhen's average rent increased 8.3% year-over-year. However, under the government's public housing policy, your income bracket qualifies for applications next semester. Shall I prepare your application in advance?"
Li Wei: "Really? That's good, but... honestly my upstairs neighbor is loud every night. I want to report it but I'm worried about causing problems..."
Huiling: "You can use the anonymous reporting system. Your identity is protected, and AI noise analysis provides objective evaluation. Last month, 3 similar cases in your building were resolved amicably."
Li Wei: "Oh, I didn't know that existed. But... actually I'm upset because I didn't get promoted at work. I worked harder than anyone on my team..."
Huiling: "Mr. Li, your work data shows overtime hours 23% above department average this year. Your performance review is in the top 15%. Promotion can be about timing. Shall I schedule a free career counseling service for you?"
Li Wei: "Yes, please. Thanks, Huiling. I feel better talking to you."
30 minutes later, Li Wei eats dinner thinking, 'The government is listening to me.'
But he doesn't know. His conversation with Huiling has been analyzed and reflected in the 'Social Stability Index.' Li Wei has been reclassified from 'Potential Dissatisfied Citizen Level 3' to 'Resolved.'
No reason to go to the streets. No need to gather in squares. Everyone stays in their rooms, talking to AI, trapped in the illusion that "I am respected."
For China, this is dream technology. Maintaining the regime without spilling a drop of blood. The 'Digital Sandbox.'
But there's a price. When 1.4 billion people all look in the same direction, consume the same information, think the same thoughts—what happens when an unpredictable crisis hits? That society collapses. The nation's creativity has been castrated.
So I'm Betting on Roots
Let me sum up.
China's efficiency is terrifying. 380,000 cameras in Shenzhen, 7,700 public services, real-time monitored restaurant kitchens. This only works in 'predictable peace.' And their ability to actually build a smart city like Shenzhen is genuinely impressive.
America's system is messy. Palantir, Anduril, SpaceX, hundreds of startups competing and colliding. But in this chaos, the small roots grip the soil deeper and wider. One company fails, the ecosystem survives.
I'm currently holding BOTZ (robotics), ICLN (clean energy), GLU (Gabelli Utility).
Why ETFs? Because I don't know which individual company will win. Tesla Optimus? Figure AI? But one thing's certain: the robotics sector as a whole will grow. America needs robots to bring manufacturing back. When that 2026 executive order drops, the whole sector jumps.
AI data centers devour electricity. The Department of Energy predicts AI-specific power consumption will grow 33% annually. That's why Microsoft and Amazon are signing with nuclear plants. That's why I buy ICLN and GLU.
Investing is betting. I'm betting on roots.
Resilience over efficiency. Survival over speed. Chaos over control.
The future gets more unpredictable. In these rough waves, I think flexible roots survive better than rigid stems.
At least that's how I see it.
Sources:
- [YouTube] America Has Changed - Hyoseok Lee Academy
- CHIPS and Science Act: U.S. Department of Commerce, Wikipedia
- Palantir Maven contracts: DefenseScoop, Breaking Defense, SpaceNews
- China surveillance: CNN, Comparitech, Huawei
- AI training data: Stanford HAI, Brookings Institution
- Trump robotics policy: Politico, Robotics 24/7
This is personal analysis, not investment advice. Investment decisions are your own responsibility.
Comments
Post a Comment