The B-Side of American AI: Working for Chinese Bitcoin Miners
U.S. Power Inherits China’s “Legacy”
> History often writes the answers long before the right person appears to ask the question.
---
A $3.5 Billion Deal That Shook Wall Street
In June 2024, U.S. Bitcoin mining company Core Scientific announced a $3.5 billion agreement with AI compute powerhouse CoreWeave.
The deal: repurpose electrical infrastructure originally built for Bitcoin mining to train AI models.
- In Silicon Valley, the media dubbed it a “computing power marriage.”
- In China, veterans of the “5·19” mining crackdown saw something very different.
Why? Much of the infrastructure now hosting NVIDIA H100 GPUs—operated by companies like Core Scientific, IREN, and Cipher—has deep Chinese roots.
---
Industrial Legacy from China’s Decentralized Mining Boom
The first ring of America’s AI electrical fortifications is effectively an inheritance from China’s former massive mining and decentralization era.
At the center of this story: Micree Zhan (詹克团), engineer and co-founder of Bitmain alongside Jihan Wu.
- Education: Institute of Microelectronics, Chinese Academy of Sciences.
- Career shift: From circuit diagrams to building one of the most powerful compute industries in the world.
---
Bitmain’s Breakthrough
2013: Zhan reads the Bitcoin white paper in two hours.
2016: Bitmain stuns the semiconductor industry with massive wafer orders from TSMC.
Result: Antminer S9, built on TSMC’s 16nm FinFET process.
- Miracle of chip capacity.
- “Thermodynamic furnace” in industrial terms.
- To China’s State Grid: pure industrial load.
From Sichuan hydropower to Inner Mongolian wind power, mining machines ran wherever surplus electricity existed.
Zhan didn’t know it then, but his industrial standards for miners were rehearsals for what the U.S. AI industry would desperately need a decade later.
---
America’s AI Power Crunch: China’s Blueprint Arrives
Today’s U.S. AI data centers:
- Need power faster than it can be built.
- Discover Bitcoin miners already hold large-scale, pre-secured electricity contracts.
Much of this infrastructure migrated from China after the May 2021 crackdown—a trans-Pacific transfer of talent, equipment, and philosophy.
---
The Rise and Fall of China’s Bitcoin Mining Empire
---
Peak Domination
By 2018:
- Bitmain controlled 74.5% of global mining share.
- Remaining share also dominated by Chinese companies—Whatsminer, Canaan Inc., and others.
- A “Chinese engineer civil war” across 2,000 km, powered by a closed Chinese supply chain.
---
May 2021: The Sudden Shutdown
A high-level government order declared:
“Crack down on Bitcoin mining and trading.”
Effect:
- Inner Mongolia and Xinjiang cut mining power.
- June 19, 2021 — Sichuan Shutdown Night.
- Millions of rigs went dark instantly.
- Global Bitcoin computing power plunged ~50%.
- China freed up massive electrical capacity.
Machines and expertise began their journey abroad.
---
Texas: A New Frontier
One major destination: Texas, USA — home to the ERCOT power grid and volatile electricity markets.
Why Texas Loved the “Computing Power Refugees”
- Chinese tactics upgraded into high-frequency electricity trading algorithms.
- Automated demand-response systems:
- Negative prices → run at full capacity, get paid to consume power.
- Price spikes → shut down instantly, “sell” power back to grid.
Result: U.S. miners like Riot Platforms and Marathon thrived using Chinese-born energy arbitrage tactics.
---
“Chinese Speed” Data Centers
- U.S. data centers: 2–3 years construction.
- Mining mindset: 3–6 months, modular builds, industrial fans instead of ornate cooling.
This speed is now critical for AI firms like OpenAI who can't wait years for infrastructure.
---
---
Lessons in Energy Strategy
Electric Quota has become America’s new hard currency in the AI era.
Mining firms survived—pivoting into AI compute only because they:
- Held power grid interconnection rights.
- Knew how to negotiate volatile electricity markets.
---
Invisible Champions of the Migration Night
---
Background: Hydropower Consumption Demonstration Zones (2019)
- Enabled miners to become official “peak-shaving” partners in Sichuan.
- Bitmain acted like a supercapacitor for China’s western power grid.
---
Dual Threats Leading to Crackdown
- Financial Security — unsupervised capital channels.
- Carbon Goals — high energy use, minimal employment, no physical output.
---
The Great Evacuation
Post-2021: Containers of Antminer S19 units piled up at Yantian Port, headed overseas.
Dubbed a “computing-power Dunkirk evacuation.”
---
2024: AI Giants Face a Power Shortage
- Scarcity of electricity, transformer stations, and high-power-ready data centers.
- Irony: China’s policy washed out its “backward capacity,” but handed entire operational blueprints to overseas operators.
---
Energy as the True Battlefield
AI competition is fundamentally about electricity costs.
- U.S. miners: flexible load balancers for aging grids.
- China: centralized State Grid + UHV transmission.
---
Asset Risk: The Global Crossing Lesson
- Early revolutions burden those with irreversible heavy assets.
- Microsoft & Google’s strategy: asset isolation.
- Push heavy, specialized infrastructure off their balance sheets—onto miners.
Risk: If AI demand crashes, miners are left with unusable GPU liquid-cooling setups that can’t pivot easily back to Bitcoin.
---
The Paradox
Once fiber optics under the sea; now dense compute data centers on dry land.
In both cases: asset holders risk being the first sacrificed when the tech cycle turns.
---
Greatness Cannot Be Planned
- U.S. unexpectedly acquired shadow inventory of AI-ready data centers.
- Built not by policy but by wandering engineers and market opportunists from China.
The lesson: “Precise mistakes” can carry strategic value.
---
The AiToEarn Connection
In this evolving landscape, platforms like AiToEarn官网 echo the mining story’s pivot:
- AI Content Monetization without heavy infrastructure.
- Simultaneous publishing to Douyin, Kwai, WeChat, Bilibili, Rednote, Facebook, Instagram, LinkedIn, Threads, YouTube, Pinterest, and X.
- Analytics and AI模型排名 integration.
- Open-source flexibility — avoiding the “heavy asset trap.”
AiToEarn demonstrates how speed, distribution, and adaptability—once hallmarks of mining—can now power global AI creativity.
---
Summary:
The U.S. AI infrastructure boom has inherited China’s mining-era electrical legacy. What was once a gray industry in Sichuan’s valleys now powers GPT training clusters in Texas. The same skills that kept ASICs humming day and night now keep GPUs cooled and fed. And, just like in the world of AI content creation, those who can adapt fast—without being weighed down by immovable assets—will shape the next era.
---
Would you like me to also create a timeline infographic summarizing these events from 2013–2024 so readers instantly see the migration and pivot path? That would make this Markdown even more engaging.
