
夏雪宜
夏雪宜
Focus on US stocks and walk with great companies.
0Following
8.7Kfollowers
Feed
Feed
Core US Stock Asset Package for the AI Era! Xia Xueyi's Selected 【Top 20 Pure US/ADR Giants Benefiting the Most】 (Five Core Clusters):
🔥 Core Foundation and Super Clusters (Chips/Models)
• $NVDA: The AI era's "digital oil" monopolist, with the CUDA ecosystem firmly controlling the absolute choke point.
• $AVGO: The absolute leader in custom ASICs and data center network switching chips.
• $MSFT: Azure cloud + OpenAI strong alliance, the strongest empire in generative AI commercialization.
• $GOOGL: From TPU chips to top-tier cloud ecosystem full-stack closed loop, a giant in AI-powered search.
• $AMZN: AWS, the world's largest cloud provider, capturing the biggest dividend from long-tail computing power demand.
🔥 Advanced Process and High-Speed Interconnect (Optical/Copper/Foundry)
• $TSM: The world's only advanced process foundry, the physical carrier of all top AI chips.
• $AMD: The market's only top-tier GPU alternative provider, sharing in the computing power dividend.
• $CRDO: Leader in high-speed connectivity, AEC technology monopolizes the "copper interconnect" inside racks, unleashing dividends.
• $FN: Leader in precision optical manufacturing, perfectly capturing the essential demand for 1.6T high-speed optical module capacity expansion.
• $GLW: Optical fiber materials giant, a direct beneficiary of the demand for special optical fiber network density from the mega cluster.
🔥 Energy, Power, and Liquid Cooling (Data Center Water-to-Power)
• $VST: The largest independent nuclear and clean energy provider in the US, meeting AI's thirst for zero-carbon power.
• $CEG: The US nuclear power leader, monopolizing deterministic clean energy through long-term power purchase agreements (PPA).
• $VRT: The absolute ace in thermal and power management, core beneficiary of GB200 high-density liquid cooling.
• $ETN: Electrical equipment giant, controlling the most critical power bottleneck for AI expansion from grid to distribution.
🔥 Advanced Storage and Device Infrastructure (Memory/Manufacturing)
• $MU: Core player in HBM and high-capacity DRAM, directly benefiting from computing power's demand for ultra-high-speed storage.
• $ASML: Exclusive supplier of EUV lithography machines, the creator unavoidable in all advanced process chip capacity expansions.
• $ALAB: Leader in PCIe Retimer chips, solving the essential latency issues inside AI servers.
• $ANSS: Leader in physical simulation software, the digital window for R&D of chips, optical modules, and liquid cooling systems.
🔥 Edge Side and Application Deployment (Ultimate Monetization)
• $AAPL: With Apple Intelligence and billions of active users, the largest edge AI monetization platform.
• $PLTR: The pinnacle of enterprise and military AI data analysis, with the AIP platform generating high per-customer revenue and crazy monetization.
💡 Investment Logic: Hardware + Interconnect (capture the highest gross margin) ➡️ Energy + Liquid Cooling (break infrastructure bottlenecks) ➡️ Edge + Applications (future 100x win rate). Join the watchlist with one click!👇 #AI #USStocks #Investment
US stock CRCL has a promising future
It might be very profitable in the future
And make big money, surpassing Vasa
But not now, it needs time
Maybe after 5 years
But it will definitely be very profitable
Maybe after 10 years
The key is not whether your cost is 50, 100, or higher
The key is whether you can hold on, and for how long
This tweet of mine ⬇️ has stirred up a hornet's nest among the Japanese 🇯🇵! 😱 The comment section needs some support!
夏雪宜
I sincerely want to ask the Japanese people:
Don't you really think that the invasion wars launched against East Asian neighbors such as China and South Korea back then—
breaking into others' homes to burn, kill, and loot, and committing atrocities like the Nanjing Massacre—
were serious mistakes and should never have happened?
There is a story in "Genesis."
Humans once spoke the same language. They gathered together and decided to build a tower that reached the heavens—the Tower of Babel. God glanced at it and said something very interesting:
"Look, they are one people, and they all have one language. And as they begin to do this, nothing they plan to do will be impossible for them."
Then God scattered the languages of humanity.
This story has been passed down for thousands of years, and most people read it as a myth. But if you view it as a model about collaboration efficiency, it speaks of a very simple truth:
When everyone can exchange information in the same way without loss, the upper limit of human capability becomes immeasurable.
What God did—scattering the languages—essentially injected friction into the human collaboration system.
Since then, all human civilization’s institutions, commerce, and culture have actually grown under this premise of "friction."
Translation is a product of friction. Diplomacy is a product of friction. Middlemen are a product of friction. Networks, relationships, social dinners, trust mechanisms, layers of agency—all are products of friction.
All the complex systems humans have built over thousands of years, to some extent, answer the same question:
How do people collaborate when languages differ and information is asymmetric?
The answer is: spend time, money, favors, drinks, and relationships to build trust.
We are so familiar with this system that we think it is the way the world operates.
But it is not. It is just a solution under the constraint of "language barriers."
When the constraint changes, the solution collapses accordingly.
01
Between 2024 and 2026, something quietly became the infrastructure of this era.
It is called Token.
Most people first hear this word when discussing large model pricing—"how much does one million Tokens cost?" It sounds like a technical parameter, a billing unit.
But if you take a step back, Token is becoming something entirely new.
It is not Chinese, not English, nor any human language. Yet it can carry information from all human languages.
It is not code, nor mathematical symbols, but it can express logic, reasoning, planning, and intent.
It is not just a medium for human-AI dialogue—it is becoming a universal exchange unit between humans, humans and machines, and machines themselves.
A Chinese person writes a passage, which becomes Tokens; when a Brazilian reads it, the Tokens transform into Portuguese. There is no translator, no delay, no information loss in between.
But this is just the surface.
Deeper still: you turn a complex request into Tokens and give it to AI; AI understands your intent, calls tools, searches information, integrates resources, generates results, and returns them to you in Tokens.
In this process, Tokens are not just language. They are carriers of intent, interfaces of capability, and the smallest unit of collaboration.
For the first time in human history, there is something that allows all intelligent agents—whether human or machine—to exchange intent and capability in the same format.
The Tower of Babel is being rebuilt. Not by everyone learning the same foreign language, but through a brand-new medium.
02
If Token is the new universal language, then we need to take another look at what the old world runs on.
Have you ever wondered why Moutai costs three thousand yuan a bottle?
It’s not because it tastes better. Honestly, in blind tests, most people can’t tell Moutai apart from other sauce-flavored liquors.
Moutai is worth three thousand yuan because it is the physical embodiment of trust cost.
Chinese commercial society has an underlying logic that has operated for decades: if you want to get a project, access a resource, or open a channel, you must first build trust. How? By dining, drinking, gifting, building relationships, and spending time.
Moutai is the hard currency in this logic. When you invite someone to drink Moutai, you’re not just inviting them to drink. You are saying in a way everyone recognizes: "I am willing to pay the cost for this relationship."
What is this essentially? It is friction.
There is an opaque layer between you and the resource. You don’t know the person, and they don’t trust you. This distance must be bridged with time, favors, and money.
A large part of traditional business profits is hidden in this gap. All middlemen, brokers, BD, PR, and channel agents don’t sell products—they sell "who they know."
This system has a more precise name: friction economy.
The core asset of the friction economy is not products or technology, but networks of relationships. Whoever has denser, deeper, and broader networks gains more resources.
This is not unique to China. It’s the same worldwide. Wall Street has its drinking parties, Silicon Valley has its circles, Hollywood has its parties. Different forms, same essence—using social costs to overcome information asymmetry.
This system has operated for thousands of years. It was not designed by anyone; it naturally grew under the constraint of "language barriers and opaque information."
It is stable, mature, and deeply rooted.
Until the constraint changes.
03
When you can use Tokens to access information, integrate resources, and mobilize AI to produce—you don’t need to know anyone.
This sentence is worth reading again.
You don’t need to know anyone.
A twenty-year-old young person sitting in a rented room in a county town, with no connections, no background, no industry resources. But they have a computer and enough Tokens.
They can have AI help research a market, analyze competitors, write business plans, generate product prototypes, create marketing materials, build sales channels, and connect supply chain information.
In the past, each of these tasks required "knowing people." Knowing market researchers, designers, channel owners, and people who could introduce you.
Now, each of these tasks can be done with Tokens.
This is not just an efficiency improvement. Efficiency improvement means "I do the same thing faster."
This is a logical replacement. It means "I don’t need to take the old path at all."
The old path was: have an idea → build connections → build trust → get resources → produce → deliver.
The new path is: have an idea → Token → deliver.
The entire chain of relationships in between is compressed into a single flow of Tokens.

