[Part 1] The AI Era: Where is the Money Flowing Next?

The core narrative is not about Bitcoin.

Rather, it is the structural migration of capital:
AI Infrastructure → Data Centers → Energy/Power → SMRs/Cooling → Physical AI.

1. The Prolonged Crypto Winter and Market Outlook

Crypto is currently in a deep winter.

The cryptocurrency market is suffering from severely weakened liquidity. Institutional buying pressure has dried up, while retail capital has rotated aggressively into AI equities. As a result, exchange trading volumes have plummeted, suggesting that Bitcoin will likely consolidate in a sluggish, low-momentum range for the short term.

Potential market bottoms are projected as follows:

• Earliest: Q4 2026

• Latest: H1 2027

In short, now is not the time to go all-in. The optimal approach is simple: Accumulate via Dollar-Cost Averaging (DCA) while monitoring regulatory shifts.

2. The Bitcoin Catalyst: Regulatory Frameworks, Not Price Pumping

While the Trump administration is undeniably pro-crypto, its primary objective is not to artificially boost Bitcoin's price overnight. Instead, the focus is on systemic institutionalization:

• Passing comprehensive crypto legislation (e.g., Clarity Act)

• Establishing legal frameworks for stablecoins

• Building a US-centric digital financial order

The big picture is about constructing an entire crypto ecosystem, not just sparking a short-term rally in Bitcoin.

3. Stablecoins: The True Strategic Target for the US

Compared to legacy financial systems, stablecoins are:

• Substantially faster

• Drastically cheaper in transaction fees

• Easily scalable as a dollar-backed global payment rails

For the US, stablecoins are a far more critical piece of financial infrastructure than Bitcoin. Ultimately, the goal is to extend and maintain US Dollar hegemony into the digital era.

4. Quantum Computing and the Viability of Bitcoin

It is true that if quantum computers advance sufficiently, legacy cryptographic security models could be compromised. However, Bitcoin functions as an evolving ecosystem capable of continuous upgrades through:

• Hard forks and soft forks

• Integrating quantum-resistant cryptography

• Security enhancements via community consensus

Panic-selling Bitcoin today due to quantum threats is an overreaction. That said, it is vital to monitor how proactively the network adopts quantum-resistant security updates moving forward.

5. Debunking the AI Bubble Theory

The AI sector is not in a bubble. Unlike the Dot-Com bubble, AI is already delivering tangible utility. We are actively leveraging AI to write, code, translate, and drive exponential gains in workplace productivity—proving it is a technology backed by real-world value.

Currently, we are only in "Stage 1," where the primary focus is training AI on the sum of human knowledge. This is exactly why capital is flooding into AI infrastructure:

GPU → HBM → Data Centers → Power

Capital is not chasing consumer AI apps right now; it is rushing to build the foundational infrastructure required to run those services. AI is at the absolute dawn of a 50-to-100-year civilizational shift. We must view the current market as a pure infrastructure-building phase.

6. The Sequential Migration of AI Infrastructure Bottlenecks

AI development does not rely solely on software. As AI scales, physical infrastructure bottlenecks will migrate sequentially through five distinct stages:

• ① GPUs: The absolute prerequisite for AI training. NVIDIA currently dominates this layer.

• ② HBM (High Bandwidth Memory): GPUs cannot operate at peak performance without advanced memory. This positions SK Hynix and Samsung Electronics as critical bottleneck bottlenecks.

• ③ Data Centers: Once GPUs and HBM are secured, the bottleneck shifts to the physical real estate needed to house and run these clusters.

• ④ Power/Energy: Data centers consume an astronomical amount of electricity. As AI models scale, power availability becomes the ultimate constraint.

• ⑤ Cooling, Land, and Grid Infrastructure: Providing power is only half the battle. Data centers generate extreme heat, making advanced cooling technologies, grid integration, and site acquisition the final structural bottlenecks.

The Flow of Capital & Bottlenecks:

GPUs -> HBM -> Data Centers -> Power Grid -> Cooling & Infrastructure

7. Data Centers: The Impending Structural Chokepoint

To expand AI services, data center capacity must scale. However, building a data center is a slow, complex physical process constrained by:

• Land acquisition

• Regulatory zoning and permits

• Power grid allocation

• Cooling infrastructure setup

While AI software evolves exponentially, physical infrastructure cannot keep pace. Ironically, this infrastructure lag may act as a natural circuit breaker, preventing the AI industry from overheating too quickly.

8. The Energy Crisis: The Defining Challenge of the AI Era

As data centers multiply, energy consumption is projected to explode. Parts of the US could face severe power deficits as early as 2027.

• Building new coal or gas-fired power plants takes years.

• Constructing traditional nuclear plants takes over a decade.

• Meanwhile, data center demand is scaling exponentially.

Consequently, the future of the AI industry will be decided by three concurrent battlegrounds: The Data War, the Semiconductor War, and the Energy War.

9. Why SMRs are a High-Conviction AI Infrastructure Play

SMRs (Small Modular Reactors) represent the next generation of nuclear energy. Compared to legacy nuclear plants, SMRs are:

• Massively scaled down in footprint

• Fully modularized for factory fabrication

• Significantly faster to deploy on-site

Because data centers require an uninterrupted, baseload power supply, SMRs are shifting from a traditional utility play into a core AI infrastructure asset. They represent a convergence of three major market themes:

• The AI Infrastructure Play

• The Data Center Power Supply Narrative

• The Energy Bottleneck Solution

SMRs will remain a highly viable, long-term secular growth sector throughout the AI era.

10. Next-Gen Cooling Technologies and Marine Data Centers

The hardware inside data centers generates immense heat, meaning a massive percentage of their energy consumption goes strictly toward cooling. To solve this, the industry is pivoting toward radical solutions like Marine Data Centers.

By submerging server infrastructure underwater, companies can leverage natural seawater as a highly efficient, constant coolant. South Korea is already planning pilot projects for marine data centers off the coast of Ulsan.

Key emerging investment themes to watch in this space include:

• Advanced Data Center Hardware & Layouts

• Liquid & Immersion Cooling Technologies

• Power Usage Effectiveness (PUE) Optimization

• Subsea and Marine Data Centers

[ View Part 2 ]

Source: Key Variables for Bitcoin Investors