The Hidden AI Trade: Land, Power, & Compute | With Mike Alfred
The Hidden AI Trade: Land, Power, & Compute | With Mike Alfred
YouTube16 min 54 sec
Watch on YouTube
Note: AI-generated summary based on third-party content. Not financial advice. Read more.
Quick Insights

Maintain a core long position in Bitcoin (BTC) as a foundational asset, prioritizing long-term holding over frequent trading to capture its historical alignment with AI sector growth. For direct exposure to AI infrastructure, Iris Energy (IREN) offers a high-conviction play due to its vertical integration of land, power, and high-performance GPU cloud services. Investors seeking "landlord" models in the power space should look at Cipher Mining (CIFR) and Terawulf (WULF), which leverage low-cost energy contracts in West Texas and Oklahoma. Focus on companies that secured power interconnects years ago, as physical infrastructure and cooling have replaced chip availability as the primary bottleneck for AI scaling. Treat any significant market drawdowns as generational buying opportunities for NVIDIA (NVDA) and other firms that wield massive compute power at scale.

Detailed Analysis

Bitcoin (BTC)

The discussion highlights Bitcoin as a foundational asset that has outperformed most other sectors over the last three years, despite periods of high volatility and negative sentiment.

  • Performance Trends: Bitcoin has largely mirrored the returns of major AI companies since Q1 2023.
  • Market Sentiment: While many investors focus on recent price fluctuations, the long-term trend remains positive.
  • The "Only" Crypto Play: The guest suggests that for most investors, a "straight long" Bitcoin position has been the only reliable way to make money in the digital asset space recently, as altcoins have struggled to keep pace.

Takeaways

  • Long-term Holding: View Bitcoin as a core holding that can withstand macro "noise" (recessions, geopolitical issues).
  • Avoid Over-trading: The guest emphasizes holding through drawdowns rather than trying to time the market.

Iris Energy (IREN)

Note: Referred to as "Iron" in the transcript. The guest is a board member and major investor in this company, which has transitioned from a pure Bitcoin miner to a high-performance computing (HPC) and AI infrastructure provider.

  • Vertical Integration: Unlike "asset-light" competitors, IREN owns the land, the power interconnects, and the physical buildings. They are now moving to own the actual GPUs (chips) to provide a full-stack AI cloud service.
  • Historical Valuation: The guest noted a "margin of safety" in 2022/2023 when the stock traded near $1.00, claiming the liquidation value of the physical land and power assets was worth 2-4x the market cap at that time.
  • Growth Potential: The stock has seen massive growth (peaking around $75-$76), but the guest believes it has not yet reached its "fundamental value" due to the insatiable demand for AI compute.

Takeaways

  • Infrastructure is the Moat: The real value is not just in the chips, but in the "scarce land, power, and infrastructure" required to plug them in.
  • Monitor Execution: The key risk/opportunity is whether the company can successfully scale its own "NeoCloud" business to compete with providers like CoreWeave.

Cipher Mining (CIFR) & Terawulf (WULF)

These companies are identified as key players in the infrastructure space that originated in Bitcoin mining.

  • Business Model: These firms are recognized for being "world-class infrastructure developers."
  • Focus: They specialize in low-cost energy environments (West Texas, Oklahoma) where excess renewable power is available.
  • Co-location Strategy: Some of these players may act as "landlords," leasing their powered data center space to others rather than running the AI compute themselves.

Takeaways

  • Secondary AI Plays: These stocks offer exposure to the "power and land" theme without necessarily requiring the massive capital expenditure of buying their own GPUs.

AI Infrastructure & "NeoClouds"

The transcript identifies a massive shift in how data centers are built and who controls the "compute" (processing power).

  • The Chokepoint: The bottleneck for AI is no longer just the availability of NVIDIA chips, but the physical locations (power and cooling) to run them.
  • Key Players Mentioned:
    • NVIDIA (NVDA): The primary chip provider, though some chips are currently "sitting in warehouses" due to a lack of data center space.
    • CoreWeave / Nebius: "NeoCloud" providers that are often asset-light, leasing space from others.
    • XAI (Elon Musk’s AI company): Noted for building the "Colossus" data center in Memphis extremely quickly.
    • Anthropic: Highlighted as a leader in the space, reportedly showing positive EBITDA (profitability) even before an IPO.

Takeaways

  • The "1999 Internet" Parallel: AI may be in a bubble-like phase similar to the late 90s. Even if a 50%+ drawdown occurs, the long-term (20-30 year) trend is expected to be 100x higher.
  • Investment Strategy: Look for companies that "wield the most compute at scale." If a major market crash occurs, use it as a generational buying opportunity for AI infrastructure.

Investment Themes: Land, Power, and Compute

The overarching investment thesis is that the "digital gold rush" is actually a "physical land grab."

  • Geographic Shift: Traditional data center hubs (like Northern Virginia) are becoming less relevant compared to middle-America (Texas, Oklahoma) where power is cheaper and more abundant.
  • First Principles Thinking: Investors should look for companies that secured power contracts and land years ago, as these are now the "gating factors" for the entire AI industry.
  • Idiosyncratic Investing: The guest suggests ignoring broad market fears (recessions, yield curves) and focusing on concentrated bets in these specific, high-demand themes.
Ask about this postAnswers are grounded in this post's content.
Video Description
Ash Bennington sits down with Mike Alfred to unpack one of the most overlooked ideas in markets right now: AI may be less about software than about scarce land, power, and physical infrastructure. In the secodn half, Alfred explains why the real bottleneck in AI is compute deployment, how bitcoin miners evolved into infrastructure developers, why he still likes Bitcoin long term, and where stablecoins and digital asset payment rails may fit into the next phase of the cycle. Check out the full interview on Real Vision! 🔥 Get 𝗙𝗥𝗘𝗘 𝗔𝗖𝗖𝗘𝗦𝗦 to Real Vision https://rvtv.io/3YOZZUe Timestamps: 0:00 - Why This Market Environment Is So Unusual 0:54 - Why AI Has Been the Only Trade That Really Worked 3:25 - From Bitcoin Miners to AI Infrastructure 8:28 - The Real AI Stack: Land, Power, and Compute 12:29 - Why Traditional Valuation Metrics Miss the Point About Real Vision™: We arm you with the knowledge, the tools, and the network to succeed in your financial journey. Connect with Real Vision™ Online: Twitter: https://rvtv.io/twitter Instagram: https://rvtv.io/instagram Website: 🔥 https://rvtv.io/3Y4t5Pw 📈 Get your Real Vision swag: https://shop.realvision.com 📣 Elevate your brand with Real Vision. Connect with us at partnerships@realvision.com to explore advertising possibilities. Disclaimer: https://media.realvision.com/wp/20231004185303/Disclaimer-1.pdf #mikealfred #ashbennington #ai #bitcoin #datacenters #infrastructure #power #stablecoins #crypto #investing #macro #realvision #land #compute #markets
About Real Vision
Real Vision

Real Vision

By @realvisionfinance

We arm you with the knowledge, the tools, and the network to succeed on your financial journey.