Can Data Centers Keep Up with AI Demand?

The TeleGeography Explains the Internet podcast welcomes Phill Lawson-Shanks, Chief Innovation Officer at Aligned Data Centers. With several decades of experience in digital infrastructure, Phill has seen the industry evolve from mainframes and green screens to the massive hyperscale cloud platforms we rely on today.

Key topics covered:

• The shift in data center geography: Why the "center of gravity" for networks is moving toward new power-rich zones and how the rise of AI is rewriting the rules of connectivity.
  
  

• The engineering challenges of AI: From floor loading for 6,000-pound racks to the transition from air to liquid cooling, Phill explains the "fungible" design needed for modern AI factories.
  
  

• The infrastructure bottleneck: Why AI is ultimately a network-dependent technology and how high-speed optical networking between buildings is pushing against the limits of physics.
  
  

• Sustainability and power: How the industry is chasing power availability and pioneering new ways to build modular, carbon-traceable infrastructure.
  
  

• The future of AGI: Phill shares his philosophical take on AGI and why he believes we're living through the most exciting time in human history. 

Chasing power, not just latency

Historically, data centers were built where the "eyeballs" were to minimize latency. AI has flipped that script. We are seeing a massive shift in data center geography toward "power-rich zones."

As Phill notes, the industry is currently treating AI like the first industrial revolution—utilizing every available inch of warehouse space until new "factories" can be built. Today, data centers are built wherever sustainable power is available. It is now often cheaper to trench hundreds of miles of fiber to a power source than to wait for a utility company to build new transmission lines.

Engineering the "AI factory"

Building for AI isn't just about adding more servers; it’s about reimagining the physical limits of a building. Phill highlights two massive engineering hurdles:

• Weight: Traditional data center floors are designed for 3,000 pounds per square foot. Modern AI racks, packed with GPUs, are hitting 6,000 pounds. This has forced a shift away from raised floors toward building directly on reinforced slabs.
• Thermal management: We are moving from air cooling to liquid cooling. Aligned Data Centers utilizes a "fungible" design, allowing customers to transition between air and liquid cooling "in-flight" as hardware evolves.
  
  

"Fungibility is critical; we design in a modular fashion so our clients can transition... as chip form factors change." — Phill Lawson Shanks

The network bottleneck and physics

While much of the hype focuses on chips, AI is ultimately a network-dependent technology. In an AI cluster, every GPU must talk to every other GPU. This requires massive, high-speed optical networking between buildings that pushes against the literal limits of physics.

Interestingly, Phill points out that within these short spans, copper can actually be faster than fiber, though fiber remains king for distance. The "Local Area Network" (LAN) has never been more important than it is inside the modern AI campus.

Sustainability through innovation

The industry is no longer just chasing the lowest cost of power; it is chasing power availability and sustainability. Aligned Data Centers is pioneering modular, carbon-traceable infrastructure, even building and "gifting" substations to utilities to ensure the grid can keep up. By using advanced cooling arrays, they can cool 50 kilowatts in a single cabinet, dramatically shrinking the physical and carbon footprint of the compute.

The future of AGI?

When asked about the future of Artificial General Intelligence (AGI), Phill offers a grounded perspective. While AI is world-class at pattern recognition, we are still far from understanding what human consciousness is. "It's hard to say if we'll get there, when we'll get there, and would we know it when we saw it?"

Despite the infrastructure bottlenecks and the "insatiable demand" for power, Phill remains incredibly optimistic. He believes we are living through the most exciting time in human history—a recursive era where software and hardware push each other to heights previously unimagined.