AI infrastructure is expanding faster than the power systems required to support it. A data center can be built in two to three years, while a new power plant or transmission line may take seven to nine years. That gap puts utilities at the center of the next phase of AI growth.

Vik Chaudhry, cofounder and CTO of Buzz Solutions, explains how utilities are using visual AI, computer vision, drones, and infrastructure data to find defects, prioritize maintenance, prevent outages, and reduce wildfire risk. He also discusses how AI can help utilities uncover existing grid capacity, forecast unpredictable demand, control operating costs, and preserve knowledge as experienced workers retire.

What You’ll Take Away

• Why electricity, not computing chips, may become the largest constraint on AI growth

• How utilities can extract more capacity from existing infrastructure while new power generation is built

• Where visual AI helps teams prioritize inspections, repairs, and maintenance spending

• How AI can improve load forecasting and transfer knowledge to the next generation of utility workers

A Moment Worth Pulling Out

“The biggest problem for AI right now is not the chips. It’s the electrons.”

Key Moments

Approximate timestamps based on the transcript.

00:45 How Buzz Solutions uses visual AI to assess power infrastructure

03:05 Why utilities began building internal AI teams and governance processes

06:45 The energy constraint behind data center and AI expansion

08:20 Why data centers can be built much faster than new power infrastructure

11:50 Balancing data center demand with affordability for consumers

26:35 Using AI for load forecasting and utility workforce knowledge transfer

Follow The Tech Trek for more conversations about how technical teams are building and operating around AI, data, platforms, product, and engineering.