ARTIFICIAL INTELLIGENCE

“Data centers used 4% of U.S. electricity two years ago and are on track to devour three times that by 2028.” – Hari Vasudevan – Utility Dive

Hari Vasudevan is the founder and CEO of KYRO AI, an AI-powered platform designed to streamline operations in utilities, vegetation management, disaster response, and critical infrastructure projects, supporting over $150 billion in program value by enhancing safety, efficiency, and cost savings for contractors and service providers.^1,3,4

Backstory and Context of the Quote

The quote—”Utilities that embrace artificial intelligence will set reliability and affordability standards for decades to come”—originates from Vasudevan’s November 26, 2025, opinion piece in Utility Dive titled “Data centers are breaking the old grid. Let AI build the new one.”^1,6 In it, he addresses the grid’s strain from surging data center demand fueled by AI, exemplified by Georgia regulators’ summer 2025 rules to protect residential customers from related cost hikes.⁶ Vasudevan argues that the U.S. power grid faces an “inflection point,” where clinging to a reactive 20th-century model leads to higher bills and outages, while AI adoption enables a resilient system balancing homes, businesses, and digital infrastructure.^1,6 This piece builds on his November 2025 Energy Intelligence article urging utilities and hyperscalers (e.g., tech giants building data centers) to collaborate via dynamic load management, on-site generation, and shared capital risks to avoid burdening ratepayers.⁵ The context reflects escalating challenges: data centers are driving grid overloads, extreme weather has caused $455 billion in U.S. storm damage since 1980 (one-third in the last five years), and utility rate disallowances have risen to 35-40% from 2019-2023 amid regulatory scrutiny.^4,5,6

Vasudevan’s perspective stems from hands-on experience. He founded Think Power Solutions to provide construction management and project oversight for electric utilities, managing multi-billion-dollar programs nationwide and achieving a 100% increase in working capital turns alongside 57% growth by improving billing accuracy, reducing delays, and bridging field-office gaps in thin-margin industries.³ After exiting as CEO, he launched KYRO AI to apply these efficiencies at scale, particularly for storm response—where AI optimizes workflows for linemen, fleets, and regulators amid rising billion-dollar weather events—and infrastructure buildouts like transmission lines powering data centers.^3,4 In a CCCT podcast, he emphasized AI’s role in powering the economy during uncertain times, closing gaps that erode profits, and aiding small construction businesses.³

Leading Theorists in AI for Grid Modernization and Utility Resilience

Vasudevan’s advocacy aligns with pioneering work in AI applications for energy systems. Key theorists include:

• Amory Lovins: Co-founder of Rocky Mountain Institute, Lovins pioneered “soft path” energy theory in the 1970s, advocating distributed resources over centralized grids—a concept echoed in maximizing home/business energy assets for resilience, as Vasudevan supports via AI orchestration.¹
• Massoud Amin: Often called the “father of the smart grid,” Amin (University of Minnesota) developed early frameworks for AI-driven, self-healing grids in the 2000s, integrating sensors and automation to prevent blackouts and enhance reliability amid data center loads.^4,6
• Andrew Ng: Stanford professor and AI pioneer (co-founder of Coursera, former Baidu chief scientist), Ng has theorized AI’s role in predictive grid maintenance and demand forecasting since 2010s deep learning breakthroughs, directly influencing tools like KYRO for storm response and vegetation management.^3,4
• Bri-Mathias Hodge: NREL researcher advancing AI/ML for renewable integration and grid stability, with models optimizing distributed energy resources—core to Vasudevan’s push against “breaking the old grid.”^1,5

These theorists provide the intellectual foundation: Lovins for decentralization, Amin for smart infrastructure, Ng for scalable AI, and Hodge for optimization, all converging on AI as essential for affordable, resilient grids facing AI-driven demand.^1,4,5,6

References

1. https://www.utilitydive.com/opinion/

2. https://www.utilitydive.com/?page=1&p=505

3. https://www.youtube.com/watch?v=g8q16BWXk4o

4. https://www.utilitydive.com/news/ai-utility-storm-response-kyro/752172/

5. https://www.energyintel.com/0000019b-2712-d02f-adfb-e7932e490000

6. https://www.utilitydive.com/news/ai-utilities-reliability-cost/805224/