scripod.com

How AI can solve its own energy crisis | Varun Sivaram

TED Talks Daily

2025/11/03
TED Talks Daily

TED Talks Daily

2025/11/03
As artificial intelligence surges forward, its hunger for energy is reaching a critical point—threatening to overwhelm aging power grids and disrupt the pace of innovation. Physicist Varun Sivaram joins the conversation to explore how AI doesn’t have to be a burden on our infrastructure, but could instead become a dynamic force for modernizing the grid and unlocking a cleaner, more resilient energy future.
The exponential rise in AI computing is placing unprecedented stress on power grids, risking higher costs and increased reliance on fossil fuels. However, flexible AI data centers offer a transformative solution by dynamically adjusting their energy consumption based on grid demand and renewable availability. By shifting workloads in time and location—pausing during peak hours or moving computations to regions with surplus power—these data centers can ease strain on infrastructure and integrate more solar and wind energy. Technologies like Emerald Conductor and initiatives such as DC Flex are pioneering intelligent orchestration systems that enable this responsiveness. When aligned with solar production peaks, flexible AI not only reduces emissions but also accelerates clean energy adoption. With strategic coordination between tech companies, utilities, and policymakers, AI could evolve from a grid challenge into a catalyst for a smarter, more sustainable energy system.
03:04
03:04
Flexible AI data centers can help power the grid by reducing consumption during peak demand
07:11
07:11
Data centers can pause or move workloads to times with cheaper, cleaner energy.
11:04
11:04
Flexible AI data centers can relieve the grid and promote clean energy use
15:06
15:06
Flexible AI data centers could match solar peaks to integrate more clean energy.