Podcast Summary: TED Talks Daily

Episode Title: The story you're not hearing about AI data centers | Ayșe Coskun
Date: February 25, 2026
Speaker: Ayșe Coskun
Host Introduction: Elise Hu

Episode Overview

This episode of TED Talks Daily features computer scientist Ayșe Coskun, who challenges the prevailing narrative around AI data centers by offering a transformative perspective: rather than merely seeing these centers as voracious “energy hogs,” they can instead be harnessed as dynamic assets for the power grid. Coskun explains how power-flexible data centers could stabilize the grid, prevent blackouts, make electricity more affordable, and accelerate clean energy adoption—if we are bold enough to rethink their role.

Key Discussion Points and Insights

1. The Growing Energy Demand of AI Data Centers

• AI Boom and Power Infrastructure Strain
  • There is an “AI race,” with companies and institutions rushing to build bigger models—and, crucially, more data centers to underpin those systems.
  • In the U.S., AI data centers are requesting energy loads equivalent to entire cities, and some utilities are unable to keep up.
• Notable Stats:
  • Training models like GPT-4 has consumed electricity comparable to thousands of homes annually.
  • In Ireland, nearly 20% of the nation’s electricity now goes to data centers.
  • In Virginia’s Data Center Alley, residents’ electricity bills have risen by 20% as utilities struggle to serve new facilities.
  • “So energy hog label seems well deserved. But that's only half the story.” (Ayșe Coskun, 06:23)

2. Rethinking the Role of Data Centers

• From Energy Hog to Flexible Asset
  • Coskun describes data centers as potential "muscles" for the grid:

    “Unlike our homes or hospitals, AI data centers run jobs that are predictable, controllable, and often delayable. That makes them ideal to help balance supply and demand on the grid.” (07:38)

  • AI data centers can be made power-flexible: adjusting load in real-time to match grid needs, absorbing surplus renewable energy, or reducing activity during peak demand.
• The Renewable Energy Connection
  • “The AI boom is arriving just as the renewable boom is also taking off. Wind and solar don’t follow our schedules, but data centers can.” (08:38)

3. The Origin of Power-Flexible Data Centers

• Early Vision and Resistance
  • Coskun recounts her own journey pioneering flexible computing systems:

    “Early in my career, I asked a question that many found unrealistic. Could computer systems adapt their behavior depending on power grid needs, but without breaking their performance promise to their users?” (08:56)

  • She faced skepticism and setbacks, but persisted in developing systems that could slow down non-urgent tasks for the grid’s benefit.
• Key Breakthroughs
  • Discovering that “not all computing tasks are urgent” opened the door for intentional slowdown and rescheduling.
  • Reframed design goal: “Instead of asking, how do we compute as fast as possible? We asked, how do we make computer systems meet the constraints of the power grid?” (10:44)
  • Prototypes on real servers proved the concept.

4. Grid Challenges and the Flexibility Opportunity

• Limitations of Current Grid Solutions
  • Issues: Renewable production fluctuates (solar peaks at noon, demand in evenings); nuclear is slow and expensive to scale; batteries are costly and may not be environmentally sustainable.
  • Data centers sometimes wait 5-7 years to connect to the grid, far outpacing “AI time,” where innovation cycles happen every 6 months.
• Real-world Example: Texas Blackout Risk
  • During a 2023 Texas heatwave, electricity demand spiked prices by 800%. Flexible data centers could have helped prevent emergency alerts and reduce costs.

5. AI as Both Problem and Solution

• AI Conducting the Grid
  • The variability and challenge of management (dynamic prices, unpredictable workloads, inconsistent regulations) outpaces static human decision-making.
  • AI itself is “probably the only thing smart enough to tame it”—becoming the “conductor” orchestrating harmony between the grid and AI workloads:

    “Imagine a data center, or a whole network of them as an orchestra...bringing a conductor, suddenly all that noise turns into music...AI can direct data center operation so that the data center can precisely match power constraints...” (13:27)

• Practical Impact
  • Coskun’s team has built software that slows, speeds up, pauses, or shifts workloads in real-time, always preserving user guarantees.

6. The Imperative to Act

• Dual Opportunity
  1. Make current data centers flexible to help prevent blackouts and reduce costs.
  2. Build future centers flexibly from the start, avoiding long connection wait times and allowing faster, cleaner AI adoption.
• Warning
  • Sticking with old paradigms “wastes renewable energy,” raises costs, and delays AI progress.
• Vision for the Future
  • “AI is already reshaping how we compute, but it could also reshape how we power the world...The real question is how much flexibility, resilience and clean power can AI unlock. If we are bold enough to rethink AI data centers, the very machines that now seem like a burden could be our greatest assets in building a sustainable AI future.” (15:05-15:30)

Notable Quotes & Memorable Moments

• On Transforming the Narrative:

  “So energy hog label seems well deserved. But that's only half the story. Here is the new view. These facilities are not just energy hungry brains. They can also be the muscles of the grid, flexing on demand.” (06:23)

• On Skepticism and Persistence:

  “Bold ideas require persistence, because change almost always looks impossible before it looks obvious.” (11:48)

• On Orchestrating Flexibility:

  “The very technology driving this unforeseen demand is also probably the only thing smart enough to tame it. AI can learn patterns, anticipate grid needs, and coordinate across data centers, across utilities, even nations in real time.” (13:05)

• On the Call to Action:

  “The question isn’t how much energy AI consumes. The real question is how much flexibility, resilience and clean power can AI unlock.” (15:05)

Important Timestamps

• 03:34 – Introduction to AI data centers' impact on the grid
• 06:23 – Shifting from “energy hogs” to flexible grid assets
• 08:56 – Early ideas on adaptive computing and initial resistance
• 10:44 – Reframing system design for grid constraints
• 12:10 – Challenges with traditional grid upgrades vs. data center agility
• 13:05 – AI as conductor of data center power flexibility
• 15:05 – 15:30 – Conclusion and vision for a sustainable AI future

Tone and Delivery

Ayșe Coskun’s delivery is inspiring, clear, and grounded in firsthand research experience. She combines technical insight with compelling storytelling, emphasizing persistence, innovation, and the urgent need for boldness in sustainable tech development.

Summary Takeaway

While AI data centers are often maligned for their growing electricity demands, Ayşé Coskun persuasively argues that—if we leverage their flexibility and use AI itself as a management tool—these centers can actually become allies in grid stabilization and clean energy adoption. The future requires bold rethinking, but it’s within reach—and could turn the very symbols of energy overuse into engines for resilience and sustainability.