Could AI Go Green?

Summary

Short Wave: Could AI Go Green?

Released May 9, 2025 | Host: Emily Kwong and Regina Barber | NPR

---

Introduction

In the episode titled "Could AI Go Green?" from NPR's Short Wave, hosts Emily Kwong and Regina Barber delve into the environmental challenges posed by the rapid expansion of artificial intelligence (AI). This episode, the second part of a miniseries, explores the significant energy demands of AI technologies, particularly large language models, and investigates innovative strategies aimed at reducing their ecological footprint.

---

Sasha Luccioni’s Personal Crisis and Mission

The episode opens with the compelling story of Sasha Luccioni, a computer scientist who experienced a profound climate anxiety that led her to leave her role as an AI researcher at Morgan Stanley in 2018.

“Essentially was getting more and more climate anxiety. I was really feeling this profound disconnect between my job and my values and the things that I cared about.” — Sasha Luccioni [01:52]

Motivated by her concerns, Sasha joined Hugging Face as the climate lead, dedicating her efforts to developing more sustainable AI models. Her journey underscores a growing movement within the tech community that seeks to balance AI innovation with environmental responsibility.

---

The Energy Footprint of AI

Emily and Regina highlight alarming projections regarding the energy consumption of data centers, which are the backbone of AI technologies.

By 2028, data centers in the United States alone could consume up to 12% of the nation’s electricity, equating to 580 terawatt hours ([02:32]). To put this into perspective:

“U.S. data centers alone could someday use a Canada size amount of energy.” — Regina Barber [02:46]

This surge is largely driven by the prevalence of large language models powering virtual assistants like ChatGPT, Microsoft Copilot, and Google Gemini. These models require extensive computational resources, leading to increased energy and water usage for cooling data centers.

---

Tech Giants’ Commitment to Nuclear Energy

In response to the escalating energy demands, major tech companies—Google, Meta, Microsoft, and Amazon—have set ambitious goals to achieve net-zero carbon emissions, with most aiming for 2030 and Amazon targeting 2040.

The discussion turns to three primary strategies these companies are employing to power their data centers sustainably:

1. Nuclear Power Expansion

   The tech giants are investing in nuclear energy to significantly boost global supply by 2050. For instance, Microsoft plans to revive the Three Mile Island nuclear plant in Pennsylvania.

   “Now you're just cooling the surface of whatever the cold plate is covering rather than just blowing air through the entire machine.” — Benjamin Lee [09:42]

   This shift marks a transformation where AI companies are stepping into roles traditionally held by energy firms, striving to secure a more sustainable energy future.

---

Innovations in Data Center Cooling

Beyond energy sourcing, improving the efficiency of cooling systems in data centers is crucial. The episode explores two main cooling innovations:

1. Free Air Cooling Systems

   This method involves designing data centers to utilize ambient cool air from the environment, eliminating the need for energy-intensive chillers.

2. Liquid Cooling Technologies

   Liquid cooling uses synthetic fluids to absorb and dissipate heat more efficiently than traditional air cooling. David Craig, the retired CEO of the company Isotope, explains:

   “We are always that kid who does touch the very hot ring on the cooker when our mum said don't. We are always the people who touch the wet paint.” — David Craig [15:26]

   Liquid cooling can reduce energy usage by up to 40% compared to air cooling and eliminates water consumption, making it a pivotal innovation for green data centers.

Additionally, the concept of district heating—where excess heat from data centers is redirected to warm nearby communities—demonstrates practical applications of these cooling technologies. For example, Google’s data center in Finland, Homina, supplies heat to 2,000 residents ([11:36]).

---

The Push for Smaller AI Models

While improving energy sources and cooling systems are essential, another critical strategy discussed is the development of smaller, more efficient AI models. Sasha Luccioni advocates for:

“Nowadays, more and more, I think companies especially are like, for our intents and purposes we want to do X, like whatever, summarize PDFs, but you don't need a general purpose model for that.” — Sasha Luccioni [13:39]

Smaller models require fewer computational resources, thereby reducing energy consumption. Companies like Deep Seq are pioneering this approach by creating AI models with fewer parameters and training requirements:

“The idea behind a mixture of experts is you don't need a single huge model with a trillion parameters to answer every possible question under the sun.” — Benjamin Lee [13:10]

Despite these innovations, large language models remain dominant, as evidenced by the development of smaller variants by tech giants like Meta and Microsoft.

---

Towards an Industry-Wide Energy Rating for AI Models

Sasha Luccioni proposes the creation of an industry-wide energy rating system for AI models, akin to the Energy Star ratings for appliances. However, she notes resistance from tech companies:

“There's like such a blanket ban on any kind of transparency because it could either like make you look bad, open you up for whatever legal action, or just kind of give people a sneak peek behind the curtain.” — Sasha Luccioni [14:48]

This lack of transparency hampers efforts to assess and improve the environmental impact of AI technologies systematically.

---

Conclusion: Balancing AI Advancement with Environmental Responsibility

Emily Kwong emphasizes the urgency of addressing AI's environmental impact, likening the current AI revolution to the discovery of electricity but with a known climate cost.

“Google CEO Sundar Pichai compared it to the discovery of electricity. Except unlike the people during the Industrial Revolution, we know that this has a climate cost.” — Emily Kwong [16:20]

Regina Barber adds a touch of humor while highlighting the impracticality of halting technological progress entirely:

“Where would we get our cat videos in the mail?” — Regina Barber [15:57]

The episode concludes on a hopeful note, suggesting that with continued innovation and conscious efforts, it is possible to steer AI development towards a more sustainable and environmentally friendly future.

---

Key Takeaways

• AI’s Growing Energy Demand: Large language models are significantly increasing data centers' energy consumption, projected to reach 12% of U.S. electricity usage by 2028.

• Tech Companies’ Strategies: Major tech firms are investing in nuclear energy and developing more efficient cooling systems to mitigate AI’s environmental impact.

• Innovation in Cooling Technologies: Liquid cooling and free air cooling systems offer substantial energy savings and reduce water usage in data centers.

• Advocacy for Smaller AI Models: Reducing the size and complexity of AI models can lead to lower energy consumption without sacrificing functionality.

• Need for Transparency: Implementing an energy rating system for AI models could drive industry-wide improvements, though resistance from tech companies poses a significant challenge.

---

Notable Quotes with Timestamps

• “Essentially was getting more and more climate anxiety...” — Sasha Luccioni [01:52]

• “U.S. data centers alone could someday use a Canada size amount of energy.” — Regina Barber [02:46]

• “Now you're just cooling the surface...” — Benjamin Lee [09:42]

• “We are always that kid who does touch the very hot ring...” — David Craig [15:26]

• “Nowadays, more and more, I think companies...” — Sasha Luccioni [13:39]

• “There's like such a blanket ban on any kind of transparency...” — Sasha Luccioni [14:48]

• “Google CEO Sundar Pichai compared it to the discovery of electricity...” — Emily Kwong [16:20]

• “Where would we get our cat videos in the mail?” — Regina Barber [15:57]

---

This episode of Short Wave provides a thorough examination of the environmental challenges posed by AI and the multifaceted approaches being explored to make AI technologies more sustainable. Through engaging storytelling and expert insights, Emily Kwong and Regina Barber illuminate the path toward a greener AI future.

Transcript

[Sponsor Announcer]

This message is from NPR sponsor, Spectrum Business. From one employee to thousands, Spectrum delivers connectivity solutions at a great value for any size business. And it's all backed by the Spectrum commitment. Learn [e-mail] business A quick note before.

[Regina Barber]

We start today's show. You may have heard that President Trump has issued an executive order seeking to block all federal funding to npr. This is the latest in a series of threats to media organizations across the country. The executive order is an affront to the First Amendment rights of public media organizations. It's also an affront to the First Amendment rights of the American people. NPR remains committed to serving the public. That's you. We lay out the facts and bring you stories that spark your curiosity that you won't find anywhere else. This is a pivotal moment. It's more important than ever that every supporter who can contribute comes together to pitch in as much as they are able to. Visit donate.NPR.org now to give. And if you already support us via NPR or another means, thank you. Your support means so much to us. Now more than ever, you help make NPR shows freely available to everyone. We are proud to do this work for you and with you.

[Emily Kwong]

You're listening to Shortwave from npr.

[Regina Barber]

Hey, short wavers. Regina Barber here with my co host, Emily Kwong with the second half of a miniseries. She reported on the environmental footprint of AI. Hey, Em.

[Emily Kwong]

Hi, Gina. So today I am bringing you a story of a personal crisis.

[Regina Barber]

It's very relatable. Go on.

[Emily Kwong]

Okay. So in 2018, computer scientist Sasha Luccioni took a new job, AI researcher for Morgan Stanley. She was excited to learn something new in the field of AI, but she couldn't shake this wor.

[Sasha Luccioni]

Essentially was getting more and more climate anxiety. I was really feeling this profound disconnect between my job and my values and the things that I cared about. And so essentially I was like, oh, I should quit my job and go plant trees. I should, you know, I should do something that's really making a difference in the world. And then my partner was like, well, you have a PhD in AI. Maybe you can use that.

[Emily Kwong]

So Sasha quit her job.

[Regina Barber]

Wow.

[Emily Kwong]

And she joined this growing movement to make AI more sustainable.

[Regina Barber]

Yeah, you were saying that, like, AI innovation was causing this, like, surge in energy and water use, like, to cool data centers. And the construction of those data centers was only going to increase.

[Emily Kwong]

Yes, some think exponentially. Gina. By 2028, Lawrence Berkeley National Laboratory forecasts that data centers could consume as much as 12% of the nation's electricity. That's 580 terawatt hours.

[Regina Barber]

Okay, can you give me like a different way to kind of think about how much that actually like the amount.

[Emily Kwong]

Of energy that Canada consumed two years ago?

[Regina Barber]

Okay, so U.S. data centers alone could someday use a Canada size amount of energy.

[Emily Kwong]

They could.

[Regina Barber]

Wow.

[Emily Kwong]

So Sasha is on a quest to find AI models that are smaller and use less energy. She is now the climate lead at Hugging Face, which is an online community for AI developers to share models and data sets.

[Regina Barber]

And a model is just like an AI program that's trained to take in data and like output data. Yes.

[Emily Kwong]

So virtual assistants such as ChatGPT, Microsoft Copilot, Google, Gemini, they are all powered by what's known as large language models. And Sasha, as she made quite plain in her 2023 TED Talk, is not a fan.

[Sasha Luccioni]

In recent years, we've seen AI models balloon in size because the current trend in AI is bigger is better. But please don't get me started on why that's the case.

[Regina Barber]

Wait, so I actually do want her to get started. Like, why are these big players all using these huge models?

[Emily Kwong]

I'm glad you asked.

[Regina Barber]

Thank you.

[Emily Kwong]

Because today on the show we're going to talk about why bigger isn't always better when it comes to generative AI. In part two of our series, we'll talk about how this big sprawling industry is looking to shrink its environmental footprint with everything from small models, clean energy, and a back to the future way of keeping data centers cool. I'm Emily Kwong.

[Regina Barber]

And I'm Regina Barber. You're listening to Short Wave, the science podcast from npr. Don't worry, you won't be lost if you haven't heard part one.

[Sponsor Announcer]

This message comes from Charles Schwab. When it comes to managing your wealth, Schwab gives you more choices like full service, wealth management and advice when you need it. You can also invest on your own and trade on thinkorswim. Visit schwab.com to learn more.

[State Farm Announcer]

This message comes from NPR sponsor State Farm. In the market for small business insurance, State Farm agents can help you create a personalized plan that fits your business needs and budget. Like a good neighbor, State Farm is there. Talk to your local agent today. This message comes from Schwab at Schwab. How you invest is your choice, not theirs. That's why when it comes to managing your wealth, Schwab gives you more choices. You can invest and trade on your own. Plus get advice and more comprehensive wealth solutions to help meet your unique needs. With award winning service, low costs and transparent advice, you can manage your wealth your way at Schwab. Visit schwab.com to learn more.

[Regina Barber]

Okay, Em, you've been talking with, like, four of the biggest tech companies. Google, Meta, Microsoft and Amazon, which I should say are like, all financial supporters of NPR.

[Emily Kwong]

It's true. Amazon also pays to distribute some of NPR's content.

[Regina Barber]

Right. And these four companies, like all, have ambitious goals for hitting net zero carbon emissions. Most by 2030. Amazon by 2040. How are they going to get there?

[Emily Kwong]

There are three paths, as far as I can tell. But before we talk about small AI models, you know what Sasha's describing. Let's talk about two solutions to make large language model computing more green, and that is more efficient. Data centers and nuclear power. What do you want to start with, Gina?

[Regina Barber]

I'm a physicist. Nuclear, obviously.

[Emily Kwong]

Of course. Of course, nuclear. Because Amazon Meta and Alphabet, which runs Google and made a big announcement in March. As reported by Straight Arrow News, three of the world's largest tech companies are promising to help triple global nuclear power supply by 2050. They're going to build new nuclear power plants and along with Microsoft, purchase nuclear energy.

[Regina Barber]

Okay.

[Emily Kwong]

And Microsoft plans to get its nuclear energy by reviving a plant in Pennsylvania.

[Regina Barber]

Yeah, our colleague Jeff Brumfield, he came on the show in December to talk about how Microsoft purchased Three Mile island. Like the site of a partial nuclear meltdown in 1979.

[Sponsor Announcer]

Yes.

[Emily Kwong]

Only one of the reactors melted down, by the way. The whole site was shut down in 2019, and now Microsoft wants to bring it back.

[Regina Barber]

Okay, so are AI companies turning into energy companies?

[Emily Kwong]

They are turning into energy movers and shakers for sure. But Jeff sees a discrepancy in this, you know, between the AI people and the nuclear energy people.

[Regina Barber]

Yeah.

[Emily Kwong]

Silicon Valley loves to go fast and break things. The nuclear industry has to move very.

[Regina Barber]

Very, very slowly because nothing can ever break.

[Emily Kwong]

Nuclear is also extremely expensive. Yes. And while solar and wind energy combined with batteries is quicker to build and more inexpensive than nuclear or gas power plants, it still takes time.

[Regina Barber]

I mean, like, do we need to move that quickly to grow AI?

[Emily Kwong]

Well, depends on who you ask. Kevin Miller, who runs global infrastructure at Amazon Web Services, says, yes, I think.

[Regina Barber]

You have to look at the world.

[Emily Kwong]

Around us and say, and we're moving.

[Regina Barber]

Towards a more digital economy overall.

[Emily Kwong]

And that is ultimately kind of the biggest driver for the need for data centers and cloud computing. But Sasha Luccioni, the computer scientist who we met earlier, feels this rush for AI is coming from industry, not from consumers.

[Sasha Luccioni]

It's unfair to say that users want more because users Aren't given the choice.

[Regina Barber]

Yeah. I mean, like, I hear Sasha here because, like, I'm a big fan of, like, AI's benefits. It's totally changed science and medicine and business and banking, all these things that affect our lives. Lives. But it does feel like opting out of AI is, like, becoming more and more difficult.

[Emily Kwong]

Absolutely, yes. And until nuclear power catches up with AI's energy demand, data centers will, for the foreseeable future, continue to use fossil fuel sources.

[Regina Barber]

Yeah.

[Emily Kwong]

So the question becomes, you know, is there a way to make data centers themselves more efficient? And the tech giants are trying through better hardware, better chips, and this really captured my attention. But more efficient cooling systems. So that's solution number two.

[Regina Barber]

I love a tech solution to a tech problem. What are some of these strategies?

[Emily Kwong]

Well, one method that's become quite popular is to design a data center to bring in cool air from outside the facility. No chilling required.

[Regina Barber]

So they just, like, pull in this cold air.

[Emily Kwong]

Yeah. This is what's known as a free air cooling system. And then there's a design paradigm that's getting a bit of buzz. Folks in the industry call it liquid cooling.

[Regina Barber]

Okay. And this is a different kind of liquid cooling evaporation we talked about in the first episode.

[Emily Kwong]

Yes. This does not water. Liquid cooling uses a special synthetic fluid that runs through the hottest parts of the server to take the heat away.

[Regina Barber]

Okay.

[Emily Kwong]

Or whole servers are immersed in this cool liquid bath.

[Regina Barber]

Okay. So the idea of, like, running coolant through, like, a car engine.

[Emily Kwong]

The very same. You can think of this, like, coolant, but for computers. Okay. Benjamin Lee, who studies computer architecture at the University of Pennsylvania, said this is just a much more efficient way to. To cool off a hot computer, because.

[Benjamin Lee]

Now you're just cooling the surface of whatever the cold plate is covering rather than just blowing air through the entire machine.

[Emily Kwong]

So I wanted to talk to someone who's trying to bring liquid cooling to the market, and I found this company called Isotope. David Craig is their recently retired CEO.

[David Craig]

I definitely come from the point of view that, you know, we literally have just one planet, and I cannot understand why anybody would want to do anything other than care for it.

[Emily Kwong]

David says the older way of cooling data centers, that daisy chain of moving heat with air and water, is just completely consumptive.

[Regina Barber]

Yeah.

[Emily Kwong]

And while he couldn't tell me which tech companies have struck agreements with Isotope.

[David Craig]

So there are a number of confidentiality clauses that sit around kind of customers.

[Emily Kwong]

Isotope has announced public partnerships with Hewlett Packard and Intel and Ashley Settle. A spokesperson at Meta told me that Meta anticipates some of its liquid cooling enabled data centers will be up and running by 2026.

[Regina Barber]

Wow. Okay. But em, because I'm a numbers person, like how much energy is being saved by liquid cooling versus, like air or water cooling?

[Emily Kwong]

Depends on the data center. You can say that the very best liquid cooling system uses about 40% less energy than a traditional air cooling system and it uses no water.

[David Craig]

And then what we're doing is we're capturing that heat in a closed water loop. So it's a bit like a domestic central heating system at that point. And in huge parts of the planet, particularly the north, we can return that heat and do useful things with it in much more intelligent ways.

[Emily Kwong]

David is talking about something called district heating. And that's where the heat from a data center, any data center, doesn't have to be liquid cooled is then diverted to a local neighborhood. And that is starting to happen at some data centers in Europe. Google has a data center in Finland that is providing heat to 2,000 people.

[Regina Barber]

That's so cool. I think I've actually read about this. Like, I think it's called Homina Data Center.

[Emily Kwong]

That's the one. Yeah. No, Homina does not use liquid cooling, but it is kind of a poster child for a green data center. Hana runs on 97% renewable energy and pumps in seawater from the Bay of Finland to keep cool.

[Regina Barber]

Wow, that's really cool. Literally. Okay, so Hana is just like one of these data centers, like out of thousands, right?

[Emily Kwong]

Yes. And this is the challenge. Most data centers are not situated by bodies of water in Northern Europe. Right. So I want to talk about a third and final innovation, and it's the one that the tech companies I spoke to were kind of quiet about.

[Regina Barber]

Oh, okay.

[Emily Kwong]

But the one that scientists and engineers outside the industry could not stop mentioning. And that is smaller AI models.

[Regina Barber]

I mean, of course, right?

[Emily Kwong]

One's good enough to complete a lot of the tasks we care about, but in a much less energy intensive way. So a third and final solution to AI's climate problem is just to use less AI. One kind of disruptor in this space is Deep Seq.

[Sponsor Announcer]

Right?

[Regina Barber]

That's the chatbot out of a company in China and it is claiming to use less energy.

[Emily Kwong]

Yes, I did reach out to Deepseek for comment. I didn't hear him back. But here's the thing, Gina. Large language models like ChatGPT are often trained with really large data sets. Deep Seq, on the other hand, appears to have been trained with fewer chips and consists of Smaller models that run fewer parameters. Benjamin Lee at UPENN says this is called a mixture of experts.

[Benjamin Lee]

The idea behind a mixture of expert is you don't need a single huge model with a trillion parameters to answer every possible question under the sun. Rather you would like to have a collection of experts, smaller models, and then you just sort of route the request of the right expert. And because each expert is so much smaller, it's going to cost less energy to invoke.

[Emily Kwong]

But deep seek, here's the thing about it, it's still a big general purpose model. And Sasha Luccione at Hugging Face wants to walk away from large models entirely.

[Sasha Luccioni]

Nowadays, more and more, I think companies especially are like, for our intents and purposes we want to do X, like whatever, summarize PDFs, but you don't need a general purpose model for that. You can use a model that's task specific and a lot smaller and a lot cheaper.

[Emily Kwong]

Basically, Sasha wants to see companies develop small language models, models that have far fewer parameters and are trained for a specific task.

[Regina Barber]

So like, are tech companies experimenting with this?

[Sponsor Announcer]

A few.

[Emily Kwong]

Okay, here's what I found. Last year, Meta announced a smaller quantized version of some of their models and Microsoft announced a family of small models called Phi 3. But honestly, in all of my conversations and emails with big tech, it's clear to me that large language models are here to stay. So if that's the case, Sasha has one last idea. An industry wide score for AI models.

[Regina Barber]

Okay, so like Energy Star was created to like rank the energy efficiencies of appliances. You see that little star on a lot of those appliances?

[Emily Kwong]

Yes, something like that for AI models. But at least according to Sasha, tech companies are not embracing a rating system.

[Sasha Luccioni]

There's like such a blanket ban on any kind of transparency because it could either like make you look bad, open you up for whatever legal action, or just kind of give people a sneak peek behind the curtain.

[Emily Kwong]

And as a science reporter for npr, my question was just, do we really need all of this computing power when we don't know how much it's costing us environmentally and when it could imperil our climate or goals? And David Craig, the recently retired CEO of Isotope, he chuckled when I said this and he's like, emily, you know, human nature is against us.

[David Craig]

We are always that kid who does touch the very hot ring on the cooker when our mum said don't. We are always the people who touch the wet paint. We're not good at learning until bad things happen to us. The truth is with data. You know, this stuff has just grown up in the background. People just haven't known about it.

[Regina Barber]

I mean, yeah, I mean, I certainly had no idea about all of this while I was growing up. All this was happening in the background. But em, you can't like disinvent the Internet, right?

[Emily Kwong]

Nor should we.

[Regina Barber]

Where would we get our cat videos in the mail?

[Emily Kwong]

There's something I think that we as consumers can think about. The AI revolution is fairly new. Google CEO Sundar Pichai compared it to the discovery of electricity. Except unlike the people during the Industrial Revolution, we know that this has a climate cost.

[Regina Barber]

Wow. Yep.

[Emily Kwong]

And there's still time to adjust how and how much we use AI.

[Regina Barber]

Em, thank you so much for bringing us this reporting.

[Emily Kwong]

Thanks, Gina.

[Regina Barber]

If you like this episode, you can check out part one of our series on AI and the environment. It's already out. This episode was produced by Hanichin, edited by our showrunner Rebecca Ramirez, and fact checked by Tyler Jones. Robert Rodriguez was the audio engineer. Special thanks to Brent Baughman, Johannes Durgi, and our incredible standards team. And Special thanks to TED Conferences, LLC.

[Emily Kwong]

Special thanks also to Julia Simon on NPR's Climate Desk. Beth Donovan is our senior director and Colin Campbell is our senior vice president of podcasting strategy. I'm Emily Kwong.

[Regina Barber]

And I'm Regina Barber. Thanks for listening to Shortwave, the science podcast from npr.

[Sponsor Announcer]

This message comes from Warby Parker. What makes a great pair of glasses at Warby Parker? It's all the invisible extras without the extra cost, like free adjustments for life. Find your [e-mail] or visit one of their hundreds of stores around the country. This message comes from Mint Mobile. Mint Mobile took what's wrong with wireless and made it right. They offer premium wireless plans for less and all plans include high speed data, unlimited talk and text, and nationwide coverage. See for [e-mail] Switch this message comes.

[State Farm Announcer]

From Ezcator, a business tool for food helping organizations order food for meetings and events from favorite restaurants, set up meal programs for their employees and manage food. Spend all in one [e-mail].