AI: Is It Ruining the Environment?

Summary

Science Vs — "AI: Is It Ruining the Environment?"

Hosted by Rose Rimler
Release Date: November 13, 2025

---

Episode Overview

In this episode of Science Vs, host Rose Rimler, sitting in for Wendy Zuckerman, investigates the environmental impact of artificial intelligence (AI), focusing on its energy and water usage. The show dissects the claim that AI data centers are "ruining the environment," examines popular memes and viral statistics, and contrasts AI's ecological footprint with other industries and habits. Through interviews with climate journalists, AI researchers, and sustainability-focused engineers, the episode explores facts, context, and misconceptions around AI's real cost to the planet.

---

Key Discussion Points & Insights

1. Why Is AI’s Environmental Impact in the Spotlight?

• Growing concerns about AI’s energy- and water-intensive data centers.
• Community protests against new data center construction over resource competition ([01:03]).
• Cultural meme: Using AI for trivial tasks is mocked as "ruining the planet for a f***ing Instagram post" ([01:32]).
• Rebuttals from tech companies note that per-user impact is small, and some critics argue it’s less than other behaviors (e.g., eating meat, flying) ([01:32]).

---

2. How Do AI Data Centers Differ from Regular Computing?

• AI utilizes GPUs (graphic processing units) instead of CPUs (central processing units), enabling massive parallel processing but at higher energy costs ([05:12], [06:09]).
• Memorable Analogy:
  • CPUs: "One paintball gun drawing a happy face pellet by pellet."
  • GPUs: "200 paintball guns firing at once to create the Mona Lisa."
  • — Rose Rimler ([05:57])

---

3. Quantifying the Energy Use of AI Models

Interviewees: James O'Donnell and Casey Crownheart, MIT Technology Review

• Real numbers for energy use are hard to come by—big AI providers are secretive ([07:47]; “Companies ... not willing to share the details ... so, you know, we weren't going to get it from them." — James O'Donnell [07:47]).
• MIT team tested open-source models themselves. Takeaways:
  • "What is AI's energy burden?... One of my biggest takeaways was just how much it depends."
    — Casey Crownheart ([08:58])
  • Different models, different sizes: Smaller models (8B parameters); larger ones up to 400B; leading products likely in the trillions ([10:44]).

How Much Energy Does a Prompt Use?

• Small language model ("Llama," 8B parameters):
  • ~114 Joules (about one-tenth of a second in a microwave) ([12:19])
• Huge model:
  • 8 seconds of microwave time per prompt
• Industry numbers:
  • Text prompt for major models: 1–2 seconds microwave time
  • "Probably on the order of zapping something in the microwave for less than 10 seconds." — Rose Rimler ([13:36])
• Image generation:
  • Slightly less or comparable to large text model output; ~5.5 seconds microwave ([14:39])
• Video generation:
  • Significantly more: "Over an hour in the microwave" for a 5-second, 16 fps clip ([15:57])

---

4. The Systemic Picture: AI’s Expanding Demand

• Daily usage: OpenAI reports 2.5 billion prompts per day ([17:21]).
• Organizational adoption: 78% of surveyed organizations use AI ([17:22]).
• Impact on grid:
  • Data center electricity use tripled between 2014 and 2023 (Lawrence Berkeley National Lab) ([17:56]).
  • By 2028, AI data centers could use as much power annually as a quarter of US households ([18:03])
    • "Imagine adding 25% more households to the US in 2028." — Rose Rimler ([18:23])
• Problem lies less with electricity itself, but that most grids still rely heavily on fossil fuels ([18:41]):
  • Only 9% US power is renewable; 1/3 petroleum, 1/3 natural gas, 8% coal ([19:10])
• Nuclear power considered, but new plants take decades—not scalable fast enough for demand ([20:04])

---

5. Water Use: Debunking the Viral "AI Is Drinking All Our Water" Meme

Guest Expert: Shao Le Ren, UC Riverside

• Viral claim: Every AI use "wastes a bottle of water" ([27:23])
• Reality (based on ChatGPT-3 for reference):
  • ~30 prompts exchanged = 0.5 liter (one bottle) of water ([28:34]).
  • Only ~12% is drinking water for cooling; remainder is water used at power plants (from rivers, lakes, not potable) ([29:06]).
  • Water used evaporates, but may not replenish the local supply—regional water stress matters ([26:26], [27:02]).
  • Total US data center water use: 0.3% of supply (roughly Rhode Island’s share) ([31:29]).

---

6. Regional vs. National Impact & Future Outlook

• Effects can be severe in small towns with under-built water infrastructure ([30:28]).
• Water and energy demand expected to double in coming years—but proportional national impact remains modest ([31:56]).
• Real opportunity for reduction lies in a greener energy grid and smarter AI model engineering ([21:47], [31:44]).

---

7. Personal and Societal Responsibility: Who Should Be Accountable?

• Even environmentally-conscious experts report they still use AI for convenience ([33:10]).
  • E.g.: Planning trips, editing text, technical research.
• “It's on the companies, it's on the government ... I'm not sure AI is the villain. I think the villain is our reprehensible and baffling inability to switch to renewable energy and to put any kind of real effort into getting off of fossil fuels.”
  — Rose Rimler ([34:16])
• Key reason AI gets blamed: Many don’t yet see social or personal value in it, compared to flying or eating meat; 61% of the US public perceives AI as having "more drawbacks than benefits." ([35:23])

---

Notable Quotes & Memorable Moments

• On Meme Culture and the “AI Water Bottle” Statistic:
  • "So that's a distortion of the message that we show in the paper."
    — Shao Le Ren ([28:28])
• On Model Complexity:
  • "It's like a switchboard that goes on for miles."
    — Rose Rimler ([10:50])
• On Individual vs. Systemic Impact:
  • "I'm not sure AI is the villain. I think the villain is our reprehensible and baffling inability to switch to renewable energy..."
    — Rose Rimler ([34:16])

---

Timestamps for Key Segments

• [00:33] — Framing the controversy: AI’s environmental downsides introduced
• [05:12] — How AI’s architecture (GPUs) differ from regular computing
• [07:00] — MIT Tech Review's journalists on measuring energy use
• [12:19] — Explaining energy use in "microwave seconds"
• [13:55] — Text vs. image vs. video energy comparison
• [17:21] — The scale: daily prompts, grid impact, and tripling of power usage
• [18:58] — Fossil fuel dependency: what powers the grid
• [24:23] — Profile: Shao Le Ren and introduction to water cooling in data centers
• [28:28] — Debunking the viral "one prompt = one bottle" claim
• [31:29] — Data centers = Rhode Island’s worth of national water use
• [33:10] — Do the experts personally use AI?
• [34:16] — "The real villain is fossil fuels"
• [35:23] — Public ambivalence: 61% see more drawbacks than benefits
• [36:47] — Has learning the facts changed personal AI use?

---

Conclusion & Takeaways

• Individual use of AI for text and images is relatively modest per interaction; the worry is in the exponential scaling as usage explodes.
• The environmental impact hinges less on AI itself, more on the source of energy grids. If powered by renewables, AI’s effect would shrink considerably.
• Water concerns are regional, and the viral "one email = one bottle" stat is a simplification.
• Biggest growth in harm is from increasing institutional adoption and demand—not memes or individual silly uses.
• Responsibility for managing AI’s footprint falls on governments and corporations to decarbonize energy and develop efficiency—individual restraint helps, but isn’t the solution.
• Public perception is skeptical—possibly because AI’s societal benefits are less tangible (yet) than those of established technologies or activities.

---

For more details and sources, check the episode transcript and show notes, including the MIT Technology Review article referenced throughout.

---

Quotations and key segments are labeled with timestamps and speaker attributions, directly reflecting the original tone and context of the discussion.

Transcript

[Rose Rimler]

Hi, I'm Rose Rimler. I'm filling in for Wendy Zuckerman. And this is Science Versus. This is the show that pits facts against filling the world with AI data centers. Today on the show AI and the Environment. Lately we've been hearing a lot about how power hungry AI is. AI uses a ton of electricity, straining the nation's aging power grid and creating more planet warming emissions and how thirsty it is.

[Shao Le Ren]

The amount of water that AI uses is astonishing.

[Rose Rimler]

Asking ChatGPT to write one email is the equivalent of pouring out an entire water bottle.

[Shao Le Ren]

One bottle of water.

[Rose Rimler]

Let that sink in. Let that sink in. And the major culprit here is the data centers. Warehouses full of computer servers that AI needs to function. And tech companies are trying to build more of these data centers, but people who live nearby are protesting them, often saying that they're going to compete for their electricity and use up their water.

[Shao Le Ren]

Go data center here. We do not want a data center.

[Eva Longoria]

Built in St. Charles City.

[Rose Rimler]

Because of all this around some corners of the Internet, using AI has become kind of a faux pas, especially if you use it for something silly.

[Blythe Terrell]

You are actively contributing to global warming.

[Rose Rimler]

And climate change all because you want.

[Blythe Terrell]

To Photoshop Chris Brown into your picture.

[Rose Rimler]

So next time you use AI to generate an image for a meme, think.

[Nordstrom Sponsor]

About the impact impact on the environment.

[Rose Rimler]

First, stop ruining your planet for a Instagram post. But on the flip side, we've got people pushing back against this idea. They say that these reports are skewed or misleading and that the impact of AI on the environment isn't nearly as bad as a bunch of other stuff we're already doing, like eating meat or taking international flights. In fact, recently some of the big AI companies have said that their products only use a tiny bit of power and a few drops of water for each prompt. So what's really going on here? Is AI actually ruining the planet? Or have the bots been framed? Because when it comes to AI and the environment, there's a lot of stop ruining your planet for a f Cking Instagram post. But then there's science. And that's coming up after the break.

[Nordstrom Sponsor]

This episode is brought to you by Nordstrom. Cool temps are here. Time to level up your wardrobe at Nordstrom. You'll find the best cold weather must haves, including thousands of styles under $100, shop head to toe cozy from faves like Ugg, All Saints, Nordstrom Skims the North Face and more. Plus free shipping, free returns and quick order pickup. Make it easy in stores or online. It's time to go shopping at Nordstrom.

[Spotify Portal Sponsor]

This episode is brought to you by Spotify Portal for Backstage. But you're wondering, what's Portal? Well, it's an internal developer portal built to improve developer experience and boost productivity. All software components are centralized. Documentation is automated and easy to maintain. New projects and components, just a few clicks. With your best practices already built in, think less friction, more innovation. Ready to double your productivity? Try Spotify [e-mail].

[Rose Rimler]

Welcome back. I'm Rose Rimler. I'm a senior producer at Science Versus, and I'm here with our editor, Blythe Terrell. Hi, Blythe.

[Blythe Terrell]

Hey, Rose.

[Rose Rimler]

Blythe, it seems like you're my AI buddy. I invite you to talk to me about controversies when it comes to AI.

[Blythe Terrell]

It's because I'm part robot.

[Rose Rimler]

I would say that of the team, you're the person who is most tuned into this idea about AI using up all the energy, using up all the water. You've been to this for a while.

[Blythe Terrell]

Yes. This is actually. I am one of those people who probably shared a meme, Rose, without knowing if it was true on the water use or the whatever. Like, I do remember seeing those memes and being like, is this true? And then. And then, honestly, for me, it did. It did make me take a step back from AI and be like, before I get involved in this, I. I do want to know the truth. Like, is this actually terrible for the environment? Is this actually terrible for the water? Because why would I want to integrate it into my life if it is?

[Rose Rimler]

Right. So the first question is, why do we think AI would use so much more energy than all the other stuff that we do in our digital lives? Just like messing around on the computer, posting on Instagram, watching Netflix. Like, this is all stuff that we do pretty routinely and don't think a lot about the, like, footprint of that behavior. Right.

[Blythe Terrell]

Like looking at pictures of Jeff Goldblum.

[Rose Rimler]

Yeah.

[Blythe Terrell]

How much energy is that using? For just as an example, hypothetically, Photoshopping.

[Rose Rimler]

Jeff Goldblum as your prom date.

[Blythe Terrell]

You know, I don't know who you might be talking about, Rose, but that sounds like a pretty good use of electricity and energy and water, no matter how much it takes.

[Rose Rimler]

So that kind of stuff also requires data centers and energy to run them. But the thing that's different about AI is that their servers are using a different kind of computer chip. So normal computing uses a cpu, but AI uses a gpu. And if you're familiar with video games, you might think of this as, like, a graphics card, but it's actually become the powerhouse behind machine learning. This is an extremely visual and potentially copyrighted analogy. But I was looking around on YouTube.

[Blythe Terrell]

I love a copyrighted analogy.

[Rose Rimler]

The Mythbusters guys, they did a demonstration of a CPU versus a GPU. And in their demonstration, they used paintball guns.

[Blythe Terrell]

Okay.

[Rose Rimler]

The CPU was, like, programming one paintball gun to draw a happy face with, like, one paintball pellet at a time, like, firing at a piece of paper on a wall.

[Blythe Terrell]

Okay?

[Rose Rimler]

That's a CPU. A GPU was, like, 200 paintball guns all bound together, making it like a mega, mega paintball gun. And one switch is hit, and they all fire at once. And the image that they create is the Mona Lisa.

[Blythe Terrell]

Oh, man, those guys are good Mythbusters.

[Rose Rimler]

So the point is that while CPUs are good at doing one task after another, the GPUs are good at doing a bunch of tasks at. And that requires a lot more energy.

[Blythe Terrell]

Okay. How much energy?

[Rose Rimler]

Are you ready for that? Yeah. Be ready. And I got a bit of an assist here. I talked to some journalists who have covered this stuff for years.

[James O'Donnell]

My name is James o'. Donnell. I'm a senior reporter for AI at MIT Technology Review.

[Casey Crownheart]

I'm Casey Crownheart. I'm a senior climate reporter at MIT Technology Review.

[Rose Rimler]

So James and Casey both report a lot on AI and energy use.

[Blythe Terrell]

Okay.

[Rose Rimler]

And about a year ago, they started a project trying to figure out, like, how much does an average query or prompt to say, chatgpt, how much energy does that use? And they were inspired to do that because they were seeing all these numbers out there floating around that just didn't seem all that reliable. Here's Casey.

[Casey Crownheart]

These kind of wild estimates of, you know, oh, a query to something like, ChatGPT uses this much water and this much energy, and isn't that so much? And so I think that that started to kind of get our gears turning and wondering, you know, is that right? How can we add all of this up? What does it. What does it all add up to?

[Rose Rimler]

So Casey and James looked around for the real number, but we learned very.

[James O'Donnell]

Quickly that it's not going to be so easy to know that number. Companies, they're not particularly willing to share the details of how much energy their AI models require to answer one question. And so, you know, we weren't going to get it from them.

[Rose Rimler]

And so what did they say when you reached out and asked?

[James O'Donnell]

They said, in so many words, no. So.

[Rose Rimler]

So James and Kasey went a different route. There are AI models that are not proprietary, anybody can use them, even download them, host them on their own computer, as long as they have the power to do that. These are open source models and you can kind of open the hood, poke and prod them. And so James and Casey teamed up with experts, including academics at the University of Michigan, to measure it themselves. So they ran a bunch of different prompts through an open source large language model called Llama, and then they were able to actually measure how much energy those requests required. And so they got some answers. Are you curious?

[Blythe Terrell]

Yes, I would love to know. Give me the answers.

[Rose Rimler]

Well, there's a range here I was.

[Casey Crownheart]

Really struck throughout this project of, I think we went in and I was looking for kind of one definitive answer. What is AI's energy burden? And I think that one of my biggest takeaways was just how much. It depends. It depends on the model, it depends what you're asking. And so there's just this really big range that was one of my biggest takeaways.

[Rose Rimler]

So, okay, you know, basically when. When people like us say I asked AI, you know, we kind of act like AI is this one thing, and it's totally not. There are all these different models and these models come in different sizes.

[Blythe Terrell]

So a model, Sorry, a model is like a ChatGPT or a Gemini or a Claude or whatever. Yeah.

[Rose Rimler]

And. And they're within chat, GPT, Gemini, Claude. There are multiple models within llama. There are multiple, and some are bigger, some are smaller. If you imagine that this AI model, imagine it's the command of a spaceship or actually my favorite is a switchboard with tons of knobs and dials. You can kind of imagine that's what these parameters are.

[James O'Donnell]

James says each of those knobs is helping the AI come up with a better answer. But also each of those knobs requires energy to operate.

[Rose Rimler]

So the smallest model that the team looked at for this analysis had 8 billion parameters. So. 8 billion, wow, that sounds big. The biggest one they looked at had 400 billion parameters. 400 billion. And when it comes to the big players here, we actually don't know how many parameters they have, but James said if he had to guess, it's, you.

[James O'Donnell]

Know, in the order of trillions.

[Rose Rimler]

Whoa.

[James O'Donnell]

Really big. Lot of knobs.

[Rose Rimler]

A lot of knobs. I mean, I'm imagining basically like a switchboard, but now I have to completely change it because it's like a switchboard that goes on for miles.

[James O'Donnell]

Yes, that's right.

[Blythe Terrell]

So these parameters you're talking about, which is like sort of the what Underpins the model.

[Rose Rimler]

I guess they are. Yeah, they're numbers, values. And the more of them that there are, the better the model is at learning patterns and making predictions, which is how large language models work.

[Blythe Terrell]

Okay. Okay, so say I have like one request and I pop it into a model with 8 billion parameters, and then I pop that same request into a model with like 400 billion parameters. That same request is going to use different amounts of energy based on the model that I'm using.

[Rose Rimler]

Yes. And actually we can move into. We don't even have to hypotheticalize here. I have real numbers for you.

[Nordstrom Sponsor]

Oh, nice.

[Rose Rimler]

Okay, so the smallest llama model that the team used, they fed in some prompts like teach me about quantum computing or suggest some travel tips, and then they measured, you know, how much energy that used the smallest model when it spit out an answer used, on average, 114Joul.

[Blythe Terrell]

Oh, great, great. That's very helpful. Thank you.

[Rose Rimler]

Are you being sarcastic? Do you want some other way to think about this?

[Blythe Terrell]

Yes, please give me something more concrete.

[Rose Rimler]

Well, I didn't, you know, it was James and Casey, so they came up with something for some context. One thing they converted these energy units into is a fun new type of unit called microwave seconds.

[Casey Crownheart]

I love the microwave seconds unit. It's so much more relatable than joules or watt hours.

[Blythe Terrell]

Casey gets me.

[Rose Rimler]

All right, so 114 Joules is roughly a tenth of a second in a microwave.

[Blythe Terrell]

Okay, so that's one query small model. A tenth of a second in the microwave.

[Rose Rimler]

Yeah.

[Blythe Terrell]

Okay.

[Rose Rimler]

The biggest model, which was 50 times bigger, that was like zapping something in a microwave for eight seconds.

[Blythe Terrell]

Oh, so that's. The biggest model was still only for one query. The biggest model was still only eight seconds. Okay, that doesn't even get my rice remotely hot.

[Rose Rimler]

Right? I mean, and after James and Casey published their article, frustratingly for them, OpenAI and Google did release a little bit of information on how much energy their text prompts use on average. And what they said suggests that a text query is equivalent to one or two seconds in the microwave. Okay, so basically what we can tell you is like a text prompt to a large language model is probably on the order of zapping something in the microwave for less than 10 seconds.

[Blythe Terrell]

Okay.

[Rose Rimler]

And then for images, this is a different kind of machine learning, but it also uses a fair amount of energy. And I would have assumed that this image making thing is inherently more energy sucking than text making, but as it turns out, that is not necessarily the case.

[James O'Donnell]

Here's James, if you have a really big, large language model that's generating text and answers. It may actually use more energy than generating an image. And that was kind of counterintuitive for me because, you know, you think about like these AI models that come up with fantastical images that we've all seen over the past few years, and it just seems like such an intense process to kind of create that from scratch.

[Rose Rimler]

Yeah, because it always takes longer too, than getting your text back.

[James O'Donnell]

Yeah, exactly. But what we found was that if you have a really large text model, it has so many parameters, so it has so many knobs and dials, that it actually can use up more energy than generating, you know, certain types of images.

[Rose Rimler]

They found that making an image was like running a microwave for five and a half seconds.

[Blythe Terrell]

So, like a big language model can be like 8 seconds of microwave time. So that that's a little less. Okay, right, I'm with you.

[Rose Rimler]

And you know, for all this stuff, if you don't like microwave time, you could also think about it in light bulb time. So it's like running an LED light bulb for somewhere between 10 seconds and two minutes.

[Blythe Terrell]

So I guess, Rose, what this maybe tells me is that if I wanted to make my Jeff Goldblum prom picture with AI, that is a slightly less energy intensive process than perhaps using AI to write romantic Jeff Goldblum fanfiction.

[Rose Rimler]

Possibly if you use a really big model to write your Jeff Goldblum fanfiction.

[Blythe Terrell]

Obviously I would need a very large model for this work, Rose. Okay, got it.

[Rose Rimler]

And then there's video generation. This might not surprise you to hear that that used the most energy. So the team, what they did was they looked at an open source video generation model and they just made like a crap video. It was 16 frames a second, 5 seconds long. They compare it to like the quality of a silent film era type film.

[Blythe Terrell]

Okay.

[Rose Rimler]

And that one, that would be the.

[Casey Crownheart]

Equivalent of over an hour in the microwave.

[Rose Rimler]

I don't think I've ever microwaved anything for an hour.

[Casey Crownheart]

I don't think I have either. A long time in the microwave, for sure.

[Rose Rimler]

That one scares me more because I'm seeing a lot of AI generated videos out there.

[Blythe Terrell]

Yeah, that's what's huge right now. And getting bigger. Right. There's a ton of. What is it? Sora?

[Rose Rimler]

Yeah, Sora is big right now.

[Blythe Terrell]

Yeah.

[Rose Rimler]

And they can look incredibly realistic. Right, right. We don't know if that's also using up as much energy. We asked OpenAI, which makes Sora, and they didn't give us Any information on Sora's energy use? And James and Casey didn't want to speculate. It's a different model, but it's probably using a fair amount of electricity. I think that's safe to assume.

[Blythe Terrell]

So, I mean. But. Okay, so what we have so far is all about, like, individual use.

[Rose Rimler]

Yeah, but I want to.

[Blythe Terrell]

But what I want to know, though, is, like, obviously lots of us are doing this. Lots of us are using this. Like, what is the impact? If you, like, add it all up, if you, like, scoop up all the AI use that we're doing, like, what do we know about that?

[Rose Rimler]

Right. And it's interesting, like, on an individual level, certainly, like the texting and image generation stuff, they're not that crazy energy intensive. But that doesn't leave AI off the hook, because when you do zoom out and to answer your question, really adds up fast because OpenAI says it receives two and a half billion prompts per day from people around the world.

[Blythe Terrell]

Wow.

[Rose Rimler]

And AI in general is getting integrated into all these institutions, which I think a lot of us are noticing. In fact, one survey of a variety of organizations around the world found that 78% of them are now using AI to some extent. That is a lot. And so nerds have looked at how much electricity is going to data centers to see if AI has made an impact. And they saw that from 2014 to 2023, the electricity consumption of data centers tripled. That's according to a report from the Lawrence Berkeley National Laboratories.

[Blythe Terrell]

Oh, wow. So, like, in this. And that's the AI period. Like, that's like sort of.

[Rose Rimler]

That is like the. Yeah, basically the age of AI, like, taking off.

[Blythe Terrell]

Taking off.

[Rose Rimler]

And the energy suck is expected to keep sucking up more and more. One analysis predicts that by 2028, AI data centers will use as much electricity as. As a quarter of US households use per year.

[Blythe Terrell]

Wow.

[Rose Rimler]

Imagine adding 25% more households to the US in 2028. That's what the prediction is that these AI data centers are gonna use up.

[Blythe Terrell]

That doesn't sound good.

[Rose Rimler]

Well, I mean, Casey, who's a climate reporter, she was like, you know, it's not the electricity per se that's the problem here.

[Casey Crownheart]

That's kind of the crucial thing that I like to bring up and really harp on is that, you know, if we had abundant solar and wind power and batteries, you know, we might be less concerned about some of this. This energy demand. But the reality is that grids around the world are still largely relying on fossil fuels.

[Rose Rimler]

So it's not Good. Right now in the U.S. only 9% of the country's power comes from renewable sources. It's still mostly fossil fuels that we use to power our electric grid.

[Blythe Terrell]

Right.

[Rose Rimler]

A third of our energy comes from petroleum. A third comes from natural gas, which is another fossil fuel. Both are greenhouse gas emitters.

[Blythe Terrell]

And coal.

[Rose Rimler]

Coal is in the mix too. It's 8%.

[Blythe Terrell]

So just a lot of this energy is, is dirty.

[Casey Crownheart]

Mm.

[Rose Rimler]

And of course, there are other countries with cleaner energy grids than the US but more than half of the data centers for the world are here in.

[Blythe Terrell]

The U.S. well, and you know, I feel like the headlines, some of the headlines I've seen around this rows have been like, related to nuclear energy. Because there were headlines a while back that one of these companies was gonna reopen Three Mile island, which is this nuclear plant that was shut down because of an accident. And so there was talk of like, that being reopened. And like, you know, really a lot of these companies being very interested in what's going on with nuclear. So it does make me wonder, could nuclear help if we can get that ramped up?

[Rose Rimler]

I asked Casey about that and she was like, the thing about nuclear reopening or building a new nuclear plant, it takes so long. The last nuclear plant that we built in the US took 15 years to complete. Yeah. And companies are just not going to wait for that to happen. And they're not, they're not waiting for it. I mean, look at X AI they brought in gas burning generators to run their data center in Tennessee. Right.

[Blythe Terrell]

Okay. Okay. Well, that sucks.

[Rose Rimler]

Yeah. So I reached out to XAI and I didn't hear back. I also contacted Google and Anthropic just to ask about all this stuff that we've been talking about. I didn't get answers from them by our deadline. OpenAI did get back to me. They mostly pointed me to stuff that's already publicly available, open letters and blog posts, that kind of thing, talking about their energy use and how they see that in the future. And basically what OpenAI is saying is that they want to work with the government to act, add capacity to the grid. And they say that they want that energy to come from all kinds of sources, including renewables.

[Blythe Terrell]

Uh huh.

[Rose Rimler]

Okay. And just overall, I will say there might be some changes coming for the positive. So the energy that AI requires to answer your query or make your image or your video, that could be going down because a lot of the tech companies are trying to make their models more efficient. One way they're doing that is by turning off some of the parameters that we talked about earlier when they don't necessarily need them to answer a particular question or do a task. So that's like shrinking the switchboard essentially as needed.

[Blythe Terrell]

So there is some evidence that like the tech companies are like trying to adjust to make this thing.

[Rose Rimler]

Yes, it might get better.

[Blythe Terrell]

Okay, so that's energy Rose. But. But I know there's another piece to this. Yeah. What about water? What is going on with water?

[Rose Rimler]

Right. So we're going to talk about that after the break.

[Nordstrom Sponsor]

This episode is brought to you by Nordstrom. Oh, what fun. Nordstrom has tons of gifts under $100 for all your favorite people all in one place. Like beauty and grooming sets, Ugg gifts, jewelry and toys, neat ideas. Check out gifts from Ugg, Skims, Diptyque, Free people, Stanley and more. Plus, explore their amazing gift shop in stores and online. Free gift finding help, free shipping and order pickup. Make it all easy at Nordstrom.

[Rose Rimler]

So good, so good, so good.

[Nordstrom Sponsor]

Score holiday gifts. Everyone wants for way less at your Nordstrom rack store. Save on Ugg, Nike, Rag and Bone, Vince Frame, Kurt Geiger, London and more.

[Rose Rimler]

Because there's always something new.

[Nordstrom Sponsor]

I'm giving all the gifts this year with that extra 5% off when I use my Nordstrom credit card. Santa who join the Nordiclub at Nordstrom Rack to unlock our best deals. It's easy. Big gifts, big perks. That's why you rack. This episode is brought to you by Nordstrom. Oh, what fun. Nordstrom has tons of gifts under $100 for all your favorite people all in one place. Like beauty and grooming sets, Ugg gifts, jewelry and toys. Need ideas? Check out gifts from Ugg, Skims, Diptyque, Free people, Stanley and more. Plus explore their amazing gift shop in stores and online. Free gift finding help, free shipping and order pickup. Make it all easy at Nordstrom.

[Eva Longoria]

Hi, it's Eva Longoria. And let's be real. After 40, we should ask for more from our skincare. I swear by Revitalift triple power moisturizer by l' Oreal Paris. With vitamin C pro, retinol and hyaluronic acid. It reduces my wrinkles, firms and brightens. And it's not a procedure. It's just a hard working moisturizer.

[Spotify Portal Sponsor]

Revitalift triple power moisturizer by l' Oreal Paris. Grab it today and fragrance free or with SPF 30.

[Rose Rimler]

Available at your local Walmart. Welcome back, I'm Rose Rimler. I'm here with Blythe Terrell.

[Blythe Terrell]

Hello.

[Rose Rimler]

So let's talk about water and AI. And if you want to talk about water and AI, you call up Shao Le Ren. He is a professor at UC Riverside. He's actually in the computer engineering department. But he focuses on sustainability. Most people in his field look at, you know, energy, greenhouse gases, like we were just talking about. But Chalet has kind of forged his own path because he's thought about conserving water for just a lot of his life.

[Shao Le Ren]

I spent my first few years in a small town back in China. We just had access to fresh water, drinking water, water for half an hour each day. During those half an hour, we had to use a big bucket to collect the water and use it for the rest of the day. So in my memory, I never thought water is something unlimited. It's just. It's a finite resource. Mm.

[Blythe Terrell]

He's never taken it for granted.

[Rose Rimler]

Right. And so one reason AI uses a lot of water is something that you've probably heard before. The data centers get really hot because they're running all these fancy chips doing all this computation like we were talking about earlier. And so these buildings, they often use a cooling tower that uses water to cool everything down.

[Shao Le Ren]

Just like our human bodies. We sweat and we feel cooler. For data center, if you use water evaporation, you can take away the heat very naturally, very efficiently.

[Rose Rimler]

And where do they get that water from?

[Shao Le Ren]

Most typically it's from the municipal water infrastructure system.

[Rose Rimler]

So the same as where, if I lived there, if I were to turn my tampon.

[Shao Le Ren]

Yeah.

[Blythe Terrell]

So they get water where everyone gets water from the faucet, basically.

[Rose Rimler]

And the reason for that is they want like clean, filtered water. Because if there was salt or minerals or like gunk in it, then it could gum up this system, basically.

[Blythe Terrell]

Mm. Okay.

[Rose Rimler]

So as the water cools the data centers, it, you know, it evaporates away. It evaporates. And if I remember my, like kindergarten, you know, the water cycle, when water evaporates, it eventually comes back as rain. Right. So why do we need to worry about this?

[Shao Le Ren]

So the evaporated water. Yeah, it still stays within our global water cycle system. It doesn't go away from the earth, but still when the water will be coming back and where it will be coming back, that's highly uncertain. And it's very unevenly distributed across the globe. So due to the long term climate change, we're seeing more and more uneven distribution of the water resources. So essentially, the wetter regions are getting wetter and drier regions are getting drier.

[Rose Rimler]

So even if the water is evaporated in Say Arizona. That doesn't mean it'll come back as rain in Arizona. At least not anytime soon.

[Shao Le Ren]

You're correct.

[Blythe Terrell]

Okay, okay. So the argument is it's using a bunch of water, it's drawing it out from where everyone else is getting their water. And it's not. That's not necessary. And it's not necessarily going to be replenished that easily.

[Rose Rimler]

Well, yeah, yeah, it's going to evaporate the drinking water in Tucson, and that water might next show up as a flood in Shanghai. Right, okay, so let's talk about how much water is actually getting used here. Shalay and his team, they went down this rabbit hole fairly recently and they published a paper that kind of went viral. In fact, a lot of people turn their results into a meme, basically saying that every time you use AI, they'll say in different ways. Like every time you chat with ChatGPT, every time you write an email with AI, you're consuming a bottle of water. Have you seen this, Blythe?

[Blythe Terrell]

Yes, yes. This was one of the memes I first saw and shared without evidence.

[Rose Rimler]

I've seen videos of people filming themselves with a nice, beautiful fresh bottle of water from the store, opening it up and pouring it down the drain and saying this is what you're doing when you use AI or someone dressed up and pretending to be AI dressed as a robot, and they're just guzzling water. But that's not quite accurate.

[Shao Le Ren]

Yeah, so that's a distortion of the message that we show in the paper.

[Blythe Terrell]

A distortion.

[Rose Rimler]

Here's what they actually found. So they found that if you have a back and forth conversation with. In this case, the model they looked at was ChatGPT3. That's a slightly older model, but if you have a back and forth with ChatGPT three medium length messages, if you go back and forth for on average about 30 times, that uses up essentially the volume of a bottle of water. A half liter of water.

[Casey Crownheart]

Mm.

[Blythe Terrell]

Okay, so it's like a. It's a decent conversation that gets you to that half liter.

[Rose Rimler]

Yeah. And that's where the meme comes from. So it's not super duper wrong. But what they're getting wrong or misunderstanding is that the fresh drinking water that's used to cool the data center, that's actually only a small part of this calculation. So out of this half liter of water that we're talking about, only about 12% of it is drinking water that's used directly by the data center for cooling. Oh, and the rest of it is non Potable water from elsewhere. It's drawn out of rivers, lakes, whatever. It's used in the process of making electricity. So that brings us back again to the power plants. You know that old chestnut.

[Blythe Terrell]

Okay, wait, wait. So it's talking. So some of this is drinking water, but some of this is like.

[Rose Rimler]

But most of it is not.

[Blythe Terrell]

But most of it's not. But I mean, but still, like, that's water in the environment could eventually become drinking water. Right? So, like, why does. So why does that distinction actually really matter?

[Rose Rimler]

Well, if you think the data center moving into your town is a threat because it's gonna turn on a bigger tap than yours, that's not quite right. And I asked Chalet about that. Do you think it's possible that a town will accept a data center and it uses up all the town's water? Essentially? Like, you live next door to a data center, you turn your tap and no water comes out?

[Shao Le Ren]

I think in certain towns it could be possible, but in most towns, I think the US Infrastructure tends to be. At least for the water infrastructure, it's. They should be able to have the capacity available for data centers.

[Rose Rimler]

He said that the biggest problems here might be likely to happen in really small towns with really old or limited water infrastructure. Okay, so when I see people talking about how data centers are using up water, I think, like, we might be ignoring the bigger issue here, which is the water used by power plants. And by the way, if we had more wind and solar on the grid, the water use would go down. But anyway, as of right now, overall, taking into account the water used by power plants and the water used for cooling, we know that Data centers consume 0.3% of the nation's water supply. I asked Shelley about this. I don't know what to make of that. Is that a lot or is that a little 0.3%?

[Shao Le Ren]

So it's roughly the same amount of total public water supply in Rhode Island. So whether this 0.3% is high or not. Not. I would say it's modest. It's not that much.

[Rose Rimler]

Brings up the question, should we be letting Rhode island use all that water? I mean, what has Rhode island done for anyone else lately?

[Blythe Terrell]

You know, it's. You know, finally the podcast is getting around to that question, which I've also had for years.

[Rose Rimler]

What is the point of Rhode Island? Yeah, I mean, and the water used for the data centers for the power generation and cooling is projected to go up. It's actually expected to double in the next few years. But ultimately, Chalet and another Expert I spoke to said that whether or not this becomes a problem is a regional question. It makes more sense to be granular about this. Like, is the water being taken from an area that doesn't have the capacity? You just can't paint with a broad brush here.

[Blythe Terrell]

So complicated, I guess, is where we so often land. Okay, so taking all this together, Rose, where do you land? Like, how evil is AI when it comes to the environment?

[Rose Rimler]

I asked all of our guests basically that same question. I kind of put it in terms of, like, well, do you personally use AI knowing about all these environmental impacts? Because these are people. All these people care a lot about the environment and these issues. And all of them, Casey, Chalet, James, they all said that, yes, they do still use AI.

[Casey Crownheart]

I'm awful at planning trips, so asking for an itinerary for going on a road trip or something, that's. I found that that's really helpful.

[Shao Le Ren]

I use it to polish my text writing to help me answer some questions. And also my students use AI generate paper summaries.

[James O'Donnell]

So I've used AI for technical things like how to do certain repairs on my bike. But I've also used it for seeing what people have said on a certain topic, like hikes in New England with the best views.

[Rose Rimler]

But everybody agreed that we should be thoughtful about how we use it, given this energy and water requirement as well.

[Blythe Terrell]

So it's annoying because part of me is like, you know, the companies that make this and that are using this, like, they're and that are, like, using it for their products and services that I'm using, like, they're the ones who I want to think about their AI use. Right. I want them to be thinking about whether they really need to use this or not. And I want them to be thinking about that in the context of energy use, water use, climate change. Right. Like, that's my dream, right? Yes.

[Rose Rimler]

It's on the companies, it's on the government. I mean, I think that the. My takeaway here is that, like, I'm not sure AI is the villain. I think the villain is our reprehensible and baffling inability to switch to renewable energy and to put any kind of real effort into getting off of fossil fuels.

[Blythe Terrell]

Right.

[Rose Rimler]

It's the same enemy we've been fighting for 50 years or whatever.

[Blythe Terrell]

Right? Right.

[Rose Rimler]

Also, I think that one reason AI is getting people riled up, as opposed to, like, those old climate offenders, flying, eating meat, you know, that kind of thing, is people see the value in the trade off of the environmental impact of something like taking a flight or eating a burger. There's an obvious benefit to those things with AI. Yes, some people have found it really useful, but a lot of people haven't, and they just don't think it has much value at all. In fact, one Survey found that 61% of people in the US think that AI has more drawbacks than it has benefits.

[Blythe Terrell]

Okay. So more than half of us are just like, no, we hate this. Yeah, no, thank you. Exactly. Okay.

[Rose Rimler]

I think that's one reason AI is our current villain, when in fact, I think the villain is. I think that's like a nostril on the larger villain, which is the evil monster that is keeping us glued to fossil fuels.

[Blythe Terrell]

Right, the nostril. Okay, I appreciate that picture. Okay, So I do want to know one last thing, though. Has learning this and digging into all of this AI and energy and water stuff, has it changed how you use AI?

[Rose Rimler]

Yeah, a little bit. Also, I think the novelty is wearing off a bit. And I was never, like, using it a of ton, but I don't know, I was asking people about, like, what's some stupid stuff that you've seen generated by AI and you're like, oh, my God, that wasn't worth the energy. And I thought of my own playing around with it, and I was like, remember that time I had AI generate an image of my boyfriend cuddling with my cat because my cat doesn't like him? So I was like, oh, this is what it would be like if you guys got along, you know, And I sent it to him and I was like, I don't think I would do that again. I don't think that was worth the energy.

[Blythe Terrell]

So it has changed a little bit. How you would make that value assessment kind of. Is this. Is this a using. Is using AI for this thing going to, like, add value? Is it actually really useful for this or.

[Rose Rimler]

Or could I just glue salmon to his fingers and then the cat would actually maybe come over?

[Blythe Terrell]

Yeah. You know? Yes. Rose, let's go back to the.

[Rose Rimler]

Let's go back to basics.

[Blythe Terrell]

Let's go back to the basics of gluing Sam into our boyfriend's fingers to get our cat to like him. Yep.

[Rose Rimler]

That's science versus thanks, Bly.

[Blythe Terrell]

Thanks, Rose. Oh, and while we're here, how many citations are in this week's episode?

[Rose Rimler]

There are 66 citations.

[Blythe Terrell]

Where can people find them?

[Rose Rimler]

They can find them in our transcript. The link to the transcript is in our show notes. Also in our show notes, we'll put a link to the article that James and Casey wrote for MIT Technology Review. It's really good. People should go read it. And people should also check out our Instagram. We've got some interesting stuff there. Maybe even a little Jeff Goldblum content for you.

[Blythe Terrell]

Give the people what they want.

[Rose Rimler]

Exactly.

[Casey Crownheart]

Great.

[Nordstrom Sponsor]

Love it.

[Rose Rimler]

This episode was produced by Rose Rimler and Blythe Turrell with help from Meryl Horne and Michelle Dang. We're edited by Blythe Turrell, fact checking by Diane Kelly, Mix and sound design by Bobby Lord Music written by Emma Munger so Wiley, Peter Leonard, Bumi Hidaka and Bobby Lord. Thanks to all the researchers we reached out to, including Professor Melissa Scanlon, and special thanks to Andrew Pouliot and Jesse Rimler. Science Versus is a Spotify Studios original. Listen for free on Spotify or wherever you get your podcasts. Follow us and tap the bell for new episode notifications. We'll fact you soon.