Pod Insights Stuff You Should Know Data Centers: Can't Live With Em, Can't Live Without Em About this episode Data centers are the backbone of our digital world. All the sites you visit and social media you post happen there. A boom in huge data centers to support AI is underway, and they’re taking a big toll on the environment and life for the people around them.
See omnystudio.com/listener https://omnystudio.com/listener for privacy information.
Transcript Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.
0:00 Speaker 00 This is an iHeart Podcast. Guaranteed Human.
0:04 Josh Clark? Buy smart, drive for less and trade for more. That's Toyota affordability and it starts with a great deal on an off-road ready Tacoma or half ton Tundra, two of Toyota's longest lasting trucks, combining power for work, tech for adventures and style in between. And they're built on a reputation for quality and reliability, which is why Toyota trucks are known for holding their value, which could mean big savings when it's time for your next Toyota. So get a great deal today and a great value tomorrow. Shop toyota.com for more. Toyota, let's go places.
0:36 Speaker 00 Welcome to Stuff You Should Know, a production of iHeart Radio.
0:46 Josh Clark? Hey, and welcome to the podcast. I'm Josh and there's Chuck and Jerry's here too. And we've got our pocket protectors and tape on the bridge of our glasses. And this is Stuff You Should Know.
0:58 Speaker 02 Nice work. Did you just hear something?
1:02 Josh Clark? No.
1:03 Speaker 02 Oh, weird.
1:03 Josh Clark? I heard you say, nice work.
1:05 Speaker 02 Yeah.
1:05 Josh Clark? And then stop abruptly.
1:06 Speaker 02 Well, I stopped abruptly because I thought I heard a little digital glitch.
1:11 Josh Clark? Oh, no, I didn't hear anything.
1:14 Speaker 02 I might be losing my mind then.
1:16 Josh Clark? I, you know, I'm curious whether we end up editing this out or not. Any other podcast on the planet would edit that out without even thinking about it. But there's like a 50% chance it'll stay in with us.
1:27 Speaker 02 Uh, I mean, this is why we didn't get that Golden Globe nomination.
1:30 Josh Clark? That's right.
1:31 Speaker 02 This kind of classic Stuff You Should Know unprofessionality.
1:34 Josh Clark? That's exactly right. That and the Italian accent, the mispronunciations, there's a whole laundry list.
1:40 Speaker 02 Yeah.
1:40 Josh Clark? That's okay, Chuck. I think we're, we're golden regardless.
1:44 Speaker 02 Agreed.
1:46 Josh Clark? So we're talking about data centers, which I had a very, very rough idea about, but actually, no, I knew that they existed essentially and that they were
1:55 Speaker 02 Yeah.
1:55 Josh Clark? becoming a problem with the rise of AI.
1:58 Speaker 02 Yes.
1:59 Josh Clark? But that was about it. How about you? Are you, are you data, data center philiak?
2:07 Speaker 02 No, you know me, I'm not, I'm not super technology minded, so I don't know a lot about this stuff. I remember walking by our server room back in the day when we were at Ponce City Market and seeing our colleague Izzy in there, hard at work.
2:21 Josh Clark? Yeah.
2:21 Speaker 02 And when that door was unlocked and open, hearing the whir of the, of the servers and the cooling machines and, you know, that, that's a, on a smaller scale, that's a data center.
2:32 Josh Clark? Absolutely. A hundred percent that's a data center. It was also a great place to curl up and take a nap in the middle of the day.
2:39 Speaker 02 The warmth of the server.
2:40 Josh Clark? Yeah. And the whir puts you right to sleep.
2:44 Speaker 02 Yeah.
2:45 Josh Clark? So, uh, yeah, that definitely counts as a data server. If you have like one of those, um, little home networking setups and like a closet in your house, data center.
2:54 Speaker 02 Sure.
2:54 Josh Clark? Technically, the PC is a data center. Anywhere you can store and access data, that's technically a data center. And you're like, well, that's stupid. Why'd you even say that? Josh did squibbling, that's quotidian. Shut up and get on with data centers. Whoa, whoa, whoa. First of all, don't use the S word. And then secondly. Yes. I, wait, what did I say? That's the Q word?
3:17 Speaker 02 Quotidian.
3:19 Josh Clark? Oh.
3:20 Speaker 02 No, that's KW.
3:21 Josh Clark? Right.
3:24 Josh Clark? So the reason that I bring that up though, Chuck, is because that technically is part of the progression of data centers.
3:30 Speaker 02 Yeah.
3:31 Josh Clark? Probably goes without saying, but it's evolved along with computing and as computing's kind of gotten bigger and bigger, the need to, to store and access more data has gotten bigger. So much so Chuck, that just wrap your head around this one. In 2024, just over a year ago.
3:48 Speaker 02 Yeah.
3:48 Josh Clark? We used 150 zettabytes of data. That's what we consumed. And consuming is anything from making a video and uploading it to TikTok or putting a post up on Instagram. It's, um, browsing a website, it's buying a song from iTunes, it's doing web analytics, it's buying something with your American Express card. All of that is data consumption and we consumed 150 zettabytes of data in 2024.
4:19 Speaker 02 Yeah, I don't even know how many Big Macs that is. I know you name dropped a lot of brands. It should be like the movies where every time you even just say like buy something on your Amex, the bank account grows by like $10.
4:32 Josh Clark? I agree wholeheartedly. I agree that Amex should do that. Amex? There's 20 bucks, I'll split it with you.
4:40 Speaker 02 Wow, you just bought lunch in 1997.
4:43 Josh Clark? That's right.
4:45 Josh Clark? So just real quick, a zettabyte Chuck is a trillion gigabytes.
4:48 Speaker 02 Okay.
4:48 Josh Clark? And we consumed, so we consumed 150 trillion gigabytes.
4:52 Speaker 02 Wow.
4:52 Josh Clark? That was 2024.
4:54 Speaker 02 That's worldwide, right?
4:55 Josh Clark? Yes, yeah, that's worldwide. In 2010, we consumed two zettabytes.
5:00 Speaker 02 Jeez.
5:01 Josh Clark? Yeah, so it's growing exponentially, which means that data centers are growing exponentially. And now they're about to just blow up, like truffle up essentially from, from, you know, this kind of calm plateau that they'd reached, it's about to just go in hyperdrive.
5:19 Speaker 02 Yeah, and actively is, and we're gonna, we're gonna get to some startling statistics later on in the episode, but, Kyle helped us out with this, our writer over in the UK.
5:29 Josh Clark? He did a fantastic job.
5:31 Speaker 02 He did a really good job, and there's gonna be some UK specific things in here, because Kyle's always keen, as they say, to throw that stuff in there.
5:37 Josh Clark? Yeah, for sure. Kyle likes to pepper those in.
5:40 Speaker 02 Yeah, of course.
5:42 Speaker 02 And he's not barred from doing so. We allow it.
5:47 Speaker 02 So, since Kyle, you know, is frequenter of the Wayback Machine, as are all the wonderful writers that we use, they all have the keys to the car essentially.
5:59 Speaker 02 He jumped in the Wayback Machine to sort of give us a little bit of a timeline on data centers and a bit on, you know, mainframes and PCs.
6:05 Josh Clark? He also left all of his used tea bags in there too. I don't know if you noticed.
6:08 Speaker 02 Oh, it's fine. You know, you can throw those back in some hot water and they do just a little bit weaker tea.
6:13 Josh Clark? Well, if you put like five of them together, it's like one.
6:16 Speaker 02 Yeah, and Kyle, I mean, that thing was full.
6:18 Josh Clark? It really was. He drinks a lot of tea.
6:21 Speaker 02 He really does. So if you want to talk about the earliest data centers that you could kind of call maybe a data, a data center, they were, you know, computers, they were electronic computers. Most of this stuff that we're gonna talk about early on was military in use. And as you'll see, even the first, when we talk about the UK one that was supposedly not military, they even loaned it to the military, which was kind of interesting. But these things were built with, you know, state of the art technology at the time, which meant vacuum tubes and, you know, manual switches and plugs and things like that. And the first thing that we can really talk about as the first programmable electric digital computer was the Colossus. And as we'll see, Elon Musk has now stolen that for his own purposes, that name.
7:08 Speaker 02 Probably because of this, I would imagine, but it was at Bletchley Park, of course, during World War II. And they were trying to, you know, crack into Hitler's messages at the time. And these things were huge. And kind of to me, the thing that stood out about Colossus, which is a neat little factoid, is that where Colossus was at Bletchley Park at Block H, it is now the National Museum of Computing.
7:31 Josh Clark? I want to go to that so bad. When we do that, that UK Europe tour next year.
7:36 Speaker 02 Oh.
7:37 Josh Clark? We got to go to that together, okay?
7:39 Speaker 02 Okay, are we doing that next year?
7:41 Josh Clark? That's what we were talking about.
7:42 Speaker 02 All right.
7:43 Josh Clark? We kind of already have promised it. We have to now, we're locked in the punch.
7:47 Speaker 02 Why is my voice so high then?
7:49 Josh Clark? I don't know. You practicing for the Alps?
7:52 Speaker 02 Yeah, that's right. No, that'd be a lot of fun. I'd love to go to that.
7:56 Josh Clark? So that was Colossus. Another one about the same time was the ENIAC, Electrical Numerical Integrator and Computer. So that's a quality acronym.
8:05 Speaker 02 Yeah.
8:07 Josh Clark? And it was the first general purpose electronic computer. And here's the thing. This is technically not data storage yet. It's data processing.
8:17 Speaker 02 Right.
8:17 Josh Clark? But these things, Colossus, ENIAC, you walked up to them and you said, what's the trajectory of this missile if I fire it from here? And ENIAC would go beep, bop, boop, boop. And then say like, whatever a trajectory is described in.
8:32 Speaker 02 Sure.
8:32 Josh Clark? Or Colossus, you'd be like, what is Hilter saying here to Goebbels? And the Colossus would say, Hilter is saying that he's a big fan of Goebbels' work, but he's suspicious that the rest of the world doesn't like either of them. And that was it. After that, you'd be like, hey, what was the last answer? And they'd be like, what's an answer?
8:53 Speaker 02 Yeah, you gotta just tell me what's going on with this Hilter business.
8:56 Josh Clark? You don't remember from our Art Mysteries of the Art World? That how stuff works already go-
9:03 Speaker 02 Oh, did it say Hilter?
9:04 Josh Clark? The title of that section was, "Did Hilter do these paintings?"
9:07 Speaker 02 Oh my god, that's a, that's a deep cut. I did not remember that.
9:10 Josh Clark? Yes, and I think it still says that on that, um, that listicle.
9:15 Speaker 02 Oh, probably.
9:15 Josh Clark? Yeah.
9:16 Speaker 02 Yeah, I can only hope.
9:17 Josh Clark? It's gotta be Hilter forever.
9:19 Speaker 02 All right, so, uh, we go into mainframes at this point, and this is, uh, like the 1950s, basically, when companies, um, could actually have their own computer. It wasn't just the military. Uh, these were the old punch card computers, and they were called mainframes. It wasn't made up for this term. Uh, mainframes were originally described, uh, were describing like what you would house telecommunications equipment and maybe some other sciency stuff.
9:42 Josh Clark? Mm-hmm.
9:43 Speaker 02 Uh, but it was referencing literally the cabinets that held this technology. And it became known as just, you know, it kinda took over when the computers, uh, world started using it as computer only.
9:53 Josh Clark? Yeah, but again, this is, um, like, you're a company, and this is where you store and process all of your data, and it's in this one room, but it's not going anywhere else. It's not for anybody else. And you have to physically be in the room to get your answer, process whatever data you're looking for. When the PC came along, and then the Macintosh came along, they took that thing and just made it very small so you could put it on all of your employees' desks. And now they had, like I was saying before, their own little data center right there. So if you said like, "Hey, what's the, um, uh, I need to know the Q4 reports," they'd say, "Go to Debbie's desk."
10:33 Josh Clark? Debbie's the one who's got that on her computer. And you would go over there and say, "Debbie, what's the Q4 report?" And Debbie would give it to you, right? There was no connectivity, but you could still, like, do a lot more stuff than you could when you had a mainframe.
10:48 Speaker 02 Yeah, for sure.
10:50 Josh Clark? Which makes mainframes feel like really outdated, but it turns out they're like totally still in use today.
10:55 Speaker 02 Oh yeah, absolutely. Uh, I do wanna jump back in time a little bit because I did, I promised talk of, uh, lending the military, uh, basically your equipment. And that's what happened in 1951. There was a tea shop chain. In the UK, I don't know if it's still around, Lyons, L Y O N S. And they were the very first company in the world that used a mainframe. It was called the Leo, the L E O. And it was, you know, like what you would think. They've handled like payroll and stock management and stuff like that. But there wasn't a lot for it to do at a T shop chain, except for those couple of things. And so they calculated missile trajectories like you were talking about for the Ministry of Defense.
11:34 Josh Clark? Exactly. And that actually kind of helped establish like a, I guess, a pay schedule or how people charged for data centers to come.
11:45 Speaker 02 Right.
11:45 Josh Clark? It was, you, they, like, you would charge them for the time that they used it, or you could lease it for a month. And that really started to come around when IBM got in the game. They became like the mainframe leader in the 50s, the early 50s. I think they had a unit that you could lease for $16,000 per month. That's in 1952 money. And then as the things, as like the processors got better and smaller and faster, that price came down dramatically. And then finally, in the 60s, they released the IBM System 360, which not only got Apollo 11 to the moon and back.
12:26 Josh Clark? It is a, it appears in an episode of Mad Men, apparently.
12:30 Speaker 02 Oh, really?
12:31 Josh Clark? Yeah.
12:32 Speaker 02 You didn't see that, though, right?
12:33 Josh Clark? No, I never did. I just saw a reference to it on the internet.
12:37 Speaker 02 And you knew it was the show?
12:38 Josh Clark? Yeah. But they, like, you should look up pictures of it. It's like those giant burnt orange cabinets with reel to reel magnetic tape. It's just, they're, they're cool to look at, yeah.
12:48 Speaker 02 Yeah, I remember we've referenced the movie War Games from our childhood in the 80s a lot. And the, the Whopper from War Games was, that was, you know, at that age, to see the Whopper in action and to see Matthew, I almost said Matthew Modine, Matthew Broderick, hanging up his handheld telephone receiver onto a modem to talk to the school computer. It was mind blowing.
13:12 Josh Clark? That, that, yeah, that phone in the modem made just a really big impression on me.
13:16 Speaker 02 Yeah, and a cool sound.
13:19 Speaker 02 Beep, bop, boop, boop. Should we take a break?
13:23 Josh Clark? Wait, let me talk about mainframes today, though, because I just want to give a little, I don't know if a shout out is the right term, but they are still around because they're so reliable, because they're so secure. You can make it so that there's information on those things that you, again, have to be physically present in the room to access.
13:41 Speaker 02 Sure.
13:41 Josh Clark? You can put all sorts of different layers of security. So if you're like Visa, or you're a healthcare company, or you're the Census Bureau, you're probably still using a mainframe because you're protecting information as tightly as you can. But those things are also super fast and can hold huge volumes of computation at once.
14:02 Speaker 02 Yeah, or a non-Golden Globe nominated podcast.
14:06 Josh Clark? Yeah, we've got our own mainframe.
14:08 Speaker 02 Yeah.
14:09 Josh Clark? We've got our IBM 360.
14:11 Speaker 02 That's right. What year was that from again, the 360?
14:14 Josh Clark? 64.
14:16 Speaker 02 Yeah, yeah, that's the one.
14:18 Speaker 02 I'm just making sure we didn't have the 65, because that was notoriously buggy.
14:21 Josh Clark? Yeah.
14:22 Speaker 02 All right, we can take that break now, and we're gonna jump out of the Wayback Machine and venture into the modern world, right after this. Beep, bop, boop, boop.
14:47 Speaker 03 This mood check is brought to you by State Farm. Porque tu bienestar también merece protección. ¿Qué tal, mi gente? Honey German checking in contigo. Take a second. Literalmente, un momentico para ti. ¿How's your corazón feeling today? If it's a little tired, a little oof, you're not alone. Have you eaten algo rico? Taken a deep breath? Maybe even turned off your phone for cinco minutitos? I promise, the memes will still be there. Sometimes self-care isn't candles and meditation. It's just saying no. It's drinking a little bit of agua. It's letting yourself rest sin pena. You don't have to prove anything today. Just be, breathe. Estás haciendo lo mejor que puedes. And yes, that's enough.
15:31 Speaker 03 Like a good neighbor, State Farm is there.
15:46 Speaker 02 All right, so we are out of the Wayback Machine. We're making that, we're combining all those old tea bags and making some still somewhat weaker tea.
15:54 Josh Clark? It really works, though.
15:55 Speaker 02 Yeah, it's not too bad. It's a combination of Earl Grey and chamomile, all kinds of fun stuff, but not too bad. And now we're going to talk a little bit about when things started to ramp up because it kind of happened in fits and starts. And one of the biggest, I guess, would it be a fit or a start, was the internet. Because once the internet came along, every business in the world started using it. And so, all of a sudden, you had to have a lot more data storage and bigger data centers and bigger server rooms in your companies, which was, you know, a pretty good thing at the time. After the dot com bust, there were a lot of casualties of that growth. But then things kind of, you know, the ship kind of righted itself.
16:37 Josh Clark? Yeah, and because it was like accessible to basically every business now, like you didn't have to buy a mainframe. You could, you could lease space on someone else's mainframe, like you're the Ministry of Defense or something all of a sudden. So that led to this huge proliferation. That gave the, the foundation for web commerce, e-commerce. That's what they, they used to call it. That's an old timey term now. But it created the ability for e-commerce to start and flourish. So it, this data centers scaling up to meet the needs of the internet and then to kind of give people all sorts of new space and room to come up with new stuff. That's where the digital economy came from right there.
17:19 Speaker 02 Yeah, yeah, all of a sudden you could hop on Webvan and order a sack of groceries.
17:24 Josh Clark? Mm-hmm. I have a friend who was all in about that.
17:30 Speaker 02 Webvan?
17:30 Josh Clark? Yeah.
17:31 Speaker 02 Yeah, I think I had a friend who was pretty heavily invested in Webvan.
17:36 Josh Clark? It's, but clearly it was ahead of its time. I mean, it's, uh, let's see, it's a Postmates and...
17:44 Speaker 02 Well, there's a lot, several of them now that have succeeded.
17:46 Josh Clark? We'll name them, we'll get 10 bucks each.
17:48 Speaker 02 Well.
17:50 Speaker 02 Well, you just got 10 for Postmates and you gotta split that with me, you know?
17:53 Josh Clark? I will.
17:54 Speaker 02 All right.
17:56 Speaker 02 Cloud computing was the next big jump when cloud computing came around in the early 2000s. Or do they call that the early aughts?
18:03 Josh Clark? I do.
18:04 Speaker 02 Okay. I thought I'd heard that come from your mouth, but that is when, you know, that was the real game changer because things were still, I mean, when cloud computing came along, people thought of it.
18:17 Speaker 02 If they didn't look too hard into it, they thought it was just, you know, floating up in the ether somewhere.
18:21 Josh Clark? Yeah.
18:21 Speaker 02 It's, it's still being stored on stuff. It's just not being stored locally. So all of a sudden, things were just going somewhere else for someone else to worry about all that storage. And more importantly, they could, they could link everything together and store a lot of stuff from a bunch of different people.
18:38 Josh Clark? Right. So now you have data centers, not just available to somebody like a huge bank or something like that, or the government initially, then a bank, yeah, or a tea shop. And then to e-commerce businesses now, it's available to you and me. So it's really hard to remember back because the world has changed so much. But Chuck, like 2008, 2009, they were giving us like VPN little things that, that, that like, you could go home and work and like you would, it would never work. I'd never understood how to make it work.
19:13 Speaker 02 Yeah.
19:13 Josh Clark? But that was like the very beginning of how you could take your work home with you and work from home and do things remotely like we can now, like it's nothing. But this led to the rise of businesses like, like Dropbox, right? So Dropbox goes to Amazon Web Services and say, hey, we want to buy a bunch of your cloud, right? Which means that they're going to use a bunch of like different servers and different data centers all over the place. And then Dropbox turns around to you and says, hey, if you give me $19.95 a month, you can have one terabyte of data.
19:46 Speaker 02 Right.
19:47 Josh Clark? You can consume one terabyte of data, right? And then hopefully you don't use all of that. So they don't have to pay Amazon Web Services for stuff they didn't use. But you're paying that $19.95 a month rather, whether you use that whole terabyte or not. It's a pretty smart business model. Would not exist at all if the cloud didn't, didn't exist.
20:08 Speaker 02 Yeah, it's funny as you were talking about 2008 and how quaint that is now. That's the year we started the show.
20:14 Josh Clark? I know, I know.
20:15 Speaker 02 Isn't that crazy to think about?
20:16 Josh Clark? It really is, but imagine like, like, working from home at that time. It was just, you didn't. You couldn't.
20:23 Speaker 02 It was kind of great. You went home and you homed.
20:26 Josh Clark? That's exactly right. That was a big difference. Yeah, I remember.
20:29 Speaker 02 But these data centers have now come together in such a big way now that they, the largest ones are called hyperscale. And they host more than 5,000 servers, like servers, not, not individual person's data. It's like, it's incredible how big is, how big they've gotten. Google's, and we'll go over some of the kind of square footage and then later talk about the elephant in the room, which is energy and water usage. But Google's first data center was built in 2006, just about two years before Stuff You Should Know launched. And this was in Oregon. And they are still expanding that thing beyond 1.3 million square feet. Meanwhile, in China, they're like, hold my tea, I guess.
21:12 Speaker 02 Because China Telecom has a 10.7 million square foot data center in Inner Mongolia.
21:19 Josh Clark? It's 250 acres.
21:22 Speaker 02 That is a warehouse full of whirring servers heating up and being cooled.
21:29 Josh Clark? Yeah, which is a big problem for any data center, it turns out. But the, the whole, the whole expansion, this jump starting in 2017, thanks to cloud computing. Because again, cloud computing just means all your stuff isn't on one server and one data center. It's broken up into pieces and spread all over the place. That's the cloud. That's basically it. Even though it's way more advanced and intricate than that, that's like, all you really need to know for the purposes of this episode, right? It, it led to a huge jump, a huge need in data centers. And it also expanded all the stuff we can do now. And COVID actually gave it another bump.
22:06 Speaker 02 Yeah.
22:06 Josh Clark? Made building a data center a very economically attractive thing to do if you had the money. Because remote working finally, finally established itself as like, no, we're doing this. Stop calling us back to the office.
22:22 Speaker 02 Yeah, which is what they're doing now.
22:23 Josh Clark? I know, I know. I hope it doesn't work because I remember when when all of that started and everybody was so nervous, like, management was all so nervous that people were just gonna totally, like, mess around and everything. It just, it didn't happen. I don't know anybody who's even been, like, gotten a talking to, let alone been fired for for just messing around at home. As a matter of fact, like you were saying, it just makes you work more.
22:49 Speaker 02 Yeah, I mean, do you know how many times in our old offices I would see Jonathan Strickland just wandering aimlessly through the office chatting with people?
22:55 Josh Clark? Yes, I do, because he would chat with me a lot.
22:59 Speaker 02 He did. He did that in in front of God and everybody, as they say, right there in the office. So I can't imagine what happened with him at home.
23:06 Josh Clark? Yeah, like it was a sorority mixer or something.
23:11 Speaker 02 Uh, we love Strickland. He's still around everyone, by the way. He uh, he retired from tech stuff, but he's still with the company, which is great.
23:16 Josh Clark? We love Strickland.
23:18 Speaker 02 Uh, all right. So now we're on to AI data centers. And that was the, I mean, to call it a game changer is is seems quaint compared to the rise of cloud computing and everything, because it is off to the races in a way that seemingly cannot be stopped. The genie has left the bottle, as they say. Uh, starting in 2022, when ChatGPT was released by OpenAI, all of a sudden, the need for these data centers became exponentially greater in size and the speed at which they need these things built. Because AI requires a ton of computing power to operate.
23:58 Josh Clark? So much so that they don't even use the standard, um, what's called the compute machines. So compute is like all of the processing power, the networking, all of that stuff. And traditionally with a computer that's done on a CPU, right? That's how that's how all of this gets done, right? Everything else is infrastructure. The CPU is doing all of the work. Those are so like, they still work. They like, most data centers are running on CPUs. For AI, just not fast enough. They use GPUs, graphic processing units, which are associated with video games for most people. Right? You need a good graphics card to run your video game, I guess.
24:40 Josh Clark? But, but they, the reason that for AI data centers that they use GPUs is because they're really good at parallel processing. They can run a bunch of different operations at once. So you're like, cool, you just throw a GPU in the data center and you can run an AI. No, you need hundreds of thousands of these things strung together. And instead of like a CPU running like a couple of servers or something like that at a data center. All of them are strung together to form one giant supercomputer that the AI operates on.
25:13 Speaker 02 Yeah, like ChatGPT itself was trained on 20,000 of these GPUs. A GPU, you know, the sort of the biggest name in the game. There's a couple, but the biggest one obviously is the Nvidia. But the Nvidia H100, that is the standard right now. If you look this thing up, it, it fits in your hand. It's not like some gigantic thing. 20,000 of them linked together, or 100,000 of them linked together. Who knows how many, you know, hundreds of thousands are eventually going to be linked together to end the world.
25:46 Josh Clark? Right.
25:48 Speaker 02 That's where all the power comes from, like you were saying. But it's, you know, it's just a, a little rectangular handheld thing. That's like, oh, that looks like something that maybe came out of a computer. And Nvidia is, what did their stock jumped over a couple of years? Like 900% over 2023 and 2024, something like that?
26:08 Josh Clark? Yeah, 900% increase.
26:12 Speaker 02 Yeah, and we'll talk about why all of this is super scary and dangerous because it really is.
26:18 Josh Clark? Well, yeah, if you want a really good explanation of this about, and like you said, how many GPUs you string together before we end the world. Nate Sures and Eliezer Yudkowsky in that book I keep referencing that I think everybody should read. If anyone builds it, everyone dies about the current state of AI. They talk about this in depth, but in a really understandable way. It's really fascinating. But that's essentially one of the things they say is like, we keep stringing together tens and tens of thousands more GPUs. That just makes this supercomputer smarter and smarter and more capable. And eventually, what's gonna happen? We're gonna reach some point potentially where we just put that extra last GPU in there.
27:01 Josh Clark? And all of a sudden, the balance is tipped and the thing becomes super intelligent.
27:05 Speaker 02 That's right. Also, a time for me, because you're always too shy to, to plug the end of the world with Josh Clark. Your fantastic limited series of which AI is one of the central focuses or one of, was it eight things?
27:19 Josh Clark? 10. Well, there was 10 episodes.
27:21 Speaker 02 10 episodes, right.
27:22 Josh Clark? Thanks, baby.
27:23 Speaker 02 Well, but one of the episodes was just like you talking about Jamie Buffett records and.
27:27 Josh Clark? That's right.
27:29 Speaker 02 He had to lighten the mood.
27:30 Josh Clark? Yep.
27:32 Speaker 02 Should we take a break or should we keep going for a minute? Let's keep going for a minute.
27:36 Josh Clark? Okay.
27:36 Speaker 02 Because you talked about investment and, you know, if you have the money to open one of these things. And that's what these tech companies are doing, like to, perhaps their great peril at some point. We'll see. Microsoft has invested $88 billion in data centers just in 2025. Amazon has pledged over the next 15 years, $150 billion. And Google and Meta together are about, you know, not working together, but they are expected to spend about $750 billion just on equipment over the next two years. And Stanley Morgan says.
28:11 Josh Clark? Morgan Stanley.
28:12 Speaker 02 Did I say Stanley Morgan? I think we should leave that in there.
28:15 Josh Clark? Okay.
28:16 Speaker 02 Stanley says, hey, you know, guys.
28:19 Josh Clark? Maybe you meant like Stanley comma Morgan.
28:22 Speaker 02 Yeah, Stanley comma Morgan. Over five years between 2025 and 2030, Morgan Stanley says about $3 trillion is going to be spent just on the data centers. About, I mean, half of which is the hardware and half of which is just building these things.
28:39 Josh Clark? Yeah, just in what the next four years?
28:43 Speaker 02 Yeah.
28:44 Josh Clark? So think about it, if you're Nvidia and you're the industry leader for GPUs and everybody's like, we're going to spend $1.5 trillion on this, on the infrastructure and the GPUs. You're looking pretty good down the road.
28:58 Speaker 02 Yeah, for sure. And you know, they're doing this because there's a demand right now for use at least, because things like OpenAI and other AI creators are using them like crazy. But these companies are also using them for their own AI research.
29:12 Josh Clark? Right, yeah. So like XAI has that Colossus machine that you were talking about earlier, which is 200,000 GPUs strung together. I'm not sure if it's fully online yet.
29:25 Speaker 02 In Memphis, Tennessee.
29:27 Josh Clark? Yeah, and it's just for that. It's they're not doing any, they're not calculating missile trajectories for the Ministry of Defense or anything like that. Like, it's just for that AI. And yeah, I think Meta is doing the same thing. OpenAI, I don't think is building their own because they're so in cahoots with Microsoft. I think they run their stuff on Microsoft's data centers, but yeah, if you have AI essentially right now, which means like God and everybody, you probably have your own data center dedicated to it.
29:58 Speaker 02 Yeah, I, and this isn't some, some moral stand I'm taking by saying that I have never used AI. And trust me, I know that every part of my life is now touched by AI, so I am inadvertently using it.
30:11 Josh Clark? Touched by an AI.
30:12 Speaker 02 That's right. But I've never, I've never used like, you know, chatbots or, or large language models or anything like that, just mainly because I'm just, I'm fine. Doing things like they are for now and not in a, in a Luddite sort of way.
30:29 Josh Clark? Right.
30:29 Speaker 02 I just, uh, everything's going along great for me and my job and how I live my life. So I just, I don't have a need for it.
30:35 Josh Clark? I do the same thing and I think also both of us are like, if somebody else wants to do it the other way, that's fine. Like, we're certainly not gonna criticize them or be curmudgeonly about it, or say that, you know, that's stupid.
30:48 Speaker 02 Right, but as you'll see, you know, and again, this isn't, um, yucking someone's yum, but everyone should know what they're a part of and that's part of what the episode is about.
30:58 Josh Clark? That's right.
30:59 Speaker 02 You know?
31:00 Josh Clark? Yeah, no, I totally do. Uh, before we take a break, I think, uh, it's, it's a small kind of side issue, but it's worth pointing out that it sucks. Um, because these Nvidia chips are so in demand from these massive companies, it has driven the price for just the average Nvidia graphics card sky high. So if you're a gamer and you're like trying to improve your system, like you pay way more than you used to for the same graphics card that you could have bought for like a quarter of the price, you know, a couple of years ago.
31:36 Speaker 02 Yeah, and I wasn't even looking like, I, I didn't even know that you could just buy, this is how little I know about all this before this was like, could you just buy a Nvidia GPU? But I was just researching the size and like, what do these things look like? And it, you know, it's, one was on eBay for $20,000 and I was like, oh my God, I didn't, I didn't know that was the deal.
31:55 Josh Clark? Is that right?
31:56 Speaker 02 Yeah, and I don't know if that's accurate. I don't, I don't know anything about it, so I could easily be corrected on all this, but, uh, that's what the internet told me.
32:04 Josh Clark? Okay, well, the internet never lies.
32:08 Speaker 02 Oh, one thing before we break real quick, cause we did promise a little UK specific stuff and I don't wanna short shrift our Brit listeners or Kyle. The UK is right now like the third largest nation for data centers. The US is first, I think Germany is second. And they signed what was called the tech prosperity deal with the giants, the tech giants of the United States. And right now, Microsoft has announced a $30 billion investment in UK data centers. And I think like 100 new AI data centers are planned in the UK at this point, moving forward?
32:41 Josh Clark? Yeah, and I saw there's at least one in Wales that's being smartly done. They like took an old radiator factory plant campus and they're revitalizing that as a AI data center. So it does sound like I get why the UK is doing it, but there's a lot of people in the UK and elsewhere who are like, these are not, this is not a good investment for local governments or even national governments. There's a big problem with all this. Like, there is a AI boom going on. Data centers are just one part of it. Like, people are throwing money at AI like it's 1999. And a lot of people are like, there's another, it's not a dot com bubble this time, but it's an AI bubble.
33:24 Speaker 02 Yeah.
33:25 Josh Clark? One of the reasons why some people are pointing to it as an AI bubble is that there's not, it's just not clear how much money is going to be made from AI and when that's going to start.
33:38 Speaker 02 Yeah.
33:39 Josh Clark? I think the Financial Times called OpenAI a money pit with a website on top.
33:44 Speaker 02 Yeah, not great.
33:45 Josh Clark? No, because people are just pumping money into this stuff, but they're not getting, they're not seeing results from it, not yet. It's not necessarily a bad bet that AI is going to completely revolutionize the world and like revolutionize economies and going to make some people a lot of money. But there's just no clear path to it right now, which makes some people nervous.
34:07 Speaker 02 Yeah, there's about 5%, just 5% of pilot AI programs right now in business secure returns on their investments, you know, like they make the money.
34:19 Josh Clark? But Stanley Morgan is predicting revenues of a trillion dollars by 2028.
34:26 Speaker 02 That's what they're saying. I mean, we'll see. Nvidia, I mean, Kyle's also keen to point out that the, there's sort of a circular economy within all this going on that's a little bit like troubling maybe because Nvidia is investing in OpenAI, but that depends on their purchase of those Nvidia chips. So, you know, everyone from You know, just people who are smarter than us as far as this stuff goes are warning people right down to the IMF, the International Monetary Fund, are flashing the warning signs, saying like, this could be, you know, it could make a trillion dollars by 2028, or it could like wreck the global economy.
35:04 Josh Clark? Yeah, for sure. Yeah, we have no idea. Although I have seen people argue against it that say like, this is nothing like, yeah, a lot of these AI companies are probably overinflated, but it's nothing like it was with like the 2008 meltdown or the dot com bubble. Like, this is, we're a lot, we're a lot more seasoned or investors are a lot more seasoned than they were before. The problem is, one of the problems is that the financing is expected to come in large part from private credit, which is essentially an investment vehicle for investors to go lend money to say like companies that want to build data centers, right?
35:49 Josh Clark? And this is largely unregulated, it's very shadowy. We don't know how many, like how much debt exists in the world on private credit because they don't have to report this stuff. And you know, as we learned from the 2008 meltdown, when there's like a massive speculation among finances that involves debt, that's, that can go really bad.
36:14 Speaker 02 Yeah, for sure. And then speaking of going bad, I guess we're at the sort of environmental piece of this whole thing. And this is what I was talking about when I said that, you know, people should just be aware of what they're taking part in. And again, this is not to shame anybody who uses AI for their job or just to make funny fake videos. But, but you know, everyone is sort of tied together to make this what it is, who's using that stuff. And, and I get if someone says like, hey, if I quit this thing, it's not going to make any difference. But that's sort of the- The age old, like, you know, if I don't recycle my tin can, my aluminum can, tin cans, my aluminum cans, then it's not gonna make that big of a difference. But the idea of everyone getting together to do something for the common good,
36:58 Speaker 02 that's where change happens.
37:00 Josh Clark? Right.
37:01 Speaker 02 Or where negative change happens. So as far as AI data centers go, the main, you know, aside from just, you know, the land use and everything else and the hardship on the local economies and towns in certain ways that we're gonna get to, it's really just a succubus of electricity and water usage.
37:25 Josh Clark? Yeah.
37:26 Speaker 02 Succubus is not the right word.
37:27 Josh Clark? No, but it makes sense. It's like a bunker down.
37:31 Speaker 02 Yeah, but I say succubus to mean just like a bottomless pit. But I know that's not what it means, by the way.
37:36 Josh Clark? A giant sucking thing, right?
37:39 Speaker 02 Right.
37:39 Josh Clark? And it is, it's sucking tons of electricity and water up. Like, some of these AI data plants use the same amount of electricity as a town of 50,000.
37:49 Speaker 02 Yeah.
37:49 Josh Clark? And about the same amount of water as a town of 50,000 people. This is a data center we're talking about. And it's not even necessarily an AI data center, just any hyperscale data center. It uses a ton of electricity and water. Reason it uses water is because all of these processors, the CPUs that are doing all this work and just all of the networking that's going on with it's generating heat. And computing happens faster when it's cooler. So to keep the place cool, they use evaporative cooling where they funnel waste heat air through wet pads, essentially. Like, they just buy old mattresses and dose them with water.
38:31 Josh Clark? And then they run the heat through there and it, through evaporative cooling, it cools it off. It's, it uses a little less electricity than air cooling, but it uses water, a lot of water.
38:45 Speaker 02 Yeah, I mean, I assume most people know this, but like, your laptop has a tiny fan in it. Like, every computer in the world has a little fan in it that cools it down. So when you've got all this stuff together, you know, it's gonna generate tons and tons of heat. That was the whirring of the server room that you used to sleep in.
39:03 Speaker 02 Those were all fans, you know, and you know, there's some other sounds coming, but mostly it was those fans trying to cool everything down. We got a lot of stats here that are pretty eye popping, but there are 11, roughly 11,000 data centers around the world. Most of these are not AI, obviously, but they're the most, you know, robust sort of users of the energy, but they use between 1 and 1.5%, which it doesn't sound like a lot, but of the entire world's electricity usage.
39:31 Josh Clark? I know.
39:32 Speaker 02 On planet Earth goes to data centers right now, and in certain places like Ireland, data centers use about 20% of the country's electricity.
39:40 Josh Clark? Yeah, and if you dive into different places around like the world where data centers are, like that's collectively, right? All of them in Ireland, all of them in the world. If you kind of zoom into the towns where these things are located, there's, well, there's something called Data Center Alley in Northern Virginia, outside of DC, where there's this huge concentration of large data centers, probably the biggest concentration in the world. Those data centers use about the same amount of electricity as 60% of all the households in the state of Virginia.
40:16 Speaker 02 Yeah, here's another one. By 2030, they're predicting, this is Barclay's Bank, is predicting that data center energy use in the United States would make up about 13% of the entire electricity demand of the United States. And Meta has their, they all have silly names, but their data centers called Hyperion. They're all, you know, one was, where are the, where's that list? They're all these kind of sci-fi sounding names.
40:44 Josh Clark? Yeah, Stargate.
40:46 Speaker 02 Yeah.
40:47 Josh Clark? Jupiter, Prometheus.
40:48 Speaker 02 Oh, God.
40:49 Josh Clark? I'm sure all of those nerds are like, what do you mean, silly?
40:52 Speaker 02 If I opened up a data center, I'd call it Old Bessie.
40:55 Josh Clark? Old Bessie is, that's, I hope so bad that somebody's listening to this and they open a massive hyperscale data center named Old Bessie.
41:05 Speaker 02 That would be great, but Meta's Hyperion data center will consume, by the time it's finished, about five gigawatts. And if you're like, what's five gigawatts? That is about half of the peak load of all of New York City.
41:18 Josh Clark? The most that it can possibly, that can possibly be demanded, right?
41:21 Speaker 02 Yeah, the very top load, probably, I guess, New York City on the hottest day of the year, with all the lights on at night or something.
41:29 Josh Clark? Yeah.
41:31 Speaker 02 And that Rockefeller tree, just, they just, summer version.
41:34 Josh Clark? That puts it over the, over the edge.
41:37 Speaker 02 That's right.
41:37 Josh Clark? Blackout. So you can imagine that when you're using all this electricity and using all this water, if you're starting to build these massive data centers, you're looking for places that have like cheap land, cheap electricity, and because electricity is often more expensive than water, they'll go to places, they'll build them in places that are like water scarce, that have cheap electricity. On the premise that like, we're a massive multinational corporation, we can push around this little county and use up all of their water, and what are they gonna do? Nothing.
42:14 Speaker 02 Yeah, and I mean, that's literally happening. There's one right here in Georgia, in Newton County, it's a metadata center that's using 10% of the local water use. And like you said, water is a, is a resource that isn't infinite. We've talked about the dangers in the future of like, you know, perhaps the wars of the future will be fought over water. And this could get us there. I think in Phoenix, Arizona,
42:38 Speaker 02 you know, known for their abundant water, Meta and Microsoft use 7 million gallons of water every single day for their data centers.
42:46 Josh Clark? Yep. Every day, you said?
42:49 Speaker 02 Every day, 7 million gallons of water.
42:51 Josh Clark? That's insane.
42:53 Speaker 02 Yeah.
42:53 Josh Clark? And when I saw this, I was like, oh, here we go. In the UK, data centers use 10 billion liters of drinking water every year.
43:04 Speaker 02 L I T R E S.
43:06 Josh Clark? Yeah, that's right.
43:08 Speaker 02 But, you know, you mentioned some of these towns. Not only are some of the, they're like using, let's say 10% of the local water here in Newton County. In Virginia, where data center alley is, some of these places are like, some of these towns are running out of water. Like they go to turn on their water and water doesn't come out because of this.
43:26 Josh Clark? Oh, plus also, like we talked about how gamers are getting, um, getting, uh, the short end of the stick. When it comes to buying graphic cards because they are in such high demand. Same thing happens with electricity. So in addition to this data center coming to town and using up all your water. They're also jacking up your electricity prices because there's only so much that your local electric, electrical company can produce. So because of supply and demand, your price is going to rise. And I guess around, um, data center alley in Northern Virginia. Electricity prices have increased 267% since 2020. And that also is affecting, um, Maryland, which is getting little to no benefit from data center alley.
44:09 Josh Clark? And is just helping pay the price for it. This is subsidization of these data centers. Like they are subsidized in just about every single way you can imagine.
44:20 Speaker 02 Yeah, for sure. Uh, and if you say like, oh, well, sure, but they create jobs, right? So that's great for the local economy. Kyle gives an example here of, uh, Northumberland, England. There's a 10 billion pound data center there. Uh, or I guess it's, it's coming. And you think, oh, great, that's, that's going to employ probably like 5,000 people, right? It's going to employ 400 people with full-time jobs.
44:44 Josh Clark? Yeah, a $10 billion or 10 billion pound data center, 400 jobs. Because these things are so efficient and everything is just so advanced. They don't really need that many people to keep an eye on it, right? Plus also, the money from that data center, they're not going to spread it around the UK. It's going to flow right back to the US to the parent company.
45:07 Speaker 02 Oh yeah, for sure. And, you know, we also didn't point out that a lot of these, these energy grids like are literally gonna buckle under pressure at some point. Like, they're not built for this.
45:19 Josh Clark? Yes. And we're not. So I know it sounds like we're just like, and this, and that. How terrible are data centers? Like, they're incredibly important and they support an amazing array of really great stuff, right? And they are the foundation that the next expansion of the digital economy and the world culture are going to grow on. Like, they're incredibly important, but they have a lot of problems with them that need to be addressed. They're not being addressed because every government from like the local city council up to the leaders of the free world are just giving these people whatever they want.
46:02 Josh Clark? That's what's going on now. There's no checks going on at all right now. That's the problem.
46:09 Speaker 02 Yeah. And that's because the flow of money is so great at this point to a certain segment of the population only. They're protecting their own investment, you know? They're watching their own backsides.
46:23 Josh Clark? That's definitely, I would say, 99% of it. But I think there's also, Chuck, a little factor of like gee whiz. Like, these titans of the AI industry are good at like razzle-dazzling elected officials into doing whatever they want by, I think, making them feel included in this new frontier, essentially. I think there's a certain element of that.
46:48 Speaker 02 I think you're probably right. It's, hey, maybe it'll all work out great.
46:52 Josh Clark? Sure, it probably will. It usually does. Astoundingly, it usually does work out.
46:58 Speaker 02 Well, true, as far as the world hasn't ended.
47:01 Josh Clark? That's exactly what I mean.
47:02 Speaker 02 Yeah, yeah.
47:03 Josh Clark? Yeah. So I think that's it. We said, yeah, like, four or five times in succession. I think we accidentally triggered listener man.
47:12 Speaker 02 That's right. This relates to our History of the BBC episode, and this is from Erica. And Erica says, hey, guys, I really loved the episode. It left me reflecting on how I've come to understand the country through both the content the BBC produces and the people's reactions to the BBC. But more recently, my work as an academic has enabled me to be involved in creating programs for the BBC across TV, radio, and online, because there's one awesome fact about the BBC that wasn't included. For over 50 years, the BBC has partnered with the Open University, OU, which specializes in accessible and distance education. The partnership started in the 1970s to provide learning at scale, including facilitating university-level lectures at night on public television.
47:54 Josh Clark? Neat.
47:54 Speaker 02 Today, the partnership facilitates access to academic consultants to co-produce high-quality, informed content across platforms, including some of the David Attenborough nature stuff.
48:06 Josh Clark? Nice.
48:07 Speaker 02 Additionally, the Open University creates supplementary materials to enable people to continue their learning journey and explore topics in more detail. So whether viewers or listeners realize it or not, this partnership enables the public to benefit from specialist knowledge in accessible ways. And that is from Erica from the Open University, who is a professor of medical anthropology.
48:30 Josh Clark? Oh, wow. That's an awesome... Erica, you got to send us some topic ideas, too.
48:35 Speaker 02 Totally. Right up your alley.
48:37 Josh Clark? And congratulations. That's pretty neat, making stuff in conjunction with the BBC. That's got to be a neat, high-water mark, you know?
48:44 Speaker 02 Agreed.
48:44 Josh Clark? And I think, Chuck, I'm curious to see if we go look at our account, if we'll see a little line item from Open University and one from BBC.
48:55 Speaker 02 Well, what would be, like, seven pounds or something? I don't know the exchange rate right now.
48:58 Josh Clark? That sounds about right.
48:59 Speaker 02 All right.
49:00 Josh Clark? Great. Well, thanks again, Erica. And please do send us some medical anthropology ideas, because that just sounds like it'll knock our socks off. And if you want to be like Erica and try to knock our socks off, good luck. You can send it off to us at stuffpodcast@iHeartRadio.com.
49:20 Speaker 00 Stuff You Should Know is a production of iHeartRadio. For more podcasts, MyHeartRadio, visit the iHeartRadio app, Apple Podcasts, or wherever you listen to your favorite shows.
49:35 Josh Clark? Buy smart, drive for less and trade for more. That's Toyota Fordability. And it starts with a great deal on an off-road-ready Tacoma or half-ton Tundra, two of Toyota's longest-lasting trucks, combining power for work, tech for adventures, and style in between. And they're built on a reputation for quality and reliability, which is why Toyota trucks are known for holding their value, which could mean big savings when it's time for your next Toyota. So get a great deal today and a great value tomorrow. Shop toyota.com for more. Toyota, let's go places.
50:06 Speaker 00 This is an iHeart Podcast. Guaranteed Human.
Previous Short Stuff: Did a vampire legend help win the Cold War? Next Selects: BPD: The Worst Disorder or Not a Disorder at All? Page updated October 31, 2026 · transcript and summary are machine-generated