Latest / Stuff You Should Know / Data Centers: Can't Live With Em, Can't Live Without Em

Transcript

Data Centers: Can't Live With Em, Can't Live Without Em · Stuff You Should Know

• 0:01
  Speaker 1: Welcome to Stuff you should know, a production of iHeartRadio.
• 0:11
  Speaker 2: Hey, and welcome to the podcast. I'm Josh, and there's
• 0:14
  Speaker 2: Chuck and Jerry's here too, and we've got our pocket
• 0:16
  Speaker 2: protectors and tape on the bridge of our glasses day
• 0:21
  Speaker 2: and this is stuff you should know.
• 0:23
  Speaker 1: Nice work. Did you just hear something?
• 0:27
  Speaker 2: No, not weird. I heard you say nice work. Yeah,
• 0:30
  Speaker 2: but then stop abruptly.
• 0:31
  Speaker 1: Well I stopped abruptly because I thought I heard a
• 0:34
  Speaker 1: little digital glitch.
• 0:36
  Speaker 2: Oh no, I didn't hear anything.
• 0:39
  Speaker 1: I might be losing my mind, then, I you know, I'm.
• 0:42
  Speaker 2: Curious whether we end up editing this out or not.
• 0:45
  Speaker 2: Any other podcast on the planet would edit that out
• 0:47
  Speaker 2: without even thinking about it. But there's like a fifty
• 0:50
  Speaker 2: percent chance that'll stay in with us.
• 0:52
  Speaker 1: I mean, this is why we didn't get that Golden
• 0:54
  Speaker 1: Globe nomination. That's right, it's kind of classic stuff you
• 0:57
  Speaker 1: should know. On professionality.
• 0:58
  Speaker 2: It's exactly right in the Italian accent, mispronunciations, there's a
• 1:02
  Speaker 2: whole laundry list. Yeah, that's okay, Chuck, I think we're
• 1:07
  Speaker 2: golden regardless.
• 1:09
  Speaker 1: Agreed.
• 1:11
  Speaker 2: So we're talking about data centers, which I had a
• 1:14
  Speaker 2: very very rough idea about But actually no, I knew
• 1:18
  Speaker 2: that they existed essentially and that they were becoming a
• 1:21
  Speaker 2: problem with the rise of AI. Yes, that was about it.
• 1:25
  Speaker 2: How about you are you are you d data center Feliac?
• 1:32
  Speaker 1: No? You know me, I'm not super technology minded, so
• 1:35
  Speaker 1: I don't know a lot about this stuff. I remember
• 1:37
  Speaker 1: walking by our server room back in the day when
• 1:41
  Speaker 1: we were at Pont City Market and seeing our colleague
• 1:44
  Speaker 1: Izzy in there hard at word yeah. And when that
• 1:47
  Speaker 1: door was unlocked and open, hearing the were of the
• 1:50
  Speaker 1: of the servers and the cooling machines, and you know,
• 1:53
  Speaker 1: that's on a smaller scale, that's a data center.
• 1:57
  Speaker 2: Absolutely, one hundred percent, that's a data center. It was
• 1:59
  Speaker 2: also so a great place to curl up and take
• 2:01
  Speaker 2: a nap in.
• 2:02
  Speaker 1: The middle of the day, like the warmth of the server.
• 2:05
  Speaker 2: Yeah, and the word put right the word yeah, So
• 2:10
  Speaker 2: uh yeah, that definitely counts as a data server. If
• 2:13
  Speaker 2: you have like one of those little home networking setups
• 2:16
  Speaker 2: in like a closet in your house, data center. Sure,
• 2:19
  Speaker 2: technically the PC is a data center. Anywhere you can
• 2:24
  Speaker 2: store and access data, that's technically a data center. And
• 2:28
  Speaker 2: you're like, well, that's stupid. Why did you even say
• 2:30
  Speaker 2: that Josh's quibbling, that's quotitian. Shut up and get on
• 2:33
  Speaker 2: with data centers. Whoa, whoa, whoa. First of all, don't
• 2:36
  Speaker 2: use the S word and then secondly word yes, Wait,
• 2:41
  Speaker 2: what did I say? That's the keyword Quoteitian?
• 2:43
  Speaker 1: Oh no, that's kW right.
• 2:49
  Speaker 2: So the reason that I bring that up, though, Chuck,
• 2:51
  Speaker 2: is because that technically is part of the progression of
• 2:55
  Speaker 2: data centers. Yeah, it probably goes without saying, but it's
• 2:57
  Speaker 2: evolved along with computing, and his computer's kind of gotten
• 3:01
  Speaker 2: bigger and bigger. The need to store and access more
• 3:05
  Speaker 2: data has gotten bigger so much so, Chuck that just
• 3:07
  Speaker 2: wrap your head around this one. In twenty twenty four,
• 3:11
  Speaker 2: just over a year ago. Yeah, we used one hundred
• 3:14
  Speaker 2: and fifty zetabytes of data. That's what we consumed. And
• 3:19
  Speaker 2: consuming is anything from making a video and uploading it
• 3:22
  Speaker 2: to TikTok or putting a post up on Instagram. It's
• 3:27
  Speaker 2: browsing a website, it's buying a song from iTunes, it's
• 3:31
  Speaker 2: doing web analytics, it's buying something with your American Express card.
• 3:36
  Speaker 2: All of that is data consumption. And we consumed one
• 3:39
  Speaker 2: hundred and fifty zetabytes of data in twenty twenty four.
• 3:44
  Speaker 1: Yeah, I don't even know how many big Max that is.
• 3:47
  Speaker 1: I know you name dropped a lot of brands. It
• 3:49
  Speaker 1: should be like the movies where every time you even
• 3:52
  Speaker 1: just say, like buy something on your amex, the bank
• 3:55
  Speaker 1: account grows by like ten dollars.
• 3:57
  Speaker 2: I agree wholeheartedly. I agree that amex should do that. Amex,
• 4:03
  Speaker 2: there's twenty bucks. I'll split it with you.
• 4:05
  Speaker 1: Wow, you just bought lunch in nineteen ninety seven.
• 4:08
  Speaker 2: That's right. So just real quick, as zeta byte, chuck
• 4:11
  Speaker 2: is a trillion gigabytes, So we consumed one hundred and
• 4:15
  Speaker 2: fifty trillion gigabytes. That was twenty twenty four.
• 4:19
  Speaker 1: That's worldwide, right, yes, yeah.
• 4:21
  Speaker 2: That's worldwide. In twenty ten, we consumed two zeta bytes. Jeez. Yeah.
• 4:26
  Speaker 2: So it's growing exponentially, which means that data centers are
• 4:29
  Speaker 2: growing exponentially. And now they're about to just blow up,
• 4:34
  Speaker 2: like truffle up essentially from you know, this kind of
• 4:40
  Speaker 2: calm like plateau that they'd reached. It's about to just
• 4:42
  Speaker 2: go in hyper drive.
• 4:44
  Speaker 1: Yeah, and actively is, and we're going we're gonna get
• 4:46
  Speaker 1: to some startling statistics later on in the episode. But
• 4:50
  Speaker 1: Kyle helped us out with this our writer over in
• 4:53
  Speaker 1: the UK.
• 4:54
  Speaker 2: He did a fantastic job.
• 4:56
  Speaker 1: He did a really good job. And there's going to
• 4:57
  Speaker 1: be some UK specific things in here because Kyle's always keen,
• 5:00
  Speaker 1: as they say, to throw that stuff in there.
• 5:02
  Speaker 2: Yeah, for sure, Kyle likes to pepper those in Yeah, of.
• 5:05
  Speaker 1: Course, and he's not barred from doing so we allow it.
• 5:12
  Speaker 1: So since Kyle, you know, is frequenter of the Wayback Machine,
• 5:17
  Speaker 1: as are all the wonderful writers that we use, they
• 5:21
  Speaker 1: all had the keys to the car. Essentially, he jumped
• 5:24
  Speaker 1: in the Wayback Machine to sort of give us a
• 5:26
  Speaker 1: little bit of a timeline on data centers and a
• 5:28
  Speaker 1: bit on you know, mainframes and PCs.
• 5:30
  Speaker 2: He also left all of his used tea bags in
• 5:32
  Speaker 2: there too. I don't know if you noticed that.
• 5:33
  Speaker 1: Oh it's fine, you know, you can throw those back
• 5:35
  Speaker 1: in some hot water and they do just a little
• 5:37
  Speaker 1: bit weaker tea.
• 5:38
  Speaker 2: Well, if you put like five of them together, it's
• 5:40
  Speaker 2: like one.
• 5:41
  Speaker 1: Yeah, and Kyle, I mean that thing was full.
• 5:43
  Speaker 2: Maybe really what he drinks a lot of tea, he
• 5:46
  Speaker 2: really does.
• 5:47
  Speaker 1: So if you want to talk about the earliest data
• 5:49
  Speaker 1: centers that you could kind of call maybe a data center.
• 5:52
  Speaker 1: They were you know, computers, they were electronic computers. Most
• 5:56
  Speaker 1: of this stuff that we're going to talk about early
• 5:58
  Speaker 1: on was military and and as you'll see, even the
• 6:02
  Speaker 1: first when we talk about the UK one that was
• 6:05
  Speaker 1: supposedly in that military, they even loaned it to the military,
• 6:08
  Speaker 1: which was kind of interesting. But these things were built
• 6:12
  Speaker 1: with you know, state of the art technology at the time,
• 6:14
  Speaker 1: which meant vacuum tubes and you know, manual switches and
• 6:18
  Speaker 1: plugs and things like that. And the first thing that
• 6:21
  Speaker 1: we can really talk about as the first programmable electric
• 6:24
  Speaker 1: digital computer was the Colossus. And as we'll see, Elon
• 6:28
  Speaker 1: Musk has now stolen that for his own purposes that
• 6:31
  Speaker 1: name probably because of this, I would imagine, but it
• 6:35
  Speaker 1: was at Bletchley Park, of course, during World War Two,
• 6:38
  Speaker 1: and they were trying to you know, crack into Hitler's
• 6:42
  Speaker 1: messages at the time, and these things were huge and
• 6:45
  Speaker 1: kind of to me. The thing that stood out about Colossus,
• 6:47
  Speaker 1: which is a neat little factoid, is that where Colossus
• 6:51
  Speaker 1: was at Bletchley Park at Block H, it is now
• 6:54
  Speaker 1: the National Museum of Computing.
• 6:56
  Speaker 2: I want to go to that so bad when we
• 6:58
  Speaker 2: do that UK up tour next year. Oh, we got
• 7:03
  Speaker 2: to go to that together.
• 7:04
  Speaker 1: Okay, okay, are we doing that next year?
• 7:06
  Speaker 2: That's what we were talking about. All right, we kind
• 7:08
  Speaker 2: of already have promised it. We have to. Now we're
• 7:11
  Speaker 2: locked in the punch.
• 7:12
  Speaker 1: Why's my voice O high?
• 7:13
  Speaker 2: Then? I don't know. You're practicing for the alph Yeah,
• 7:17
  Speaker 2: that's right now.
• 7:18
  Speaker 1: No, that that'd be a lot of fun. I'd love
• 7:20
  Speaker 1: to go to that.
• 7:21
  Speaker 2: So that was Colossus. Another one about the same time
• 7:24
  Speaker 2: was the Aniac Electrical Numerical Integrator and Computer. So that's
• 7:28
  Speaker 2: a quality acronym. Yeah, and it was the first general
• 7:33
  Speaker 2: purpose electronic computer. And here's the thing. This is technically
• 7:38
  Speaker 2: not data storage yet, it's data processing, right, But these
• 7:42
  Speaker 2: things Colossus Aniac, you walked up to them and you said,
• 7:47
  Speaker 2: what's the trajectory of this missile if I fire it
• 7:50
  Speaker 2: from here? And Aniac would go peep pop, poop, poop,
• 7:52
  Speaker 2: and then say, like, uh, whatever a trajectory is described
• 7:56
  Speaker 2: in Sure or Colossus. You'd be like, what is Hilter
• 7:59
  Speaker 2: saying here to Goebbels And the Colossus would say, Hilter
• 8:04
  Speaker 2: is saying that he's a big fan of goebels work,
• 8:08
  Speaker 2: but he's suspicious that the rest of the world doesn't
• 8:11
  Speaker 2: like either of them. And that was it. After that,
• 8:14
  Speaker 2: you'd be like, hey, what was the last answer, and
• 8:16
  Speaker 2: they'd be like, what's an answer?
• 8:18
  Speaker 1: Yeah, you gotta just tell me what's going on with
• 8:20
  Speaker 1: this Hilter business.
• 8:21
  Speaker 2: You don't remember from our art Mysteries of the Art World,
• 8:26
  Speaker 2: how stuff works are to go? Oh, the title of
• 8:29
  Speaker 2: that section was did Hilter do these pain? Oh?
• 8:32
  Speaker 1: My god, that's a deep cut. I did not remember that.
• 8:35
  Speaker 2: Yes, and I think it still says that on that
• 8:39
  Speaker 2: oh babilistical Yeah, yeah, I can only hope it's got
• 8:42
  Speaker 2: to be Hilter forever. All right.
• 8:44
  Speaker 1: So we go into mainframes at this point, and this
• 8:47
  Speaker 1: is like the nineteen fifties basically when companies could actually
• 8:52
  Speaker 1: have their own computer. It wasn't just the military. These
• 8:55
  Speaker 1: were the old punch card computers, and they were called mainframes.
• 8:58
  Speaker 1: It wasn't made up for this. Mainframes were originally described
• 9:02
  Speaker 1: We're describing like what you would house telecommunication equipment and
• 9:05
  Speaker 1: maybe some other sciencey stuff, but it was referencing literally
• 9:09
  Speaker 1: the cabinets that held this technology. And it became known
• 9:12
  Speaker 1: as just you know, it kind of took over when
• 9:15
  Speaker 1: the computer's world started using it as computer only.
• 9:18
  Speaker 2: Yeah, but again, this is like you're a company and
• 9:21
  Speaker 2: this is where you store and process all of your data,
• 9:24
  Speaker 2: and it's in this one room, but it's not going
• 9:27
  Speaker 2: anywhere else. It's not for anybody else, and you have
• 9:29
  Speaker 2: to physically be in the room to get your answer,
• 9:33
  Speaker 2: process whatever data you're looking for. When the PC came along,
• 9:37
  Speaker 2: and then the Mac and Tosh came along, they took
• 9:40
  Speaker 2: that thing and just made it very small so you
• 9:44
  Speaker 2: could put it on all of your employee's desks. And
• 9:46
  Speaker 2: now they had, like I was saying, before, their own
• 9:48
  Speaker 2: little data center right there. So if you said, like, hey,
• 9:52
  Speaker 2: what's the I need to know the Q four reports,
• 9:56
  Speaker 2: they'd say, go to Debbie's desk. Debbie's is the one
• 9:59
  Speaker 2: who's got that on her computer. And you would go
• 10:01
  Speaker 2: over there and say, WI, what's the Q four report?
• 10:04
  Speaker 2: And w would give it to you. Right, there was
• 10:06
  Speaker 2: no connectivity, but you could still like do a lot
• 10:10
  Speaker 2: more stuff than you could when you had a mainframe. Yeah,
• 10:14
  Speaker 2: for sure, which makes mainframes feel like really outdated, but
• 10:17
  Speaker 2: it turns out they're like totally still in use. Today.
• 10:20
  Speaker 1: Oh yeah, absolutely, I do want to jump back in
• 10:22
  Speaker 1: time a little bit, because I did. I promised talk
• 10:24
  Speaker 1: of lending the military, oh basically your equipment, and that's
• 10:29
  Speaker 1: what happened. In nineteen fifty one, there was a T
• 10:31
  Speaker 1: shop chain in the UK. I don't know if it's
• 10:34
  Speaker 1: still around, Lions Lyns, and they were the very first
• 10:37
  Speaker 1: company in the world that used a mainframe. It was
• 10:39
  Speaker 1: called the LEO, the l EO, and it was you know,
• 10:42
  Speaker 1: like what you would think they've handled, like payroll and
• 10:45
  Speaker 1: stock management and stuff like that, but there wasn't a
• 10:49
  Speaker 1: lot for it to do at a T shop chain
• 10:52
  Speaker 1: except for those couple of things. And so they calculated
• 10:56
  Speaker 1: missile trajectories like you were talking about for the Ministry
• 10:59
  Speaker 1: of Defense.
• 11:00
  Speaker 2: And that actually kind of helped establish like a I
• 11:05
  Speaker 2: guess a pay schedule. How people charged for data centers
• 11:09
  Speaker 2: to come. Yeah, it was they like you would charge
• 11:12
  Speaker 2: them for the time that they used it, or you
• 11:15
  Speaker 2: could lease it for a month. And that really started
• 11:17
  Speaker 2: to come around when IBM got in the game. They
• 11:20
  Speaker 2: became like the main frame leader in the fifties, the
• 11:23
  Speaker 2: early fifties, I think they had a unit that you
• 11:25
  Speaker 2: could lease for sixteen thousand dollars per month. That's in
• 11:29
  Speaker 2: nineteen fifty two money. And then as the things as
• 11:32
  Speaker 2: like the processors got better and smaller and faster, that
• 11:37
  Speaker 2: price came down dramatically. And then finally in the sixties
• 11:41
  Speaker 2: they released the IBM six System three sixty, which not
• 11:46
  Speaker 2: only got Apollo eleven to the moon and back, it
• 11:51
  Speaker 2: is a It appears in an episode of mad Men apparently.
• 11:55
  Speaker 1: Oh really, yeah, you usould have though, right.
• 11:58
  Speaker 2: No, I never did. I just saw reference to it
• 12:00
  Speaker 2: on the government and you knew it was the show. Yeah,
• 12:05
  Speaker 2: but they like you should look up pictures of it.
• 12:07
  Speaker 2: It's like those giant burnt orange cabinets with the real
• 12:10
  Speaker 2: magnetic tape. It's just beautiful. They're cool to look at.
• 12:13
  Speaker 1: Yeah, yeah, I remember. We've referenced the movie War Games
• 12:15
  Speaker 1: from our childhood in the eighties a lot, and that
• 12:18
  Speaker 1: the whopper for War Games was that was you know,
• 12:21
  Speaker 1: at that age, to see the whopper and action and
• 12:25
  Speaker 1: to see Matthew almost aid, Matthew Modine, Matthew Broderick hanging
• 12:30
  Speaker 1: up his handheld telephone receiver onto a modem to talk
• 12:34
  Speaker 1: to the school computer.
• 12:35
  Speaker 2: It was mind blowing that, Yeah, that phone in the
• 12:38
  Speaker 2: modem made just a really big impression on me.
• 12:41
  Speaker 1: Yeah, and a cool sound.
• 12:44
  Speaker 2: Peepoop boop.
• 12:46
  Speaker 1: Should we take a break?
• 12:48
  Speaker 2: Wait, let me talk about mainframes today though, because I
• 12:50
  Speaker 2: just want to give a little I don't know if
• 12:52
  Speaker 2: a shout out's the right term, but they are still
• 12:55
  Speaker 2: around because they're so reliable, Because they're so secure. You
• 12:59
  Speaker 2: can make it so that there's information on those things
• 13:02
  Speaker 2: that you again have to be physically present in the
• 13:05
  Speaker 2: room to access. Sure, you can put all sorts of
• 13:07
  Speaker 2: different layers of security. So if you're like Visa, or
• 13:11
  Speaker 2: you're a healthcare company, or you're the Census Bureau, you're
• 13:16
  Speaker 2: probably still using a main frame because you're protecting information
• 13:20
  Speaker 2: as tightly as you can. But those things are also
• 13:22
  Speaker 2: super fast and can hold huge volumes of computation at once.
• 13:27
  Speaker 1: Yeah, or a non Golden Globe nominated podcast.
• 13:31
  Speaker 2: Yeah, we've got our own main frame. Yeah, we've got
• 13:34
  Speaker 2: our IBM three sixty that's right.
• 13:37
  Speaker 1: What year was that from again, the three.
• 13:39
  Speaker 2: Sixty sixty four.
• 13:40
  Speaker 1: Yeah, yeah, that's the one. I'll just making sure we
• 13:43
  Speaker 1: didn't have the sixty five because that no, no, no,
• 13:45
  Speaker 1: the sixty buggy. Yeah, yeah, all right, we can take
• 13:48
  Speaker 1: that break now and we're gonna jump out of the
• 13:51
  Speaker 1: wayback machine and venture into the modern world.
• 13:53
  Speaker 2: Right after this.
• 14:24
  Speaker 1: All right, so we are out of the way back machine.
• 14:27
  Speaker 1: We're making that uh we're combining all those old tea
• 14:30
  Speaker 1: bags and making some still somewhat weaker tea. Yeah, it's
• 14:34
  Speaker 1: not too bad. It's a combination of Earl Gray and
• 14:37
  Speaker 1: Camma meal, all kinds of fun stuff, but not too bad.
• 14:41
  Speaker 1: And now we're going to talk a little bit about
• 14:43
  Speaker 1: when things started to ramp up, because it kind of
• 14:45
  Speaker 1: happened in fits and starts. And one of the biggest, uh,
• 14:49
  Speaker 1: I guess would it be a fitter a start was
• 14:51
  Speaker 1: the Internet. Because once the Internet came along, every business
• 14:54
  Speaker 1: in the world started using it, and so all of
• 14:57
  Speaker 1: a sudden you had to have a lot more data
• 15:01
  Speaker 1: storage and bigger data centers and bigger server rooms in
• 15:05
  Speaker 1: your companies, which was, you know, a pretty good thing
• 15:08
  Speaker 1: at the time. After the dot com bus. There were
• 15:10
  Speaker 1: a lot of casualties of that growth, but then things
• 15:13
  Speaker 1: kind of you know, the ship kind of righted itself.
• 15:15
  Speaker 2: Yeah, and because it was like accessible to basically every
• 15:19
  Speaker 2: business now, like you didn't have to buy a mainframe
• 15:23
  Speaker 2: you could lease space on someone else's main frame, like
• 15:26
  Speaker 2: you're the Ministry of Defense or something, all of a sudden.
• 15:29
  Speaker 2: So that led to this huge proliferation that gave the
• 15:34
  Speaker 2: foundation for web commerce. E commerce, that's what they used
• 15:38
  Speaker 2: to call it. That's an old timey term now, but
• 15:41
  Speaker 2: it created the ability for e commerce to start and flourish.
• 15:44
  Speaker 2: So this data center's scaling up to meet the needs
• 15:48
  Speaker 2: of the Internet and then to kind of give people
• 15:51
  Speaker 2: all sorts of new space and room to come up
• 15:53
  Speaker 2: with new stuff. That's where the digital economy came from,
• 15:57
  Speaker 2: right there.
• 15:58
  Speaker 1: Yeah, all of a sudden, you could hop on webvan
• 16:01
  Speaker 1: for an order a SACA groceries.
• 16:04
  Speaker 2: I have a friend who was all in about that.
• 16:08
  Speaker 1: Well man, yeah, yeah, I think I had a friend
• 16:11
  Speaker 1: who was pretty heavily invested.
• 16:15
  Speaker 2: It's but clearly it was ahead of its time. I mean, sure,
• 16:18
  Speaker 2: let's see, it's a post Mates and.
• 16:22
  Speaker 1: Well there's a lot of several of them now that
• 16:24
  Speaker 1: have succeeded.
• 16:25
  Speaker 2: Well name them, will get ten bucks each?
• 16:27
  Speaker 1: Well, well you just got ten for post Mates, and
• 16:30
  Speaker 1: you got to split that with me.
• 16:31
  Speaker 2: You know, I will, all right.
• 16:34
  Speaker 1: Cloud computing was the next big jump. When cloud computing
• 16:37
  Speaker 1: came around the early two thousands, or do they call
• 16:40
  Speaker 1: that the early oughts? I do, okay, I thought i'd
• 16:44
  Speaker 1: heard that come from your mouth. But that is when
• 16:47
  Speaker 1: you know, that was the real game changer because things
• 16:50
  Speaker 1: were still I mean, when cloud computing came along, people
• 16:53
  Speaker 1: thought of it if they didn't look too hard into it.
• 16:56
  Speaker 1: They thought it was just you know, floating up in
• 16:58
  Speaker 1: the ether somewhere. Yeah, it's it's still being stored on stuff,
• 17:01
  Speaker 1: it's just not being stored locally. So all of a sudden,
• 17:05
  Speaker 1: things were just going somewhere else for someone else to
• 17:09
  Speaker 1: worry about all that storage. And more importantly, they could
• 17:12
  Speaker 1: they could link everything together and store a lot of
• 17:15
  Speaker 1: stuff from a bunch of different people.
• 17:17
  Speaker 2: Right, So now you have data centers not just available
• 17:21
  Speaker 2: to somebody like a huge bank or something like that,
• 17:24
  Speaker 2: or the government in a shop then a bank, yeah,
• 17:27
  Speaker 2: or a t shop, and then to e commerce businesses.
• 17:31
  Speaker 2: Now it's available to you and me. So it's really
• 17:34
  Speaker 2: hard to remember back because the world has changed so much.
• 17:38
  Speaker 2: But Chuck, like two thousand and eight, two thousand and nine,
• 17:42
  Speaker 2: they were given us like VPN, little things that like
• 17:46
  Speaker 2: you could go home and work and like you would
• 17:48
  Speaker 2: it would never work. I'd never understood how to make
• 17:51
  Speaker 2: it work. Yeah, but that was like the very beginning
• 17:54
  Speaker 2: of how you could take your work home with you
• 17:56
  Speaker 2: and work from home and do things remotely like we
• 17:59
  Speaker 2: can now. Like it's nothing, but this led to the
• 18:02
  Speaker 2: rise of businesses like Dropbox. Right. So Dropbox goes to
• 18:06
  Speaker 2: Amazon Web Services and say, hey, we want to buy
• 18:09
  Speaker 2: a bunch of your cloud, right, which means that they're
• 18:12
  Speaker 2: going to use a bunch of like different servers and
• 18:15
  Speaker 2: different data centers all over the place. And then Dropbox
• 18:17
  Speaker 2: turns around to you and says, hey, if you give
• 18:20
  Speaker 2: me nineteen ninety five a month, you can have one
• 18:24
  Speaker 2: terabyte of data. Right. You can consume one terabyte of data, right,
• 18:28
  Speaker 2: and then hopefully you don't use all of that, so
• 18:31
  Speaker 2: they don't have to pay Amazon Web Services for stuff
• 18:35
  Speaker 2: they didn't use. But you're paying that nineteen ninety five
• 18:37
  Speaker 2: a month rather whether you use that whole terabyte or not.
• 18:41
  Speaker 2: It's a pretty smart business model. Would not exist at
• 18:44
  Speaker 2: all if the cloud didn't exist.
• 18:47
  Speaker 1: Yeah, as funny as you were talking about two thousand
• 18:49
  Speaker 1: and eight and how quaint that is, now, that's what
• 18:51
  Speaker 1: the year we started the show I know, I know,
• 18:53
  Speaker 1: it's crazy to think about it really.
• 18:56
  Speaker 2: Is, but imagine like working from home at that time,
• 18:59
  Speaker 2: it was just didn't you could know.
• 19:02
  Speaker 1: It was kind of great you went home and you homed.
• 19:04
  Speaker 2: That's exactly right. That was a big difference I remember.
• 19:08
  Speaker 1: But these data centers have now come together in such
• 19:11
  Speaker 1: a big way now that they the largest ones are
• 19:13
  Speaker 1: called hyper scale and they host more than five thousand servers,
• 19:18
  Speaker 1: like servers, not individual persons. Data. It's like, it's incredible,
• 19:23
  Speaker 1: how big is how big they've gotten Googles. And we'll
• 19:25
  Speaker 1: go over some of the kind of square footage and
• 19:27
  Speaker 1: then later talk about the elephant in the room, which
• 19:30
  Speaker 1: is energy and water usage. But Google's first data center
• 19:34
  Speaker 1: was built in two thousand and six, just but two
• 19:36
  Speaker 1: years before stuff you should have launched. And this is
• 19:38
  Speaker 1: an Oregon and they are still expanding that thing beyond
• 19:43
  Speaker 1: one point three million square feet. Meanwhile, in China they're like,
• 19:48
  Speaker 1: hold my tea, I guess, because China Telecom has a
• 19:53
  Speaker 1: ten point seven million square foot data center in Inner Mongolia.
• 19:58
  Speaker 2: It's two hundred and fifty acres.
• 20:01
  Speaker 1: That is a warehouse full of worrying servers heating up
• 20:06
  Speaker 1: and being cool.
• 20:07
  Speaker 2: Yeah, which is a big problem for any data center,
• 20:10
  Speaker 2: it turns out. But the whole expansion this jump starting
• 20:14
  Speaker 2: in twenty seventeen thanks to cloud computing, because again, cloud
• 20:18
  Speaker 2: computing just means all your stuff isn't on one server
• 20:20
  Speaker 2: and one data center, it's broken up into pieces and
• 20:23
  Speaker 2: spread all over the place. That's the cloud. That's basically it.
• 20:26
  Speaker 2: Even though it's way more advanced and intricate than that.
• 20:30
  Speaker 2: That's like all you really need to know for the
• 20:32
  Speaker 2: purposes of this episode, right. It led to a huge jump,
• 20:36
  Speaker 2: a huge need in data centers, and it also expanded
• 20:39
  Speaker 2: all the stuff we can do now, and COVID actually
• 20:43
  Speaker 2: gave it another bump, made building a data center very
• 20:48
  Speaker 2: economically attractive thing to do if you had the money.
• 20:51
  Speaker 2: Because remote working finally finally established itself as like, no,
• 20:56
  Speaker 2: we're doing this, stop calling us back to the office, ye,
• 21:00
  Speaker 2: which is what they're doing now. I know, I hope
• 21:04
  Speaker 2: it doesn't work because I remember when all of that
• 21:07
  Speaker 2: started and everybody was so nervous, like management was all
• 21:10
  Speaker 2: so nervous that people were just going to totally like
• 21:12
  Speaker 2: mess around and everything. It just it didn't happen. I
• 21:15
  Speaker 2: don't know anybody who's even been like gotten a talking to,
• 21:19
  Speaker 2: let alone, been fired for just messing around at home.
• 21:23
  Speaker 2: As a matter of fact, like you were saying, it
• 21:25
  Speaker 2: just makes you work more.
• 21:27
  Speaker 1: Yeah, I mean, do you know how many times in
• 21:29
  Speaker 1: our old offices I would see Jonathan Strickland just wandering
• 21:32
  Speaker 1: aimlessly through the office chatting with people.
• 21:34
  Speaker 2: Yes, I do, because he would chat with me a lot.
• 21:37
  Speaker 2: He did.
• 21:38
  Speaker 1: He did that in front of God and everybody, as
• 21:41
  Speaker 1: they say, right there in the office. So I can't
• 21:43
  Speaker 1: imagine what happened with him at home.
• 21:44
  Speaker 2: Yeah, like it was a sorority mixer or something.
• 21:49
  Speaker 1: We love Strickland. He's still around everyone. By the way,
• 21:51
  Speaker 1: he retired from tech stuff, but he's still with a company.
• 21:54
  Speaker 2: Which is great. We love Strickland all right.
• 21:57
  Speaker 1: So now we're onto AI data centers and that was
• 21:59
  Speaker 1: the I mean to call it a game changer. Seems
• 22:04
  Speaker 1: quaint compared to the rise of cloud computing and everything,
• 22:07
  Speaker 1: because it is off to the races in a way
• 22:10
  Speaker 1: that seemingly cannot be stopped. The genie has left the bottle,
• 22:14
  Speaker 1: as they say. Starting in twenty twenty two, when chat
• 22:17
  Speaker 1: GPT was released by open Ai, all of a sudden,
• 22:22
  Speaker 1: the need for these data centers became exponentially greater in
• 22:26
  Speaker 1: size in the speed at which they need these things
• 22:29
  Speaker 1: built because AI requires a ton of computing power to operate.
• 22:36
  Speaker 2: So much so that they don't even use the standard
• 22:40
  Speaker 2: what's called the compute machine. So compute is like all
• 22:44
  Speaker 2: of the processing power, the networking, all of that stuff,
• 22:47
  Speaker 2: and traditionally with a computer that's done on a CPU. Right,
• 22:52
  Speaker 2: that's how all of this gets done, right, everything else
• 22:54
  Speaker 2: is infrastructure. The CPU is doing all of the work.
• 22:58
  Speaker 2: Those are so like they still work like. Most data
• 23:01
  Speaker 2: centers are running on CPUs for AI, just not fast enough.
• 23:06
  Speaker 2: They use GPUs graphic processing units, which are associated with
• 23:11
  Speaker 2: video games for most people, right, you need a good
• 23:14
  Speaker 2: graphics card to run your video game, I guess, but
• 23:20
  Speaker 2: they the reason that for AI data centers that they
• 23:23
  Speaker 2: use GPUs is because they're really good at parallel processing.
• 23:26
  Speaker 2: They can run a bunch of different operations at once.
• 23:29
  Speaker 2: So you're like, cool, you just throw a GPU in
• 23:31
  Speaker 2: a data center and you can run an AI. No,
• 23:34
  Speaker 2: you need hundreds of thousands of these things strung together,
• 23:38
  Speaker 2: and instead of like a CPU running like a couple
• 23:41
  Speaker 2: of servers or something like that. At data center. All
• 23:45
  Speaker 2: of them are strung together to form one giant supercomputer
• 23:49
  Speaker 2: that the AI operates on.
• 23:52
  Speaker 1: Yeah, like chat GPT itself was trained on twenty thousand
• 23:55
  Speaker 1: of these GPUs a GPU, you know, the sort of
• 24:00
  Speaker 1: the biggest name in the game. There's a couple, but
• 24:02
  Speaker 1: the biggest one obviously is the Navidia. But the Navidia
• 24:06
  Speaker 1: H one hundred that is the standard right now. If
• 24:09
  Speaker 1: you look this thing up, it it fits in your hand.
• 24:13
  Speaker 1: It's not like some gigantic thing. H twenty thousand of
• 24:17
  Speaker 1: them linked together or one hundred thousand of them linked together.
• 24:20
  Speaker 1: Who knows how many, you know, hundreds of thousands are
• 24:22
  Speaker 1: eventually going to be linked together to end the world.
• 24:26
  Speaker 1: That's where all the power comes from, like you were saying,
• 24:29
  Speaker 1: But it's you know, it's just a little rectangular handheld
• 24:33
  Speaker 1: thing that's like, oh, that looks like something that maybe
• 24:36
  Speaker 1: came out of a computer. And Navidia is what are
• 24:40
  Speaker 1: their stock jumped over a couple of years, like nine
• 24:43
  Speaker 1: hundred percent over twenty twenty three and twenty twenty four
• 24:46
  Speaker 1: something like.
• 24:47
  Speaker 2: That, Yeah, nine hundred percent increase.
• 24:51
  Speaker 1: Yeah, and we'll talk about why all of this is
• 24:54
  Speaker 1: super like scary and dangerous because it really is.
• 24:57
  Speaker 2: Well, yeah, if you want a really good explanation of
• 24:59
  Speaker 2: this about and like you said, how many GPUs you
• 25:01
  Speaker 2: string together before we end the world? Nate Sores and
• 25:06
  Speaker 2: Eliezer Yukowski In that book, I keep referencing that I
• 25:08
  Speaker 2: think everybody should read. If anyone builds it, everyone dies.
• 25:12
  Speaker 2: About the current state of AI. They talk about this
• 25:16
  Speaker 2: in depth, but in a really understandable way. It's really fascinating.
• 25:21
  Speaker 2: But that's essentially one of the things they say is
• 25:23
  Speaker 2: like we keep stringing together tens and tens of thousands
• 25:26
  Speaker 2: more GPUs, that just makes the supercomputer smarter and smarter
• 25:31
  Speaker 2: and more capable. And eventually, what's going to happen. We're
• 25:35
  Speaker 2: going to reach some point potentially where we just put
• 25:38
  Speaker 2: that extra last GPU in there and all of a sudden,
• 25:41
  Speaker 2: the balance is tipped and the thing becomes super intelligent.
• 25:44
  Speaker 1: That's right. Also a time for me because you're always
• 25:47
  Speaker 1: too shy to to plug the end of the world
• 25:48
  Speaker 1: with Josh Clark, your fantastic limited series, of which AI
• 25:53
  Speaker 1: is one of the central focuses or one of was
• 25:55
  Speaker 1: it eight things?
• 25:57
  Speaker 2: Ten? Well, there was ten episodes.
• 26:00
  Speaker 1: Ten episodes, right, thanks baby? Well but one of the
• 26:03
  Speaker 1: episodes was just like you talking about Jimmy Buffett Records.
• 26:06
  Speaker 2: And that's right.
• 26:07
  Speaker 1: You had to lighten the mood yep Should we take
• 26:11
  Speaker 1: a break or should we keep going for a minute.
• 26:13
  Speaker 1: Let's keep going for a minute, okay, because you talked
• 26:16
  Speaker 1: about investment, and you know, if you have the money
• 26:18
  Speaker 1: to open one of these things, and that's what these
• 26:20
  Speaker 1: tech companies are doing, like to perhaps they're great peril.
• 26:24
  Speaker 1: At some point we'll see. Microsoft has invested eighty eight
• 26:27
  Speaker 1: billion dollars in data centers just in twenty twenty five.
• 26:32
  Speaker 1: Amazon is pledged over the next fifteen years one hundred
• 26:34
  Speaker 1: and fifty billion dollars. And Google and Meta together about
• 26:39
  Speaker 1: you know, not working together, but they are expected to
• 26:41
  Speaker 1: spend about seven hundred and fifty billion dollars just on
• 26:44
  Speaker 1: equipment over the next two years. And Stanley Morgan says
• 26:49
  Speaker 1: to Morgan, Stanley, what I say, Stanley Morgan, Yeah, I
• 26:53
  Speaker 1: think we should leave that in there.
• 26:54
  Speaker 2: Okay.
• 26:55
  Speaker 1: Stanley says, Hey, you know, guys.
• 26:58
  Speaker 2: Maybe you might like Stanley Kamma boar.
• 27:00
  Speaker 1: Yeah, Stanley Kumo Morgan. Over five years between twenty five
• 27:04
  Speaker 1: and twenty thirty, Morgan, Stanley says about three trillion dollars
• 27:10
  Speaker 1: is going to be spent just on the data centers
• 27:13
  Speaker 1: about I mean, half of which is the hardware and
• 27:15
  Speaker 1: half of which is just building these things.
• 27:17
  Speaker 2: Yeah, just in what the next four years. Yeah, So
• 27:22
  Speaker 2: think about it. If you're in Nvidia and you're the
• 27:24
  Speaker 2: industry leader for GPUs and everybody's like, we're gonna spend
• 27:27
  Speaker 2: one point five trillion dollars on this on the infrastructure
• 27:32
  Speaker 2: in the GPUs, you're looking pretty good down the.
• 27:35
  Speaker 1: Road, Yeah, for sure. And you know they're doing this
• 27:38
  Speaker 1: because there's a demand right now for use at least
• 27:41
  Speaker 1: because things like OpenAI and other AI creators are using
• 27:45
  Speaker 1: them like crazy. But these companies are also using them
• 27:48
  Speaker 1: for their own AI research, right.
• 27:51
  Speaker 2: Yeah. So like Xai has that Colossus machine that you
• 27:56
  Speaker 2: were talking about earlier, which is two high hundred thousand
• 28:00
  Speaker 2: GPUs strung together. I'm not sure if it's fully online yet, Tennessee. Yeah,
• 28:06
  Speaker 2: And it's just for that. It's they're not doing any
• 28:08
  Speaker 2: they're not calculating missile trajectories for the Ministry of Defense
• 28:13
  Speaker 2: or anything like that. Like, it's just for that AI.
• 28:16
  Speaker 2: And yeah, I think Meta is doing the same thing.
• 28:19
  Speaker 2: Open ai I don't think is building their own because
• 28:22
  Speaker 2: they're so in cahoots with Microsoft. I think they run
• 28:25
  Speaker 2: their stuff on Microsoft's data centers. But yeah, yeah, if
• 28:28
  Speaker 2: you have an AI essentially right now, which means like
• 28:31
  Speaker 2: God and everybody, you probably have your own data center
• 28:35
  Speaker 2: dedicated to it.
• 28:36
  Speaker 1: Yeah. I And this isn't some some moral stand I'm
• 28:41
  Speaker 1: taking by saying that I have never used AI. And
• 28:44
  Speaker 1: trust me, I know that every part of my life
• 28:47
  Speaker 1: is now touched by AI, so I am inadvertently using.
• 28:49
  Speaker 2: It touched by an AI, that's right.
• 28:53
  Speaker 1: But I've never I've never used like you know, chatbots
• 28:56
  Speaker 1: or or large language models or inning like that, just
• 29:01
  Speaker 1: mainly because I'm just I'm fine doing things like they
• 29:04
  Speaker 1: are for now, and not in a luddite sort of way.
• 29:07
  Speaker 1: I just everything's going along great for me in my
• 29:10
  Speaker 1: job and how I live my life, so I just
• 29:12
  Speaker 1: I don't have a need for it.
• 29:13
  Speaker 2: I do the same thing. And I think also both
• 29:15
  Speaker 2: of us are like if somebody else wants to do
• 29:18
  Speaker 2: it the other way, that's fine. Like we're certainly not
• 29:20
  Speaker 2: gonna criticize them or be crimudgeony about it or say
• 29:24
  Speaker 2: that you know that's stupid.
• 29:27
  Speaker 1: Right, But as you'll see. You know, and again this
• 29:31
  Speaker 1: isn't yucking someone's yum, but everyone should know what they're
• 29:35
  Speaker 1: a part of, and that's part of what the episode
• 29:36
  Speaker 1: is about.
• 29:37
  Speaker 2: That's right. Yeah, no, I totally do. Before we take
• 29:41
  Speaker 2: a break. I think it's a small kind of side issue,
• 29:45
  Speaker 2: but it's worth pointing out that it sucks because these
• 29:49
  Speaker 2: Nvidia chips are so in demand from these massive companies.
• 29:53
  Speaker 2: It has driven the price for just the average Nvidia
• 29:57
  Speaker 2: graphics card sky high. So if you're a gamer and
• 30:01
  Speaker 2: you're like trying to improve your system, like you pay
• 30:05
  Speaker 2: way more than you used to for the same graphics
• 30:08
  Speaker 2: card that you could have bought for like a quarter
• 30:11
  Speaker 2: of the price, you know, a couple of years ago.
• 30:14
  Speaker 1: Yeah, and I wasn't even looking like I didn't even
• 30:17
  Speaker 1: know that you could just buy This is how little
• 30:18
  Speaker 1: I know about all this before. This was like, could
• 30:21
  Speaker 1: you just buy a Navidia Gpu? But I was just
• 30:24
  Speaker 1: researching the size and like what do these things look like?
• 30:27
  Speaker 1: And it, you know, it's one was on eBay for
• 30:29
  Speaker 1: twenty thousand dollars and I was like, oh my god,
• 30:31
  Speaker 1: I didn't Oh I didn't know that was the deal?
• 30:33
  Speaker 2: Is that right?
• 30:35
  Speaker 1: Yeah, and I don't know if that's accurate. I don't
• 30:37
  Speaker 1: know anything about it, so I could easily be corrected
• 30:39
  Speaker 1: on all this, but that's what the Internet told me.
• 30:43
  Speaker 2: Okay, well the Internet never lies.
• 30:46
  Speaker 1: Oh one thing before we break real quick, because we
• 30:48
  Speaker 1: did promise a little UK specific stuff and I didn't
• 30:50
  Speaker 1: want to short shrift our brit listeners or Kyle. The
• 30:55
  Speaker 1: UK is right now like the third largest nation for
• 30:58
  Speaker 1: data centers. The US is first, I think Germany a second,
• 31:01
  Speaker 1: and they signed what was called the Tech Prosperity Deal
• 31:04
  Speaker 1: with the giants, the tech giants of the United States,
• 31:07
  Speaker 1: And right now Microsoft has announced a thirty billion dollar
• 31:11
  Speaker 1: investment in UK data centers, and I think like one
• 31:15
  Speaker 1: hundred new AI data centers are planned in the UK
• 31:18
  Speaker 1: at this point moving forward.
• 31:19
  Speaker 2: Yeah, and I saw there's at least one in Wales
• 31:21
  Speaker 2: that's being smartly done. They took an old radiator factory
• 31:26
  Speaker 2: plant campus and they're revitalizing that as AI data center.
• 31:32
  Speaker 2: So it does sound like I get why the UK
• 31:35
  Speaker 2: is doing it, but there's a lot of people in
• 31:37
  Speaker 2: the UK and elsewhere who are like, these are not
• 31:39
  Speaker 2: this is not a good investment for local governments or
• 31:42
  Speaker 2: even national governments. There's a big problem with all this,
• 31:46
  Speaker 2: Like there is a AI boom going on. Data centers
• 31:50
  Speaker 2: are just one part of it. Like people are throwing
• 31:52
  Speaker 2: money at AI like it's nineteen ninety nine, and a
• 31:57
  Speaker 2: lot of people are like, there's another it's not a
• 32:00
  Speaker 2: bubble this time, but it's a AI bubble. Yeah. One
• 32:04
  Speaker 2: of the reasons why some people are pointing to it
• 32:06
  Speaker 2: as an AI bubble is that it's just not clear
• 32:12
  Speaker 2: how much money is going to be made from AI
• 32:14
  Speaker 2: and when that's going to start. Yeah. I think the
• 32:18
  Speaker 2: Financial Times called Open AI a money pit with a
• 32:21
  Speaker 2: website on top. Yeah, not great, no, because people are
• 32:25
  Speaker 2: just pumping money into this stuff, but they're not getting
• 32:28
  Speaker 2: they're not seeing results from it, not yet. It's not
• 32:32
  Speaker 2: necessarily a bad bet that AI is going to completely
• 32:35
  Speaker 2: revolutionize the world and like revolutionize economies and going to
• 32:39
  Speaker 2: make some people a lot of money, but there's just
• 32:43
  Speaker 2: no clear path to it right now, which makes some
• 32:45
  Speaker 2: people nervous.
• 32:46
  Speaker 1: Yeah, there's about five percent, just five percent of pilot
• 32:50
  Speaker 1: AI programs right now in business secure returns on their investments,
• 32:56
  Speaker 1: you know, like they make the money.
• 32:58
  Speaker 2: But Stanley Morgan is predicting revenues of a trillion dollars
• 33:03
  Speaker 2: by twenty twenty eight, that's.
• 33:05
  Speaker 1: What they're saying. I mean, we'll see Navidia. I mean,
• 33:08
  Speaker 1: Kyle's also keen to point out that there's sort of
• 33:11
  Speaker 1: a circular economy within all this going on. That's a
• 33:14
  Speaker 1: little bit like troubling maybe because Navidia is investing in
• 33:18
  Speaker 1: open AAI, but that depends on their purchase of those
• 33:21
  Speaker 1: Navidia chips. So, you know, everyone from you know, just
• 33:26
  Speaker 1: people who are smarter than us as far as the
• 33:28
  Speaker 1: stuff goes, are warning people right down to the IMF,
• 33:31
  Speaker 1: the International Monetary Fund are flashing the warning signs saying
• 33:36
  Speaker 1: like this could be you know, it could make a
• 33:39
  Speaker 1: trillion dollars by twenty twenty eight, or it could like
• 33:41
  Speaker 1: wreck the global economy.
• 33:43
  Speaker 2: Yeah for sure. Yeah, we have no idea, although I
• 33:49
  Speaker 2: have seen people argue against it, that say like this
• 33:52
  Speaker 2: is nothing like Yeah, a lot of these AI companies
• 33:55
  Speaker 2: are probably overinflated, but it's nothing like it was with
• 34:01
  Speaker 2: like the two thousand and eight meltdown or the dot
• 34:03
  Speaker 2: com bubble, Like this is we're a lot we're a
• 34:07
  Speaker 2: lot more seasoned, or investors are a lot more seasoned.
• 34:09
  Speaker 2: Than they were before. The problem is one of the
• 34:12
  Speaker 2: problems is that the financing is expected to come in
• 34:16
  Speaker 2: large part from private credit, which is essentially an investment
• 34:21
  Speaker 2: vehicle for investors to go lend money to say like
• 34:25
  Speaker 2: companies that want to build data centers. Right, and this
• 34:29
  Speaker 2: is largely unregulated. It's very shadowy. We don't know how
• 34:33
  Speaker 2: many how much debt exists in the world on private
• 34:37
  Speaker 2: credit because they don't have to report this stuff. And
• 34:41
  Speaker 2: you know, as we learn from the two thousand and
• 34:42
  Speaker 2: eight meltdown, when there's like a massive speculation among finances
• 34:48
  Speaker 2: that involves debt's that can go really bad.
• 34:53
  Speaker 1: Yeah for sure. And speaking of going bad, I guess
• 34:57
  Speaker 1: we're at the sort of environmental piece of this whole thing.
• 35:00
  Speaker 1: And this is what I was talking about when I
• 35:01
  Speaker 1: said that, you know, people should just be aware of
• 35:03
  Speaker 1: what they're taking part in. And again this is not
• 35:05
  Speaker 1: to shame anybody who uses AI for their job or
• 35:09
  Speaker 1: just to make funny fake videos, but but you know,
• 35:13
  Speaker 1: everyone is sort of tied together to make this what
• 35:16
  Speaker 1: it is. Who's using that stuff, And I get if
• 35:19
  Speaker 1: someone says, like, hey, if I quit this thing, it's
• 35:21
  Speaker 1: not going to make any difference. But that's sort of
• 35:23
  Speaker 1: the the age old. Like, you know, if I don't
• 35:26
  Speaker 1: recycle my ten can, my aluminum can, ten cans, my
• 35:30
  Speaker 1: aluminum cans, then it's not gonna make that big of
• 35:33
  Speaker 1: a difference. But the idea of everyone getting together to
• 35:36
  Speaker 1: do something for the common good, that's where change happens,
• 35:39
  Speaker 1: or where negative change happens. So as far as AI
• 35:45
  Speaker 1: data centers go, the main you know, aside from just
• 35:49
  Speaker 1: you know, the land use and everything else in the
• 35:52
  Speaker 1: hardship on the local economies and towns in certain ways
• 35:55
  Speaker 1: that we're going to get to, it's really just a
• 35:59
  Speaker 1: sucubus of electricity and water usage. Yeah, psychobus is not
• 36:05
  Speaker 1: the right word.
• 36:06
  Speaker 2: No, but it makes sense. It's like a bunker down.
• 36:10
  Speaker 1: Yeah, But I say psychibis to mean just like a
• 36:12
  Speaker 1: bottomless pit. But I know that's not what it means,
• 36:14
  Speaker 1: by the way.
• 36:15
  Speaker 2: A giant sucking thing, right right, And it is. It's
• 36:18
  Speaker 2: sucking tons of electricity and water up. Like some of
• 36:21
  Speaker 2: these AI data plants use the same amount of electricity
• 36:25
  Speaker 2: as a town of fifty thousand yeah, and about the
• 36:28
  Speaker 2: same amount of water is a town of fifty thousand people.
• 36:31
  Speaker 2: This is a data center we're talking about, and it's
• 36:34
  Speaker 2: not even necessarily an AI data center. Just any hyper
• 36:37
  Speaker 2: scale data center uses a ton of electricity and water.
• 36:43
  Speaker 2: The reason it uses water is because all of these processors,
• 36:47
  Speaker 2: the CPUs that are doing all this work, and just
• 36:50
  Speaker 2: all of the networking that's going on with it, it's
• 36:52
  Speaker 2: generating heat, and computing happens faster when it's cooler. So
• 36:57
  Speaker 2: to keep the place cool they use evaporative cooling, where
• 37:00
  Speaker 2: they funnel waste heat air through wet pads essentially, like
• 37:06
  Speaker 2: they just buy old mattresses and doze them with water
• 37:10
  Speaker 2: and then they run the heat through there and through
• 37:12
  Speaker 2: evaporative cooling, it cools it off. It uses a little
• 37:17
  Speaker 2: electricity than air cooling, but it uses water a lot
• 37:22
  Speaker 2: of water.
• 37:23
  Speaker 1: Yeah, I mean, I assume most people know this. But
• 37:26
  Speaker 1: like your laptop has a tiny fan in it, Like
• 37:28
  Speaker 1: every computer in the world has a little fan in
• 37:31
  Speaker 1: it that cools it down. So when you've got all
• 37:33
  Speaker 1: this stuff together, you know it's going to generate tons
• 37:35
  Speaker 1: and tons of heat. That was the whirrying of the
• 37:38
  Speaker 1: server room that you used to sleep in. Those were
• 37:41
  Speaker 1: all fans, you know, And you know there's some other
• 37:44
  Speaker 1: sounds coming but mostly those fans trying to cool everything down.
• 37:48
  Speaker 1: We've got a lot of stats here that are pretty
• 37:49
  Speaker 1: eye popping. But there are eleven roughly eleven thousand data
• 37:53
  Speaker 1: centers around the world. Most of these are not AI obviously,
• 37:57
  Speaker 1: but they're the most you know, robust sort of u
• 38:00
  Speaker 1: users of the energy. But they use between one and
• 38:03
  Speaker 1: one point five percent, which it doesn't sound like a lot,
• 38:06
  Speaker 1: but of the entire world's electricity usage. I know on
• 38:11
  Speaker 1: planet Earth goes to data centers right now, and in
• 38:13
  Speaker 1: certain places like Ireland, data centers use about twenty percent
• 38:17
  Speaker 1: of the country's electricity.
• 38:19
  Speaker 2: Yeah, and if you dive into different places around like
• 38:22
  Speaker 2: the world where data centers are like that's collectively, right,
• 38:26
  Speaker 2: all of them in Iran and all of them in
• 38:27
  Speaker 2: the world. If you kind of zoom into the towns
• 38:30
  Speaker 2: where these things are located, there's well, there's something called
• 38:35
  Speaker 2: Data Center Alley in northern Virginia outside of DC, where
• 38:40
  Speaker 2: there's this huge concentration of large data centers, probably the
• 38:45
  Speaker 2: biggest concentration in the world. Those data centers use about
• 38:49
  Speaker 2: the same amount of electricity as sixty percent of all
• 38:52
  Speaker 2: the households in the state of Virginia.
• 38:55
  Speaker 1: Yeah, here's another one. By twenty thirty, they're predicting. This
• 38:59
  Speaker 1: is Barclay's Bank is predicting that data center energy use
• 39:03
  Speaker 1: in the United States would make up about thirteen percent
• 39:06
  Speaker 1: of the entire electricity demand of the United States. And
• 39:11
  Speaker 1: Meta has there. They all have silly names, but they're
• 39:14
  Speaker 1: data centers called Hyperion. They're all you know, one was
• 39:18
  Speaker 1: where are the where's that list? They're all these kind
• 39:20
  Speaker 1: of sci fi sounding names.
• 39:23
  Speaker 2: Yes, Stargate, Yeah, Jupiter, Prometheus. God, I'm sure all of
• 39:28
  Speaker 2: those nerds are like, what do you mean silly?
• 39:31
  Speaker 1: If I opened up a data center, I'd call it
• 39:32
  Speaker 1: Old Bessie.
• 39:34
  Speaker 2: Bessie's I hope so bad that somebody's listening to this
• 39:39
  Speaker 2: and they open a massive hyper scale data center named
• 39:42
  Speaker 2: Old Bessie.
• 39:43
  Speaker 1: That would be great. But Meta's Hyperion data center will
• 39:46
  Speaker 1: consume by the time it's finished, about five gigawatts. And
• 39:49
  Speaker 1: if you're like, what's five gigawatts? That is about half
• 39:54
  Speaker 1: of the peak load.
• 39:55
  Speaker 2: Of all of New York City, the most that it
• 39:57
  Speaker 2: can possibly it can possibly be demanded.
• 40:00
  Speaker 1: Right, Yeah, the very toplope probably, I guess in New
• 40:02
  Speaker 1: York City on the hottest day of the year, with
• 40:06
  Speaker 1: all the lights on at night or something. Yeah, and
• 40:10
  Speaker 1: that Rocketfeller tree just they just there's a summer version.
• 40:13
  Speaker 2: That puts it over the edge that's right blackout. So
• 40:17
  Speaker 2: you can imagine that when you're using all this electricity
• 40:20
  Speaker 2: and using all this water, if you're starting to build
• 40:24
  Speaker 2: these massive data centers, you're looking for places that have
• 40:27
  Speaker 2: like cheap land, cheap electricity, and because electricity is often
• 40:32
  Speaker 2: more expensive than water, they'll go to places. They'll build
• 40:36
  Speaker 2: them in places that are like water scarce, that have
• 40:39
  Speaker 2: cheap electricity. I'm the premise that, like, we're a massive
• 40:43
  Speaker 2: multinational corporation, we can push around this little county and
• 40:49
  Speaker 2: use up all of their water and what are they
• 40:51
  Speaker 2: gonna do? Nothing?
• 40:52
  Speaker 1: Yeah, And I mean that's literally happening. There's one right
• 40:55
  Speaker 1: here in Georgia, in Newton County. It's a metadata center
• 40:58
  Speaker 1: that's using tenth of the local water use. And like
• 41:03
  Speaker 1: you said, water is a is a resource that isn't infinite.
• 41:06
  Speaker 1: We've talked about the dangers in the future of like
• 41:08
  Speaker 1: you know, perhaps the wars of the future will be
• 41:10
  Speaker 1: fought over water, and this could get us there. I
• 41:14
  Speaker 1: think in Phoenix, Arizona, you know, known for their abundant water.
• 41:19
  Speaker 1: Meta and Microsoft use seven million gallons of water every
• 41:23
  Speaker 1: single day for their data centers.
• 41:25
  Speaker 2: Yeah, every day, you said.
• 41:28
  Speaker 1: Every day, seven million gallons of water.
• 41:30
  Speaker 2: That's insane. Yeah. And when I saw this, I was like, oh,
• 41:33
  Speaker 2: here we go. In the UK, data centers used ten
• 41:37
  Speaker 2: billion leaders of drinking water every year. L I. T R. E. S. Yeah,
• 41:45
  Speaker 2: that's right. Uh.
• 41:46
  Speaker 1: But you know you mentioned some of these towns. Not
• 41:50
  Speaker 1: only are some there like using let's say ten percent
• 41:52
  Speaker 1: of the local water here in Newton County in Virginia
• 41:56
  Speaker 1: where data center Alley is, some of these places are
• 41:59
  Speaker 1: like some of these towns are running out of water,
• 42:01
  Speaker 1: Like they go to turn on their water and water
• 42:03
  Speaker 1: doesn't come out because of this.
• 42:04
  Speaker 2: Well. Plus also, like we talked about how gamers are
• 42:07
  Speaker 2: getting the short end of the stick when it comes
• 42:11
  Speaker 2: to buying graphic cards because they are in such high demand,
• 42:15
  Speaker 2: same thing happens with electricity. So in addition to this
• 42:18
  Speaker 2: data center coming to town and using up all your water,
• 42:20
  Speaker 2: they're also jacking up your electricity prices because there's only
• 42:24
  Speaker 2: so much that your local electrical company can produce, So
• 42:28
  Speaker 2: because of supply and demand, your price is going to rise,
• 42:31
  Speaker 2: and I guess around Data center Alley in Northern Virginia,
• 42:36
  Speaker 2: electricity prices have increased two hundred and sixty seven percent
• 42:40
  Speaker 2: since twenty twenty. And that also is affecting Maryland, which
• 42:45
  Speaker 2: is getting little to no benefit from Data center Alley
• 42:48
  Speaker 2: and is just helping pay the price for it. This
• 42:51
  Speaker 2: is subsidization of these data centers, like they are subsidized
• 42:56
  Speaker 2: in just about every single way you can imagine.
• 42:59
  Speaker 1: Yeah, for sure. And if you say like, oh, well sure,
• 43:02
  Speaker 1: but they create jobs, right, so that's great for the
• 43:04
  Speaker 1: local economy. Kyle gives an example here of Northumberland, England.
• 43:10
  Speaker 1: There's a ten billion pound data center there or I
• 43:13
  Speaker 1: guess it's coming and you'd think, oh, great, that's that's
• 43:16
  Speaker 1: going to employ probably like five thousand people, right, it's
• 43:19
  Speaker 1: going to employ four hundred people with full time jobs.
• 43:22
  Speaker 2: Yeah, a ten billion dollar or ten billion pound data
• 43:25
  Speaker 2: center four hundred jobs. Because these things are so efficient
• 43:29
  Speaker 2: and everything is just so advanced, they don't really need
• 43:33
  Speaker 2: that many people to keep an eye on it. Right. Plus, Also,
• 43:37
  Speaker 2: the money from that data center, if they're not going
• 43:40
  Speaker 2: to spread it around the UK, it's going to flow
• 43:42
  Speaker 2: right back to the US, to the parent company.
• 43:46
  Speaker 1: Oh yeah, for sure. And you know, we also didn't
• 43:49
  Speaker 1: point out that a lot of these these energy grids
• 43:53
  Speaker 1: like are literally going to buckle under pressure at some point,
• 43:56
  Speaker 1: like they're not built for this.
• 43:57
  Speaker 2: Yes, and we're not so we're I know, it sounds
• 44:00
  Speaker 2: like we're just like and this and that. How terrible
• 44:03
  Speaker 2: are data centers like There's they're they're incredibly important and
• 44:08
  Speaker 2: they support an amazing array of really great stuff, right
• 44:13
  Speaker 2: and they they are the foundation that the next expansion
• 44:17
  Speaker 2: of the digital economy and the world culture are going
• 44:20
  Speaker 2: to grow on. Like, they're incredibly important, but they have
• 44:24
  Speaker 2: a lot of problems with them that need to be addressed.
• 44:27
  Speaker 2: They're not being addressed because every government from like the
• 44:31
  Speaker 2: local city council up to the leaders of the free world,
• 44:37
  Speaker 2: like are just giving these people whatever they want. That's
• 44:41
  Speaker 2: what's going on now. There's no checks going on at all,
• 44:45
  Speaker 2: right now, that's the problem.
• 44:47
  Speaker 1: Yeah, And that's that's because the flow of money is
• 44:50
  Speaker 1: so great at this point to a certain segment of
• 44:54
  Speaker 1: the population. Only they're protecting their their own investment, you know,
• 45:00
  Speaker 1: they're watching their own backsides.
• 45:01
  Speaker 2: That's definitely I would say ninety nine percent of it.
• 45:04
  Speaker 2: But I think there's also chuck, a little factor of
• 45:06
  Speaker 2: like g whiz, Like these these titans of the AI
• 45:12
  Speaker 2: industry are good at like razzle, dazzling elected officials into
• 45:17
  Speaker 2: doing whatever they want by I think making them feel
• 45:20
  Speaker 2: included in this new frontier. Essentially, I think there's a
• 45:25
  Speaker 2: certain element of that.
• 45:26
  Speaker 1: I think you're probably right. It's hey, maybe it'll all
• 45:30
  Speaker 1: work out great.
• 45:30
  Speaker 2: Sure, it probably will. It usually does astoundingly, it usually
• 45:35
  Speaker 2: does work out well.
• 45:37
  Speaker 1: True as far as the world hasn't ended.
• 45:39
  Speaker 2: That's exactly what I mean. Yeah, yeah, yeah, So I
• 45:43
  Speaker 2: think that's it. We said yeah like four or five
• 45:45
  Speaker 2: times in secession. I think we accidentally triggered listener, ma'am.
• 45:51
  Speaker 1: That's right. This relates to our history of the BBC episode,
• 45:55
  Speaker 1: and this is from Erica, and Erica says, hey, guys,
• 45:59
  Speaker 1: I really love the episode and left me reflecting on
• 46:01
  Speaker 1: how I've come to understand the country through both the
• 46:03
  Speaker 1: content the BBC produces and the people's reactions to the BBC.
• 46:07
  Speaker 1: But more recently, my work as an academic has enabled
• 46:09
  Speaker 1: me to be involved in creating programs for the BBC
• 46:12
  Speaker 1: across TV, radio and online. Because there's one awesome fact
• 46:16
  Speaker 1: about the BBC that wasn't included. For over fifty years,
• 46:19
  Speaker 1: the BBC has partnered with the Open University OU, which
• 46:22
  Speaker 1: specializes in accessible and distance education. The partnership started in
• 46:26
  Speaker 1: the nineteen seventies to provide learning at scale, including facilitating
• 46:30
  Speaker 1: university level lectures at night on public television. Today, the
• 46:33
  Speaker 1: partnership facilitates access to academic consultants to co produce high
• 46:38
  Speaker 1: quality and form content across platforms, including some of the
• 46:43
  Speaker 1: David Attenborough.
• 46:44
  Speaker 2: Nature stuff nice.
• 46:46
  Speaker 1: Additionally, the Open University creates supplementary materials to enable people
• 46:50
  Speaker 1: to continue their learning journey and explore topics in more detail. So,
• 46:55
  Speaker 1: whether viewers or listeners realize it or not, this partnership
• 46:57
  Speaker 1: enables the public to benefit from special US knowledge and
• 47:01
  Speaker 1: accessible ways. And that is from Erica from the Open University,
• 47:06
  Speaker 1: who is a professor of medical anthropology.
• 47:08
  Speaker 2: Oh wow, that's an awesome Erica. You got to send
• 47:11
  Speaker 2: us some topic ideas too, totally right up your alley
• 47:15
  Speaker 2: and congratulations. That's pretty neat making stuff in conjunction with
• 47:18
  Speaker 2: the BBC. That's gotta be a neat high water mark,
• 47:21
  Speaker 2: you know, agreed, And I think, Chuck, I'm curious to
• 47:25
  Speaker 2: see if we go look at our account, we'll see
• 47:28
  Speaker 2: a little line item from open University and one from BBC.
• 47:33
  Speaker 1: Well, it would be like seven pounds or something. I
• 47:35
  Speaker 1: don't know the exchange rate, right.
• 47:36
  Speaker 2: That sounds about right? All right, great, well, thanks again, Erica,
• 47:40
  Speaker 2: and please do send us some medical anthropology ideas because
• 47:43
  Speaker 2: that just sounds like it'll knock our socks off. And
• 47:46
  Speaker 2: if you want to be like Erica and try to
• 47:47
  Speaker 2: knock our socks off, good luck, you can send it
• 47:51
  Speaker 2: off to us at stuff podcast at iHeartRadio dot com.
• 47:58
  Speaker 2: Stuff you Should Know is a production of iHeartRadio. For
• 48:01
  Speaker 2: more podcasts my heart Radio, visit the iHeartRadio app, Apple Podcasts,
• 48:05
  Speaker 2: or wherever you listen to your favorite shows.