September 4, 2026 • 64 mins
Modern civilization depends on data centers - these facilities, which can be as small as a single room or as large as gigantic, stand-alone compounds, house the servers and computers that make the digital world spin 'round. However, as Ben, Matt, and Noel discover in tonight's episode, the newest iteration of 'AI' data centers triggered a massive public outcry: they're drinking up and contaminating water. They're a profound drain on electricity networks. They're also fueling a ton of local and federal corruption. So what's the future of these data centers? Most disturbingly, what happens when all the water is gone?
See omnystudio.com/listener for privacy information.
Listen
Watch
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
From UFOs to psychic powers and government conspiracies, history is
riddled with unexplained events. You can turn back now or
learn the stuff they don't want you to know. A
production of iHeartRadio.
Speaker 2 (00:26):
Hello, welcome back to the show. My name is Matt.
Speaker 3 (00:29):
My name is Noel.
Speaker 4 (00:30):
They call me Ben. We're joined, as always, with our
super producer, Dylan, the Tennessee pal, Fagan. Most importantly, you
are you. You are here. That makes this the Stuff
They Don't Want You To Know 10101 to all our
fellow code-based friends and neighbors.
Speaker 3 (00:50):
Is that like Morse code? Binary. Binary.
Speaker 4 (00:53):
Binary code?
Speaker 2 (00:54):
Yeah.
Speaker 4 (00:54):
All the codes, most of the codes, and a little
bit of code switching. This is something we've talked about
in Strange News. We've talked about in Listener Mail. We
always wanted to do an episode on this.
Speaker 3 (01:07):
Data centers, you guys, AI booms. Is it a boom?
Speaker 4 (01:12):
Is it a bubble about to burst? What's happening with
all this construction? I remember, Matt and Noel, we were
very recently talking about how people in our own home
state of Georgia have been affected by construction of data centers,
looming huge facilities. I think they call them hyper centers.
Speaker 5 (01:33):
I'll tell you who's kind of ahead of the curve,
who often is in like sort of satirizing this is
Ari Aster, the filmmaker whose most recent film, Eddington, I
don't know, flew under the.
Speaker 3 (01:45):
Radar for a lot of people.
Speaker 5 (01:46):
It was maybe not everyone's cup of tea, but it
very much satirizes this kind of like middle America thing
sort of COVID driven paranoia and social media kind of
feedback loops. But it also at the center of it all, it's,
I think somewhere in the Midwest or maybe it's in Texas,
there's this data center that's going to be built and
its promises around all the jobs that it's going to create.
(02:07):
And delightfully, it is called the Magikarp Data Center, which
I believe is a type of Pokemon.
Speaker 2 (02:14):
Yes, a useless Pokemon.
Speaker 3 (02:16):
That's the joke.
Speaker 5 (02:16):
Thank you.
Speaker 3 (02:17):
I didn't even clock that, dude. That's incredible. Okay. I'm sorry, Matt.
I just got excited. That's a thing?
Speaker 5 (02:22):
So it's considered the joke, like neutered, nothing, do nothing Pokemon.
Speaker 2 (02:27):
But if you can make it survive, if you have
enough potions in your deck to make it survive, then
it becomes this crazy Gyarados Pokemon. creature that is like
a water snake dragon thing.
Speaker 3 (02:40):
Wait a minute. Water snake, dude.
Speaker 4 (02:44):
Yeah.
Speaker 5 (02:45):
I think Aster knows all of this. And then, and
the water thing is not for nothing. And this movie
came out, I guess a couple of years ago at
this point, but it was definitely. The conversation was being had,
of course. He didn't just pluck this out of thin air.
But it certainly hadn't escalated to the degree that it
has in terms of, you know, what he's depicting in
that film.
Speaker 4 (03:04):
We could say a lot of people saw this construction
on the horizon. Right. I mean, Astor, I think we
can argue, certainly did. There are a lot of fans
of data centers, especially the higher ups.
Speaker 3 (03:18):
People who make data centers.
Speaker 4 (03:19):
Yeah, people who make data centers. They will say, look,
this is a necessary part of innovation. We're building the
plane while we're flying it. It's not perfect, but we've
got to break some eggs to make an omelet. The
issue is, as we discussed a little bit earlier, that critics,
a growing number of a disaster waiting to happen or
(03:42):
indeed already a disaster and, uh, a nest of conspiracies
waiting to hatch. So, uh, I kept listening to, uh,
Yassine Bay or most deaths, uh, song new world water,
which just like Aster was, uh, quite prescient. Uh, maybe
we can all take a sip of some cold H2O
(04:04):
while we're still allowed to do that. And, uh, Right
into the horizon together.
Speaker 2 (04:12):
Mine was canned by Celsius and a bunch of crap
was put in it.
Speaker 4 (04:18):
Well, it's still, it's still liquid and we can still
have it for now. So what's going on with data
centers and water. We'll be right back. Here are the facts.
Speaker 2 (04:34):
Okay.
Speaker 4 (04:35):
Guys, I can't be alone in this. I think anybody
reading the recent news stories about the AI construction boom
is going to inevitably have sort of a sour taste
in their mouth when they hear that phrase data center.
It's almost become a thought terminating cliche, right?
Speaker 5 (04:52):
Well, I mean, I guess the previous version of this
that we would probably all be more familiar with in
terms of like cloud storage and these big internet companies
is would be storage right would be like all of
these server farms i guess so this is not the
same as that this is has a different goal maybe
they can be combined or these purposes can be linked
(05:14):
and they can be on the same site or in
the same facility but when we say data centers we're
not talking about just hard internet capacity storage for people's
photos and you know all of that kind.
Speaker 2 (05:25):
Of stuff no we're talking about processing exactly which again
it is funny because it is a giant Venn diagram
when it comes to the data centers that exist and
have existed for a long time and these new things.
But there is quite a bit of history when it
comes to mass computing centers as computers have been around
(05:47):
for a time.
Speaker 1 (05:48):
Oh yes.
Speaker 4 (05:48):
Yes, sir. At its most basic level, a data center
is what we're describing. It's just a physical location where,
It has servers and computer systems and networking equipment. And
like you were saying, Matt, this idea dates back to
the 1940s. And in the 1940s, the machines that were
(06:09):
part of data centers were basically kaiju. They were like
mech size. They were huge. What was that excellent film
about Alan Turing, The Imitation Game? They have a great
depiction of just the enormity. Of these different machines. These
different machines.
Speaker 5 (06:29):
Clockwork type machines in a way. And then I think
they would move on and evolve to the punch card
type systems. Or maybe there's some crossover between those two eras.
But in my mind, the earliest days that you're talking
about here, I mean, we're talking almost like these giant
mechanized metal monstrosities.
Speaker 3 (06:47):
Yeah. Not to be too alliterative.
Speaker 4 (06:49):
Yeah. As we know, technology has a bit of a
recurrent pattern. So miniaturization gets into the game, right? This
reduces the amount of real estate we need. Then we
have these smaller computers. We call them servers. For a while,
civilization thought miniaturization was, you know, nothing new and data
(07:12):
centers are going to continue the same historical pattern of
so much other technology. But there was that plot twist.
cloud computing, right? Early 2000s, like you were saying, it
rocks the game.
Speaker 5 (07:23):
Well, Ben, I mean, just because we got everything shrunk
down a little, maybe logic might dictate that, oh, that
means we don't need as much space for them.
Speaker 3 (07:30):
Not so.
Speaker 5 (07:31):
We need space for them forever because we need continuous growth,
the ability to continually upscale and hold more stuff and
process more blips, you know? And yeah, sure, the things
are getting smaller and smaller, but then demand is getting
larger and larger.
Speaker 2 (07:49):
And let's talk about that cloud computing just quickly, just
to make sure we all understand. One of the biggest
benefits of that stuff is that your machine that you
are using as an end user no longer has to
do the massive number of computations per whatever second floppy.
It's now this huge data center is doing all that stuff.
(08:10):
Your computer is just accessing that data. data center, that information,
that cloud.
Speaker 5 (08:15):
I do want to clarify, though, because, I mean, to me,
when I think of cloud, I think of cloud storage. Like,
I don't know a situation where I'm offloading the processing
of my machine to the cloud.
Speaker 2 (08:26):
Video games are doing that all the time. Certain apps
that you're accessing are doing that.
Speaker 4 (08:32):
All the time.
Speaker 3 (08:32):
See, I didn't know that exactly. That's fascinating.
Speaker 4 (08:35):
And it's huge for companies, too. Like you're saying, Matt,
we have, with these cloud services now, Let's say you
start a company and you can't afford your own warehouse
of computers. Now you have a virtual parking space and
you can pay per use. You can scale up. You
(08:56):
can scale down. You don't have the headache of maintenance.
You don't have the constant need to keep your tech
current because you're always going to be racing against obsolescence, right?
You're always going to need that new update. It's crazy.
You know, it makes me think, guys, this is this
is an old, old bit of how stuff works lore.
(09:18):
It makes me think of our little server room in
our old office where there's a true story, folks. Occasionally
we had to bring in industrial fans if something went
wrong and we had to hold the door open. and
have a ton of fans in the room just to
stop everything from melting down.
Speaker 2 (09:37):
Well, we didn't have solid state drives at the time.
We had two Pegasus systems that were just mirrored on
each other. And these are regular old hard drives that
still spin, you know, and they hold, they held a
number of gigabytes, not a number of terabytes or anything
like that. And they held the entirety of our video
library as a video team.
Speaker 4 (09:59):
And thus, our future. Yes.
Speaker 2 (10:02):
But then all of those servers were running HowStuffWorks.com, which
is crazy to think about.
Speaker 4 (10:07):
Yes. Yeah. And there were these cooling systems, power supplies,
backup generators, fire suppression systems. I just, I keep remembering
those times where we would, We would all look at
each other and then look at our legendary IT guy, Izzy,
the way that you would look at a general in
(10:29):
the front lines of a war. And the guy was like,
we need more fans.
Speaker 2 (10:33):
Well, dude, and that whole, oh, crap, Comcast went down,
so we had to switch over to AT & T
for the rest of the day or vice versa and
all that kind of weird stuff. It's a massive undertaking
for the human beings that make any– size server room
or data center function. And just from a hardware standpoint
(10:56):
and God, it's the cooling that ends up being the
thing that is most important to humanity's future and to
the data centers themselves.
Speaker 3 (11:05):
Yeah.
Speaker 4 (11:05):
And we literally could not make this show without data centers, folks.
We got to be transparent here. You probably couldn't watch
this show without those. I do suppose technically that, Guys,
we could go back to old school audio tape or Vital.
We could distribute the podcast that way, mail out the cassettes.
It would take a while.
Speaker 2 (11:27):
Amazon Web Services all day. Thank you, Amazon, I suppose.
Speaker 4 (11:33):
Oz, AWS.
Speaker 3 (11:34):
A very trepidatious thank you.
Speaker 4 (11:36):
Yes, a civil tip of the cap to thee. So
what we're saying is these things are not inherently bad. Indeed,
they are necessary for civilization as we know it. But
the controversy comes into play when we ask what a
particular data center is doing and how it affects the
communities and environment around it, right, in its continual function.
(12:02):
So that's what you're going to be hearing about in
the headlines today. overwhelmingly, the controversy is centering on data
centers and artificial intelligence, which is not a real thing.
We're talking about LLMs.
Speaker 5 (12:16):
So unlike some of these more conventional centers of yesteryear
that we've been talking about through the evolution of these
facilities and the Venn diagram of like, is it cloud
storage versus cloud computing or When we're talking about AI,
we're definitely, I think, talking about a combination of both,
because obviously all of that stuff that you enter into
(12:37):
AI and upload has to go somewhere. So it has
to live in a storage facility of some sort, but
then it has to be processed using software. I think
a lot of times they're NVIDIA or what you might
consider graphics card type processors stacked up in racks.
Speaker 3 (12:55):
Exactly. Thank you, Ben. They're the most powerful ones out there.
Speaker 5 (13:00):
And I think we're going to get to it, but
just the monopoly that companies like NVIDIA have over these
types of things and how beholden to companies like that
this whole– AI.
Speaker 3 (13:10):
Boom, uh, bust bubble, uh, is.
Speaker 5 (13:13):
Um, so we are talking about large language models that
learn by constantly analyzing petabytes of text and code and images. Again,
those require high capacity storage. Uh, this requires also extreme
computing power to sift through these massive, massive data sets
(13:33):
at scale, meaning mega, mega fast. Um, as well as,
like I was talking about, specialized hardware that you can't
just bolt onto your typical PC or laptop. So an
AI data center is very different from the typical data
centers we were talking about because the chips that they use.
Speaker 3 (13:53):
Are much, much, much more powerful.
Speaker 4 (13:56):
Yeah, yeah. Your CPU, that's going to have sequential tasks,
and it's going to have what we call cores, a
few powerful main character cores. But a GPU, a graphics
processing unit, has a bunch of complicated, repetitive math that
it does all the time. I keep doing, I guess
I'm doing Benedict Cumberbatch references because it's like a beautiful mind.
(14:20):
That's what your GPU is doing.
Speaker 5 (14:22):
No, an imitation game, right? Benedict Cumberbatch is an imitation game.
Russell Crowe is in a beautiful mind.
Speaker 3 (14:27):
Oh, yeah.
Speaker 4 (14:27):
Those are two different people. Just like a CPU and
a GPU. Yes, we got there. These repetitive complex calculations
are happening all at once, simultaneously, back and forth, distributed
across thousands of other smaller cores. And the one rule
(14:48):
is this stuff can never stop. So these, we call
them frontier AI models, right? They got thousands of processors.
They're running 24 hours a day, seven days a week,
months at a time. And it's so much energy, you guys.
It is an astounding and disturbing amount of energy required here.
(15:10):
A single, quote-unquote, AI server rack is using something like
60 kilowatts, just as some coffee table math. That is
about six times higher than the energy requirements for a
standard server rack. This means that if you put all
these AI servers in one place, depending on the size
(15:32):
of the compound, it can consume as much power as
hundreds of thousands of human homes.
Speaker 5 (15:39):
Well, and we did a story recently on Strange News
about a whole region or a municipality where I think
it was like the electric company was like saying, we're
no longer going to serve customers. We're only going to
serve AI. And that just is like evidence of how
insane the energy consumption is for these facilities. So it's
(15:59):
a twofold problem. We're mainly addressing I think water today,
but it is also a resource problem in terms of
power and like, you know, like maybe depriving regular folks
of energy. Yeah.
Speaker 4 (16:11):
But also, you know, if you're ever cold and, maybe
you could, in theory, run to a data center because...
Warm your hands.
Speaker 5 (16:19):
Eating a bean, a single bean, with your fingerless gloves
as you warm your hands over the hot, hot GPU.
Speaker 4 (16:26):
Tell stories of when water was free. It's like you
were saying there, Matt. It goes back to cooling. These
things generate so much heat because of all the work
that they're doing. Part of the heat problem is compounded because...
The way that these models work requires you to have
(16:47):
thousands of computers, not just operating simultaneously, but operating pretty
close together with a super fast network. So you don't
get bottled traffic jams. Now you get a bunch of
pretty hot things on their own that are hanging out
and being hot together. I'm sure there's a reality show
reference we could make there.
Speaker 5 (17:06):
That would be a great name for a reality show.
Just being hot together, you know, being hot together.
Speaker 3 (17:13):
It's about influencers, maybe. Rather than being hot apart.
Speaker 5 (17:18):
Yeah, I think that would work. That could be the
subtitle of Love Island.
Speaker 4 (17:21):
I like how we're going for the LCD, the lowest
common denominator there. Find us for more reality show pitches. So,
keep going back to this. Matt, as you were saying,
engineers are still trying to figure out the increasingly difficult
craft of keeping all this stuff cool and it's more
(17:43):
and more uh difficult every day.
Speaker 2 (17:46):
You should we go over the types of cooling that
have been used and are currently being used for cooling
like in the way water is used in.
Speaker 5 (17:53):
Those and i just remember when i was younger building
pcs i mean everything's fan cooled then.
Speaker 4 (18:00):
Right ginormous hvac setups or uh one of the the
big primary ones right now would be Evaporation cooling, uh,
which we've got to get into, and then what they
call closed loop cooling. Uh, we also see facilities that
have HVAC systems and cooling stuff at multiple points.
Speaker 5 (18:24):
And you can get a closed loop cooling system, like
for home, you know, I mean, people build these types
of GPU rack systems in their homes, uh, to mine cryptocurrency,
for example. That was the first time I ever heard
GPUs being used this way. And I guess, correct me
if I'm wrong, but isn't this sort of a repurposing?
Like they are graphic processors. So they were originally designed
(18:45):
just for like gaming and like really fast, you know,
responses for, you know, high quality monitors and stuff. Or
not monitors necessarily, but yeah, I guess it is monitors.
It's like, you know, you have to have a certain
speed of graphics card to run a certain type of
game on PC. And then they sort of repurposed these
or were they always kind of meant for.
Speaker 3 (19:05):
All of the things.
Speaker 4 (19:06):
I like repurpose. Let's go with evolved as well. What
we're saying here, folks, is these are thirsty, thirsty dudes
and critics are concerned. A growing number of critics is
a weirdly bipartisan issue. A growing number of critics are
concerned that the more water these things take, the less
(19:26):
water there will be for humans. We'll be right back.
Here's where it gets crazy. Let's break down some general concepts, guys.
We talked about hot hanging out together. How do we
get cool? Like the AI, the water, we all know
(19:49):
these stats. by this point. A hundred word email written
by AI uses the same amount of water as a
freshwater water bottle, right? They're not drinking water the same
way a living organism would. They're taking it for cooling.
And we were talking a little bit off air. We
got to get into the cooling.
Speaker 2 (20:10):
Yeah, and that's if they're using evaporative cooling, right? That's
when a 100-word email uses that amount of water. That's
evaporating water out of the usually municipal drinking water system
because most of these, or at least a major percentage
of them, are hooked up to local water supplies. And
it just evaporates it. Because it's literally evaporating the water
(20:32):
to cool down the equipment.
Speaker 4 (20:33):
Yeah, cooling towers, right?
Speaker 2 (20:35):
Yeah, that's one way to look at it. But it's
often written as evaporative cooling if you're seeing it in
one of these documents that's being brought up by, let's say,
a contractor that is constructing a new data center. You'll
see evaporative cooling in there to describe how they're going
to cool off their stuff. You might also see closed-loop
water cooling, which is the one that's described as the
(20:56):
coils that run through the whole system. with usually cold water, again,
from a municipal system in the same way. These are
good because they can reduce the amount of water you're evaporating. Well,
they circulate.
Speaker 5 (21:09):
So it's not like it goes away, right?
Speaker 2 (21:13):
Yes, you can cut consumption, as we're talking about there,
evaporation by up to 70% compared to evaporative cooling with
that system. However, you're introducing things into the water potentially
with that. And then the water does eventually at some
point get flushed, usually back into the system where it
would go to like a wastewater treatment plant, which has
(21:35):
ups and downs as we're going to discuss. You get
the water back, but now there's a lot of stuff
in it. There's also the air cooling.
Speaker 3 (21:43):
Yeah, exactly.
Speaker 2 (21:44):
There's also air cooling, free cooling. That's another term you'll see.
That's just bringing in cold air from an environment. Let's
say if you're in the north of the US or
in Europe somewhere, often you just pump in cold air.
And the last one, which is really interesting, that I
haven't seen much. I don't know if you guys have
seen much of this happening in new data center construction.
It's an emerging technology, but it's called direct liquid cooling
(22:08):
and immersion. You might see that written. That is where
the systems themselves are actually immersed in a liquid that
is non-conductive. It's like a coolant that the GPUs and
stuff live inside. very much like the matrix with humans
living inside those little things.
Speaker 4 (22:27):
Or like dunking, dunking someone in a bucket and not
letting them swim up to the surface.
Speaker 1 (22:34):
Yeah.
Speaker 4 (22:35):
I'd love for us to talk a little bit more
about that evaporating technique because the evaporating water can, can
in fact accelerate the cycle of thirst because it's, it
leaves behind, as we're saying, it becomes a water plus,
if you will, because it leaves behind in the evaporation
(22:56):
process bacteria, minerals, other chemical contaminants that could corrode or
clog equipment, which means that if you are a data center,
you have to constantly flush out that old water, put
it somewhere, and add in more fresh water to prevent
long-term damage over time. And like you were saying, Matt,
(23:17):
we know that Emergent technology aside, most data centers now
use water in some form. I found some stats that
said 75 to 90 percent of data centers, not just
in the U.S., but across the planet, use some kind
of water-based cooling as their primary method, which means that
(23:40):
right now, like, 10 to maybe 25% are completely water-free.
Speaker 2 (23:46):
And yeah, Ben, and I saw that too. And it was,
it was mostly based on location.
Speaker 4 (23:50):
Yes.
Speaker 2 (23:50):
Temperature of where you're at. I also saw Ben from
the university of Georgia here, the UGA, their college of
agricultural environmental sciences is they have a big part in
water usage with data centers, at least in our state.
And according to them, Around 57% of all data centers
(24:14):
use water from a potable water source. That means water
that you can drink.
Speaker 4 (24:21):
New York is drinking it. New World Water, California is
drinking it. I got to stop singing that song.
Speaker 3 (24:26):
But yes.
Speaker 2 (24:26):
57%?
Speaker 5 (24:26):
Mm-hmm.
Speaker 2 (24:30):
It seems like you could use water that is not
drinkable for this stuff.
Speaker 3 (24:34):
Gray water, right?
Speaker 2 (24:35):
Poop water. Yeah.
Speaker 3 (24:36):
Well, I mean, truly. No, but gray water. We always
talk about that.
Speaker 5 (24:38):
We used to talk about that.
Speaker 3 (24:40):
And I just remember there was.
Speaker 5 (24:42):
An eco push for a while with new construction where
you could get tax subsidies for recycling gray water and
using it for not drinkable water. Again, gray water, by
the way, is like water that comes from the shower,
not the toilet. Like it's like not the poop.
Speaker 3 (24:57):
Water and it's not pure clean water. It's somewhere in between.
Speaker 5 (25:01):
And you can use it for, I think, even like
watering plant crops.
Speaker 3 (25:06):
And things like that.
Speaker 5 (25:07):
But I do wonder if there's a world where that
could be incorporated into this system.
Speaker 1 (25:13):
Yes.
Speaker 3 (25:13):
No one talks about gray water anymore.
Speaker 4 (25:14):
Well, that's part of the issue. The reason it has
to be potable or freshwater is because you have to...
Part of the efficacy of the cooling comes from the
purity of the water. So if it has... If it
has the wrong kind of additives, we'll talk about the
chemical additives in a second. I thought water.
Speaker 5 (25:33):
Would be water when it comes to its ability to
be cool.
Speaker 3 (25:37):
You know, that's interesting.
Speaker 4 (25:38):
Yeah, if it has the wrong kind of additives, it
can spell disaster for the data center, even if the
people are going thirsty. Before we continue, though, guys, our pal, Tennessee.
That's very good. Dylan, you just laid this awesome bar out.
Could you say it on air?
Speaker 3 (25:54):
If you're making slop, you should use slop water. Exactly.
Mop water, slop water.
Speaker 4 (26:00):
Mop water, slop water, not your mom and pop water.
Speaker 3 (26:03):
Well, it is your mom and pop's water. Yeah.
Speaker 4 (26:08):
This is the issue, right? These conventional data centers are
already taking a lot of work to be pooled, to
be powered. They're drawing as much electricity as 10,000 to
25,000 households, according to often it's like the International Energy Agency,
which kind of knows what they're talking about. The issue
(26:31):
is that the newer guys, the data centers you're reading
about in your Washington Post or your Futurology subreddit or
what have you, those hyperscale data centers that are LLM-focused,
they can use as much power as 100,000 homes. Shout
out to Meta.
Speaker 2 (26:51):
I'm going to give that 10 f * * *
that's on the f * * * that scale.
Speaker 4 (26:57):
Yeah, let's give it two Louisianas, or two New Orleans
as well.
Speaker 2 (27:02):
Oh, yeah, two full New Orleans.
Speaker 3 (27:04):
You got a lot of gray water in New Orleans.
Speaker 2 (27:07):
This is super important. There is a sliding scale that
you can look at when it comes to a data
center's water usage and electricity usage, and you can see
that often... As electricity increases, water usage decreases a little
bit or vice versa. And you can really see the
(27:29):
relationship between those two things. And either way, a massive
data center that is going to use the equivalent of
100,000 homes. is going to also be using quite a
bit of water, but maybe not as much. It just
depends on which system they go with. If they go
with some of the closed looped things and the evaporative systems,
(27:52):
which is, again, the evaporative stuff is a little older,
but it is still a major player right now. It
is just the thirstiest, most power hungry thing that currently
exists or is even imagined on this planet.
Speaker 4 (28:07):
Yeah, it's the sin of gluttony. writ large, right? There
is no satiation for the data center. A data center
is not going to say, I've had enough to drink.
Or it's not going to say, I'm good. You can
turn the lights off. It's more and more.
Speaker 2 (28:22):
I'm thinking of characters I've seen in sci-fi movies that
are like the huge, just bloated creatures that are.
Speaker 5 (28:31):
Hexus from uh fern gully the last rainforest or the
kind of goo monsters in princess mononoke uh yeah and
and like you know in that i think it was
spirited away spirited away as well but princess mononoke had
a bit of a eco theme to it you know
about conservation i guess and there were those things that
were just kind of eating everything like eating the animals.
Speaker 3 (28:54):
In the woods and All of that stuff. And as
you feed this thing, and it's just, give me more.
It goes, oh, here's a picture of a cat. Oh,
here's a picture of a cat. He's a hungry boy.
Speaker 4 (29:09):
It reminds us also of that line from Apocalypto, where
people are talking about how humans have a hole inside them.
Speaker 3 (29:20):
The Matt Gibson movie?
Speaker 2 (29:22):
The Bible talks about that.
Speaker 4 (29:24):
But despite that, it's a good film.
Speaker 3 (29:26):
I never saw it.
Speaker 5 (29:27):
Yeah, that's the one where they don't speak. They speak
an indigenous language.
Speaker 4 (29:32):
It's a banger. Is it good? It's worth a watch.
I enjoyed it thoroughly. So we know these comparisons are.
by the way, folks, all on point, and we didn't
spoil anything. Check out those references. The reason we're saying
we're rating it two New Orleans in addition to 10
f * * * that's on the f * * *
(29:53):
that scale, which I love, Matt, we're referencing Meta's Hyperion.
There's another meta data center in Wyoming that at the
time of its construction was projected to use more electricity
than every home in the state of Wyoming put together.
Speaker 2 (30:15):
Yeah, Project Cosmo is, is, was, it still is messed up.
You can read all about a crazy thing they did
where they allegedly, hey, we have to say this, they
allegedly introduced some pretty gnarly bacteria into the municipal water system.
You can read all about it in The Guardian from
(30:36):
this past July.
Speaker 5 (30:38):
Where do these additives get introduced?
Speaker 3 (30:42):
I'm confused.
Speaker 5 (30:43):
Are they pumping stuff into the water to make it
more cool? Is it just stuff that's picking up along
the way because of the pipes or whatever? I'm very
confused as to why is it coming back with so
much gook in it?
Speaker 1 (30:56):
Oh, yes.
Speaker 4 (30:57):
Yeah. That goes back to the idea of the water footprint.
So we set that up real quick. We know that, okay,
so we know that the centers themselves, like we're saying,
are thirsty. They need a lot of water. They also
need a lot of electricity. And what doesn't often get
reported as much as it should is that electricity is
(31:17):
coming from fossil fuel plants or nuclear power plants. And
those folks are also infamous for needing tons of water, right,
for their own cooling systems or to generate steam. So
the water footprint changes. gets bigger and bigger and bigger
uh and the water for the living things uh gets
(31:39):
tougher and tougher to find well i.
Speaker 2 (31:41):
Mean nestle's gotta have its water so you know look
man if the data centers don't suck it up then
then nestle gets it that's how the contract that's what
it states i think it's like section 10 part a i.
Speaker 4 (31:56):
I can't i can't believe we got this far without
referencing without referencing, there will be blood. Obviously, they're drinking milkshakes, right?
This depletes the local aquifers. It could dry out wells.
And Noel, you raise such an excellent point. It is
not just pure water coming in and then pure water
(32:17):
rolling out. The stuff going in, depending on the cooling system,
is going to be mixed with chemicals like chlorine dioxide, bromine,
sodium hypochlorite, which I was unfamiliar with. And that's good
for the center. Because it mitigates or stymies the growth
of algae and mold and bacteria. But what's good for
(32:39):
the data center is not always what's good for the
community in which that center is based.
Speaker 2 (32:44):
The way I think about it, Noel, is if you've
got a river, you've got flowing water that is just
continually moving, right? And because it's continually moving, you might
have buildup of things like bacteria in certain areas of
that moving water or that lake.
Speaker 5 (32:58):
I mean, yeah, it's carrying materials. And, you know, if
you don't filter that out and or treat it, it
can start gunking up your own works. And that is
not efficient for these machines that need to be cool
24-7 forever and all time.
Speaker 2 (33:12):
But in the case of Project Cosmo, you introduce some
of that water into a closed system, even though you're
introducing a whole bunch of things into it. You're still
working with pipes and systems and pumps that are, you know,
there are areas that will be exposed. differently, which may
build up, let's say bacteria. And now that bacteria doesn't
(33:34):
matter how much bromine and crap you're pumping into it.
You've just got quite a bit of bacteria in your
wastewater system, or maybe even your wastewater container that you're
holding that stuff on site. Um, and then you flush,
if you just hit the button and flush it back
into the system, into a river, into a municipal system,
guess what? You've got bacteria now.
Speaker 4 (33:55):
Yeah, check out our earlier episode on the Turkey Point
salt plume. It's for an example of how that water
can get out of control. And don't think we're off
the hook closed-loop cooling systems. You guys use propylene glycol,
adethylene glycol, which, surprise, surprise, also can end up in
(34:17):
a community's water. Big shout-out to PFA's. fluorinated F gases,
whatever those are without being chemist, we can assure you
they're not great for people. That's why they're not, you know,
in the, uh, in the food pyramid, which is also.
Speaker 2 (34:33):
Well, and as you're saying, build up in the systems,
like heavy metals, right? The stuff that we looked at
in Flint, Michigan and small towns across the world with
lead pipes and other issues, you know, with the potable
drinking water for the humans.
Speaker 4 (34:48):
Copper zinc as well yeah because it's uh it's it's
just oh it's bad guys.
Speaker 5 (34:57):
Remember remember that jeopardy category potent potables.
Speaker 4 (35:00):
I do i do man that's all i got sorry
no that's great i always wanted to go on my first.
Speaker 5 (35:06):
Exposure to that word potable and in this case it's
just talking about a thing you drink yeah ones that
are strong yeah.
Speaker 2 (35:12):
Remember that other category Existential threats to all of humanity.
Speaker 3 (35:17):
Yeah, that was a banger.
Speaker 5 (35:19):
Always a real good time on Jeopardy. Shout out to
Ken Jennings, by the way, who we used to work with.
It's so crazy seeing that guy. He's the new Alex Trebek.
Speaker 3 (35:28):
Good job, man.
Speaker 4 (35:29):
We can assure you Ken's a super nice dude as well.
Speaker 5 (35:32):
I think his podcast is still out there that we did.
I made the theme for it, ripping off Doctor Who
pretty aggressively.
Speaker 3 (35:40):
Doctor Who?
Speaker 4 (35:42):
Yeah, exactly.
Speaker 3 (35:43):
Very good.
Speaker 4 (35:44):
Very good.
Speaker 5 (35:45):
But it's out there and it's very good. It's him
and this guy, John Roderick, talking about weird stuff. And
they did stuff like about like how Starlings made it
here because of Shakespeare and stuff like that.
Speaker 4 (35:55):
Yeah, that's a wild story. We also have to tell
you that the problems go past water with these with
these construction projects. And we talked about this a little
bit on Strange News or Listener Mail recently, which you
can find in our audio feed, folks. Similar to how
the prison industry and now the concentration camp industry targeting immigrants,
(36:20):
similar to how those industries will find small, economically struggling
towns and then come to them the same way that
guy sold monorails in that episode of The Simpsons. They'll say, hey,
let us build this thing.
Speaker 3 (36:36):
Right.
Speaker 4 (36:36):
Cast aside your NIMBY. Let us build this thing and
we will give you so many jobs. This will be
such an economic boom.
Speaker 5 (36:46):
I know I've brought this up in the past, so
please forgive me for repeating myself, but it is kind
of like a firsthand observation of the way this works.
Speaker 3 (36:51):
When I was.
Speaker 5 (36:52):
Younger and I worked for public radio after the housing collapse,
I covered this small town near where I worked in Augusta, Georgia.
I'm totally forgetting the name of the small town, but
their almost entirety of their local economy was tied to
manufacturing carpet and housing kind of stuff.
Speaker 3 (37:11):
Whatever it was.
Speaker 5 (37:12):
And so basically they had something in the neighborhood of
a 100% unemployment rate. Like it was really, really, really unusual.
And they had this private prison company that had them
on the list of sites that they would maybe build
this prison. And I went and covered this town hall,
like the monorail guy, where the folks from Corrections Corporation
(37:34):
of America, that I think has rebranded at this point,
basically pitched this thing to the townsfolk and talked about
all the different types of jobs it would generate, janitorial,
corrections officers, all this stuff. And everyone was just so like,
it was like this lifeline of, And then they built
it somewhere else.
Speaker 4 (37:51):
Right.
Speaker 5 (37:52):
I mean, but this is the same kind of thing.
Only I think in this case of the prisons, it
would have made some jobs. This one, I'm not so sure.
What are the jobs outside of the actual construction that
I think it's a bit of a bill of goods.
Speaker 4 (38:06):
Yes.
Speaker 2 (38:07):
Just on a wider scale there. Think about the complexity
of that problem. You just described Noel. We, these humans are,
living in a rather remote area, need jobs so badly
that they are willing to imprison human beings.
Speaker 3 (38:23):
They were stoked about it. They were stoked about it.
I mean, I'm sorry, you know, yeah.
Speaker 2 (38:28):
But that's what all of us, especially here in the
United States, just kind of go along with every day.
And there's not much we can do about it. The
prison industrial complex ends up being a massive part of
how the USA can make its GDP.
Speaker 3 (38:45):
Well, and let me just rephrase.
Speaker 5 (38:46):
They weren't stoked about participating in locking up humans.
Speaker 4 (38:50):
Of course.
Speaker 5 (38:50):
But it was just so removed from that, what that
is on a like large scale that it was just
like job.
Speaker 3 (38:58):
This equals job. We are stoked about job.
Speaker 5 (39:01):
The consequences or whatever the thing is, be damned.
Speaker 2 (39:05):
Because human must have job, right? That is, that's the thing.
Human must have money. Eat for eat, for food, for water,
for everything.
Speaker 5 (39:12):
To ignore the gnarly knock-on consequences because you're so desperate
that you're like, please, yes, in my backyard. And let me,
you know, be the guy that runs the freaking whatever
it is, pulls the switch.
Speaker 2 (39:25):
It's so similar to the existential threat that... the United
States in some way believes it faces with the need
for these AI data centers. If not AI data center,
then we will be destroyed by others with AI data center.
So must build AI data center. more AI data center,
most powerful AI data center, or death.
Speaker 5 (39:47):
Well, there's a space race element to it. And it almost,
you know, it really truly is an arms race as well.
It's not only about progress and about like societal, you know,
whatever superiority.
Speaker 3 (39:58):
Ben, I'm really curious your thoughts on this.
Speaker 5 (40:00):
It is a kind of arms race because of what
these things power and what they mean in terms of
the evolution, what they will do down the line.
Speaker 1 (40:08):
Yeah, yeah.
Speaker 4 (40:08):
Let's The arms race aspect is is a wonderful thing
for us to look into further. We also have got
to tell you, Bill of Goods is the perfect way
to describe the disparity between the jobs that are promised
and the jobs that are delivered. And no, you are
absolutely right that there is a lot of front end
(40:31):
employment in the construction industry. And a lot of politicians,
a lot of real estate developers make their vigorous. So
much opportunity to grip.
Speaker 3 (40:40):
So much opportunity to grip. So much.
Speaker 4 (40:42):
And often, as we'll see in specific cases, the public
of a community might not know the full extent of
what's going on. But we found this fascinating research that
dates back to 2024, where the investigators found that throughout
the United States, as few as 23,000 people held permanent
(41:08):
jobs at data centers. As few as 23,000, that is
about 0.01% of all employment across the country. And at
the same time, this industry consumed more than 4% of
all the electricity used in the United States. And those
numbers are becoming more increasingly distressing. The electricity usage is rising.
(41:34):
Automation means that you have fewer and fewer people, actual people,
who have to work at these places.
Speaker 2 (41:42):
Well, and the tech billionaire's dream is to only increase
that 23,000 people working in data centers by, you know,
just a teeny bit, just as much as they need, right?
Speaker 3 (41:51):
Yeah, their goal is to decrease it.
Speaker 2 (41:54):
Well, no, it's to increase it with the number of
new data centers, just a tiny bit, and then Get
rid of all the other jobs.
Speaker 3 (42:01):
That's what I'm saying.
Speaker 5 (42:02):
Guys, I mean, really quickly, if no one wants to
hear spoilers for the recent Wuk Chan Park film, No
Other Choice, just fast forward like 30 seconds. But it
does such an incredible job of spelling out this conundrum
where it's about this legacy industry, paper manufacturing that in
Japan is like this art. You know, there's like the
old school guys like tap on the rolls of paper
with a stick. And it's this whole almost like built
(42:22):
into the tradition of fine Japanese rice paper making by
hand and all this stuff.
Speaker 3 (42:27):
And then automation comes in. And then everyone gets laid off.
And then the.
Speaker 5 (42:31):
Whole story is about all of these qualified people fighting
for the one job at the automation center. That is
the guy that oversees what once would have been done
by 100, 150, 200 employees on a line. It's just
now dudes pacing through the facility in the dark because
they're saving money on the lights and just checking boxes
(42:52):
on an iPad. And I'm not even going to give
you the plot because in order to even get to
that place, some real gnarly shenanigans take place.
Speaker 2 (43:00):
Yes.
Speaker 5 (43:01):
Buying for those jobs, pitting, you know, colleague against colleague
to just scrap and, you know, ultimately, you know, do
some horrific things in order to be that guy.
Speaker 4 (43:11):
And speaking of horrific things, let's pause for a word
from our sponsors and then get into some specific cases. Folks,
we have returned and not to be too, you know,
self-centric here. but we are referencing our home state of
(43:33):
Georgia for a reason. There's an excellent and profoundly disturbing
example of how corruption can come into play over in Fayetteville, Georgia.
There's a great article you've got to read from Politico
about this. It's a fantastic summation, and it's the beginning
(43:54):
of dystopian sci-fi cinema. There are these residents in an
upscale subdivision named Annalise Park. And one day, these people,
who are pretty well-to-do, they notice that their water pressure
is not up to par. It's running incredibly low. And
these are people who are accustomed to things working out
(44:15):
for them because they are well-to-do. So they go to
the local authorities. Some of them are the local authorities.
And they're going like, hey, this is in Flint, Michigan.
What's up with our water? The county utility folks go
to investigate what's up with the water. And then they
happen upon the local data center, which on the books
(44:37):
has one industrial scale water hookup. But in practice, it
has two hookups of the same size.
Speaker 3 (44:45):
They sneaked in the second one?
Speaker 4 (44:47):
They just did a double. They did a twofer, right?
Oh my gosh.
Speaker 2 (44:51):
I remember we talked about this in May this year
on Strange News. The big deal is that nobody, Nobody
in the county knew, right? That was the big deal.
Speaker 3 (45:00):
The utility company didn't know.
Speaker 2 (45:02):
Nobody knew that they just snuck in a second thing.
Speaker 4 (45:05):
Yeah, yeah. Shout out to Quality Technology Services. What do
we say about innocuous names, folks? Quality Technology Services. There
are two-for-oners here. And they were able to get away
with this without being billed, without it being on record,
just full-on free water for quite some time. A better
(45:26):
way to say it, I would argue, is water theft.
They eventually owed, when they got caught in their grift
because they got too thirsty, they owed something like $ 150, 000
for using more than 29 million gallons of water. And
think about this, folks. Think about the money they were
(45:48):
making while they're stealing this water. Paying $ 150, 000. What a bargain. Yeah,
that's peanuts. That's a puddle.
Speaker 5 (45:56):
Well, So that's the question I have. I don't remember
the way this panned out. And I'm excited to read
the Politico piece, but guys, I can only imagine this
did not lead to any kind of genuine consequences for
this company.
Speaker 2 (46:07):
No. No consequences. I remember this. I remember writing all
this down, guys. There was a quote from somebody from
the city side who said, quote, they're our largest customer
and we have to be partners. It's called customer service. Yeah.
Speaker 3 (46:26):
That's one way of putting it.
Speaker 2 (46:27):
Just talking about how they were the most important thing,
but you know, we, and we joked on that episode
about the, the actual customers, you know, of the city
of the town of Fayetteville.
Speaker 3 (46:40):
Yeah.
Speaker 4 (46:40):
That's insane. Yeah. This Fayetteville facility is one of the
largest developments in the country. It's like 615 acres. Georgia
is home to more than 200 similar data centers.
Speaker 2 (46:55):
Oh, Ben, this is why the QTS campus was projected
to generate $ 150 to $ 200 million annually in property tax revenue.
Speaker 3 (47:06):
That's why.
Speaker 4 (47:06):
Yeah, that's why. That's the customer service. So think about it, folks.
If you want to go in with us on a
$ 150, 000 investment to make millions, how would you say no
to that? I also, we should mention that Meta Center
in Wyoming is in addition to the concerns about wastewater,
(47:27):
just to put a, put a bow on this one,
a bacterial bow, uh, we're talking about something called Cupria
Vetus Gilardi. Cupria Vetus Gilardi. Excuse my trivia voice there,
but it was, uh, it came into the, uh, the
Cheyenne water system because entirely because of allegedly, uh, meta's
(47:51):
activities at that data center. And they had to, the,
The thing that panicked the leadership of Cheyenne was that
they had to shut down parts of the water system
for the people, but not really shut down the data center.
Speaker 3 (48:10):
Customer service, baby.
Speaker 2 (48:11):
There's some messed up stuff in there. And I would
just point out, if you read anything, I've been, guys,
I've been seeing the best stuff in the Guardian just
because it is still independent. I can still access it,
but you can see the names of contractors that the
companies that are going in to build the data centers,
like Goat Systems LLC, that was a part of Project
(48:32):
Cosmo there in Cheyenne, or Fortis, a general contractor there,
and just how these companies are making so much money
from the construction part, and then also seeing in Congress
how the construction part ends up becoming one of the
major water pollution flashpoints in for local residents where these
(48:56):
data centers are going in, because you're talking about all
kinds of crazy manipulation of the land you have to
do to get those flat services to make an 800,000
square foot facility, 1.2 million square foot facility, and how
some of those explosions, some of that stuff ends up
getting into the water system where private wells are, especially
(49:19):
in rural rural areas.
Speaker 3 (49:22):
People have wells.
Speaker 5 (49:23):
It goes down to like, how do you own the water?
And they don't own the land. It's that milkshake argument, right?
They're like drinking it from places that they don't own.
It doesn't belong to them.
Speaker 4 (49:33):
Yeah, like we say, it's draining the local aquifers. The
wells are going dry. There was a fantastic BBC piece
about that, which, again, not to be too Georgia-centric, is
a story from a person based in Georgia.
Speaker 2 (49:47):
Well, to keep it Georgia-centric, AOC back in May held
up two jars of water. We talked about it on
Strange News. This nasty-looking, had all kinds of black sediment
and just brown poop-looking water. And it came from two
private wells in Morgan County here in Georgia. And it
was because of the construction of a data center, not
(50:08):
the use of an operational data center. And, you know,
there's a lot that has been made about that. Ben,
that's where the UGA thing came in. We find out
that the University of Georgia Cooperative Extension Office handles broader
private well testing throughout Georgia. But the EPA said that
private wells are not regulated under the Federal Safety drinking
(50:32):
water act and Georgia does not regulate private well water
at the state level. So if you do live in
a rural area where one of these things is coming
and you have a private well, you're out of luck.
You just got to hope, cross your fingers and hope
that the water you drink and bathe yourself with and
feed to your children and your pets is going to
(50:54):
be okay.
Speaker 4 (50:55):
Oh, and by the way, on that note, Folks, we
know these are incredibly divisive political times here in the
United States. There is at least one issue with broad
bipartisan support of the people in power, regardless of where
they are on the political spectrum. Broadly, the establishment political
(51:19):
parties love data centers. Did you know, folks, fun fact,
both the Biden and the Trump administrations went 10 toes
down on data centers.
Speaker 2 (51:31):
Gee, I wonder why, Ben. Why would so many people
be so into data centers, especially those who, I don't know,
get donations for their campaign or exist on ad dollars?
Why would everybody be so into these AI data centers?
Speaker 4 (51:51):
So again, that's what we have to acknowledge. Ocasio-Cortez, Bernie Sanders,
for instance, they're considered, in the Overton window of discourse,
they're considered very far left and anomalous for not supporting
these data centers. The money is rolling in, right? And
campaign donations, as you said, and pork barrel projects, as
(52:13):
we used to say. Also, it helps with the creation
of a surveillance state. It's true. That's a big part
of it is we're building the eye of Sal Rotten again.
Speaker 2 (52:26):
Let's be real about the advertising industry or the industries
that are supported by advertising. Just, guys, how many AI-related
ads have come our way that we've had to vet?
Speaker 5 (52:43):
Without being too inside baseball, it is an initiative. It
is a thing that is almost like, you know, with
companies like ours or folks that are subsidized by advertising,
it's going to be a conversation internally about guys are
going to be seeing more of this kind of stuff
and you got to have a position on it. Don't
want to get left behind, you know?
Speaker 2 (53:04):
It has to be because we've mentioned before that circular
investment that's happening, the tune of hundreds of billions of dollars,
that stuff gets cycled through marketing too. And it just,
it becomes a lifeline. Like a vital part of a
company that's based on advertising. Those companies now, right now,
all of them need it.
Speaker 5 (53:26):
As traditional advertising maybe wanes a little bit and is
at dollars or maybe not what they used to be
and it's a little tougher because the pie keeps getting
divided and divided and subdivided. Yes, of course you're going
to grab onto that with both hands if you're a
CEO or the head of a marketing agency or whatever
it might be.
Speaker 4 (53:44):
Or beholden to ill-informed shareholders, perhaps. I would also point
out that a similar thing applies to some of these
construction companies. Some of these contractors are entirely dependent upon
projects like this for their continued survival. So a lot
of the people involved, or a lot of the entities
(54:07):
involved here, do feel this is a matter of survival.
And so does the public, by the way. This is
a bipartisan issue for us, the non-powerful.
Speaker 3 (54:18):
It is, but can't you guys see it?
Speaker 2 (54:19):
We are all residents of that small town you described, Noel.
Speaker 3 (54:24):
All of us.
Speaker 2 (54:25):
We are all residents of that small town going.
Speaker 3 (54:27):
Please, please, bring it here. Protect our jobs. Yes.
Speaker 4 (54:32):
All of us.
Speaker 5 (54:33):
Even in an industry like we're in, we're not immune
from concerns around AI replacing us. Or aspects of the
production of our show. And you know what I mean?
And it's like you're almost like you're having to give
in to the AI to protect yourself from the AI.
Speaker 4 (54:52):
Right?
Speaker 3 (54:53):
It's weird.
Speaker 4 (54:53):
Burn the village to save it.
Speaker 3 (54:55):
Kind of. Yeah.
Speaker 5 (54:56):
You got to play ball with them to get their
dollars because you might not be making your dollars because
they've taken your job.
Speaker 3 (55:04):
I mean, it's a weird or a burrus, right?
Speaker 4 (55:06):
Yeah, it's a real burristic for sure. I do want
to give us a bit of good news here, folks.
We said it's weird that this is a bipartisan supported
issue by the establishment, by the people in power or
the powers that be. The flip side of this, which
I find quite inspiring, is that we're also seeing a bipartisan,
(55:30):
not unanimous, push against it from the people who are
not in power. The public is fighting back. Some politicians
are listening. Please check out an excellent article by a
guy named Andrew Howard, a journalist named Andrew Howard for Politico,
talking about how this just recently was largely a local issue, right?
(55:53):
And if you're running for a political office, you probably
wouldn't mention it a bunch, but now it is a
top-button debate in some of the most sensitive races in
American politics, Ohio, Michigan, and so on.
Speaker 5 (56:09):
Politico is writing about Fayetteville. I mean, it's definitely on
the radar in a big way.
Speaker 4 (56:16):
And isn't it inspiring to realize that there are people
in this nation who will never agree on so many things,
but here people are shaking hands across the aisle and
The disagreement, the opposition to data centers and all their
consequences is one of the biggest cross-partisan issues in modern history.
(56:42):
And then, you know, the Epstein files. So again, I'm
saying the public versus the powerful right now.
Speaker 5 (56:49):
It also goes to show that the powerful are represented
on both sides of the aisle. That's why it's so
bipartisan because this is an unimpeachably beneficial thing to, you know, folks.
Speaker 2 (57:00):
Ben, I have to just say you are absolutely right.
And it is crazy to see that we do have
so much power. Like in Hall County in Gainesville, there
was this thing called Project Turbo. Not kidding, guys. That's
not a joke.
Speaker 3 (57:12):
A $ 1.
Speaker 2 (57:12):
2 billion data center that was going to consume 225,000
gallons of water per day to run its data center.
It was going to be huge. It was going to
be a massive thing for Gainesville. The reason why the
folks there in Hall County stopped it is because human beings,
people who lived in Hall County, went to the meetings
(57:34):
and shouted enough, said no enough, and they stopped it.
They stopped it from happening. That was a year ago.
A year ago in August, it was proposed, and then
by December, the human beings spoke up enough that it
killed it.
Speaker 4 (57:50):
Yes, yeah, there are more and more inspiring stories. We're
even seeing journalism describe a paradigm shift, a sea change.
As more people learn about this, you'll see articles talking
about organizations making a U-turn on the construction of data centers.
And we want to give you a tip. I'd love
(58:10):
to give you a tip. Folks, write to us if
you play along at home and tell us how this
works out. I imagine that if you can think of
someone who, disagrees with you politically on every other issue,
go to them and ask them what they think about
data centers. Tell them you think it's * * * *
(58:30):
and then tell us what they said because it might
be common ground for you to save your friendship or
make a new rapport because fighting against the common cause forever,
it has united societies and this might be the battle
of the eight.
Speaker 2 (58:49):
Dude, we could see humanity just investing billions and billions
of dollars into this thing. And what does it really
do for us? There are a couple of cool scientific
advancements that are possible through this stuff. But most of
it is crap and a money-making opportunity. And that's all
it is. It's a scheme. It's a popular scheme. to
(59:10):
make money which is being turned into the threat to
scare you of oh AI it could bring down the
whole world because a bad actor might be able to
crack all of our systems and all this stuff it's
fear it's fear for the money makers the controllers the
owners that's all it is.
Speaker 5 (59:30):
I mean, yeah, and devil's advocate, I mean, it's with
the degree to which we've experienced it just exponentially improving
in what it's capable of. Like I do see there
being something of an investment in this idea of, well,
if it's gotten here this fast, where is it going next?
But in order to get to the next phase of it,
it requires more and more investment, more and more data centers.
(59:54):
It's a weird one because it becomes this argument of
like we can't let– be left behind by history. Yeah.
Speaker 4 (01:00:01):
It also, it's flirting with the, we're right now as
a global civilization, we're on third base of sunk cost fallacy, right?
You've walked far into a dark room. You don't know
the full direction. There is this idea of keeping up
with the Joneses globally, a new kind of, as we
said earlier, an arms race, a bag of digital badgers
(01:00:24):
to pickle for the modern age. And really the, I
don't want to sound like a Luddite, but the smart
thing to happen would be a heavier regulation and maybe
even just turn them off for a second. Turn them off,
take a breath, touch some grass.
Speaker 3 (01:00:40):
Touch some grass.
Speaker 2 (01:00:42):
They're trying to scale so that everybody can use it
for everyday things. And that's the only way they can
make money is if everybody is using these things every
day for everyday purposes.
Speaker 3 (01:00:53):
Let's add it to Google Earth. That's what I'm saying.
That's the only way they can.
Speaker 2 (01:00:59):
The only way they can be profitable is to scale
in those ways. But the only way it's going to
be good for humanity is if it's scaled down and
used for very specific purposes.
Speaker 3 (01:01:09):
Yes, that's exactly right.
Speaker 4 (01:01:10):
Like you make new medicines, you predict.
Speaker 2 (01:01:13):
Yes, that's one thing. That's what you have to do.
That's what we have to do. That's what we have
to make them do. Because right now, they are going
to kill us all.
Speaker 3 (01:01:22):
Seriously.
Speaker 2 (01:01:23):
They're going to kill us all. Because they're greedy and
they want more money. And that's what's happening right now.
Speaker 5 (01:01:30):
Yeah.
Speaker 3 (01:01:30):
I'm with you.
Speaker 4 (01:01:31):
Some extreme critics love the idea we just proposed a
second ago of turning around the ship, turning stuff off
for a second, retreating from AI, finding more focus or
scoped in uses. But it's highly unlikely that that's going
to happen. The research, the construction is going to pause, right?
(01:01:52):
And the public is increasingly opposed as they learn more
about it. But Pausing doesn't mean fully stopping. And it's also,
it's troubling. It's frightening, folks, that we as a civilization
do not know the full consequences of this reality. We
have to ask ourselves, are we going to be able
(01:02:12):
in this dark room, right? We don't know the size
of the room. Are we going to be able to
understand what is in that darkness? Are we going to
be able to comprehend the problems on the horizon? Or
will we really only learn about them once it's too
late to fix?
Speaker 5 (01:02:26):
Well, this is the kind of thing that I think
when we first started hearing about it, it was only
trickling out into the public kind of consumer spaces after
being dicked around with in exactly the way that we're
probably proposing that it should go back to being dicked
around with, like, you know, in labs, you know, in
high level computing facilities that are devoted to things like SETI,
(01:02:48):
you know, or various forms of research and analytics, right?
And then, like anything, someone had the genius idea, but
what if we sold it to everybody?
Speaker 4 (01:02:58):
Let's unscrew the lid on Pandora's jar. We can't wait
for your thoughts, folks. We know we went a little
bit long on this one. Thanks for riding with us.
Cheers to everybody. Enjoy a nice glass of water while
you can. Reach out and tell us your thoughts. We
can't wait to hear from you.
Speaker 5 (01:03:17):
Yeah, you can sure do that on the internet by
reaching out to The Handel's Conspiracy Stuff or Conspiracy Stuff Show, depending.
Speaker 2 (01:03:23):
We have a phone number. It is 1-833-STD-WYTK. See, look,
it's written right here. Just call this number. That's just
for the video team to see if they'll do it.
If you want to send us an email, we are.
Speaker 4 (01:03:38):
Let's double down on Maybe you can see that on
the bottom third, too, other Chiron.
Speaker 3 (01:03:42):
Ugh, making more work for these poor guys.
Speaker 4 (01:03:44):
We are the entities who read each piece of correspondence
we receive with great interest and be well aware sometimes
the void rides back, back, back. We'll see you here
in the dark. Conspiracy at iHeartRadio.com.
Speaker 2 (01:04:16):
Stuff They Don't Want You To Know is a production
of iHeartRadio. For more podcasts from iHeartRadio, visit the iHeartRadio app,
Apple Podcasts, or wherever you listen to your favorite shows.
From UFOs to psychic powers and government conspiracies, history is
riddled with unexplained events. You can turn back now or
learn the stuff they don't want you to know. A
production of iHeartRadio.
Speaker 2 (00:26):
Hello, welcome back to the show. My name is Matt.
Speaker 3 (00:29):
My name is Noel.
Speaker 4 (00:30):
They call me Ben. We're joined, as always, with our
super producer, Dylan, the Tennessee pal, Fagan. Most importantly, you
are you. You are here. That makes this the Stuff
They Don't Want You To Know 10101 to all our
fellow code-based friends and neighbors.
Speaker 3 (00:50):
Is that like Morse code? Binary. Binary.
Speaker 4 (00:53):
Binary code?
Speaker 2 (00:54):
Yeah.
Speaker 4 (00:54):
All the codes, most of the codes, and a little
bit of code switching. This is something we've talked about
in Strange News. We've talked about in Listener Mail. We
always wanted to do an episode on this.
Speaker 3 (01:07):
Data centers, you guys, AI booms. Is it a boom?
Speaker 4 (01:12):
Is it a bubble about to burst? What's happening with
all this construction? I remember, Matt and Noel, we were
very recently talking about how people in our own home
state of Georgia have been affected by construction of data centers,
looming huge facilities. I think they call them hyper centers.
Speaker 5 (01:33):
I'll tell you who's kind of ahead of the curve,
who often is in like sort of satirizing this is
Ari Aster, the filmmaker whose most recent film, Eddington, I
don't know, flew under the.
Speaker 3 (01:45):
Radar for a lot of people.
Speaker 5 (01:46):
It was maybe not everyone's cup of tea, but it
very much satirizes this kind of like middle America thing
sort of COVID driven paranoia and social media kind of
feedback loops. But it also at the center of it all, it's,
I think somewhere in the Midwest or maybe it's in Texas,
there's this data center that's going to be built and
its promises around all the jobs that it's going to create.
(02:07):
And delightfully, it is called the Magikarp Data Center, which
I believe is a type of Pokemon.
Speaker 2 (02:14):
Yes, a useless Pokemon.
Speaker 3 (02:16):
That's the joke.
Speaker 5 (02:16):
Thank you.
Speaker 3 (02:17):
I didn't even clock that, dude. That's incredible. Okay. I'm sorry, Matt.
I just got excited. That's a thing?
Speaker 5 (02:22):
So it's considered the joke, like neutered, nothing, do nothing Pokemon.
Speaker 2 (02:27):
But if you can make it survive, if you have
enough potions in your deck to make it survive, then
it becomes this crazy Gyarados Pokemon. creature that is like
a water snake dragon thing.
Speaker 3 (02:40):
Wait a minute. Water snake, dude.
Speaker 4 (02:44):
Yeah.
Speaker 5 (02:45):
I think Aster knows all of this. And then, and
the water thing is not for nothing. And this movie
came out, I guess a couple of years ago at
this point, but it was definitely. The conversation was being had,
of course. He didn't just pluck this out of thin air.
But it certainly hadn't escalated to the degree that it
has in terms of, you know, what he's depicting in
that film.
Speaker 4 (03:04):
We could say a lot of people saw this construction
on the horizon. Right. I mean, Astor, I think we
can argue, certainly did. There are a lot of fans
of data centers, especially the higher ups.
Speaker 3 (03:18):
People who make data centers.
Speaker 4 (03:19):
Yeah, people who make data centers. They will say, look,
this is a necessary part of innovation. We're building the
plane while we're flying it. It's not perfect, but we've
got to break some eggs to make an omelet. The
issue is, as we discussed a little bit earlier, that critics,
a growing number of a disaster waiting to happen or
(03:42):
indeed already a disaster and, uh, a nest of conspiracies
waiting to hatch. So, uh, I kept listening to, uh,
Yassine Bay or most deaths, uh, song new world water,
which just like Aster was, uh, quite prescient. Uh, maybe
we can all take a sip of some cold H2O
(04:04):
while we're still allowed to do that. And, uh, Right
into the horizon together.
Speaker 2 (04:12):
Mine was canned by Celsius and a bunch of crap
was put in it.
Speaker 4 (04:18):
Well, it's still, it's still liquid and we can still
have it for now. So what's going on with data
centers and water. We'll be right back. Here are the facts.
Speaker 2 (04:34):
Okay.
Speaker 4 (04:35):
Guys, I can't be alone in this. I think anybody
reading the recent news stories about the AI construction boom
is going to inevitably have sort of a sour taste
in their mouth when they hear that phrase data center.
It's almost become a thought terminating cliche, right?
Speaker 5 (04:52):
Well, I mean, I guess the previous version of this
that we would probably all be more familiar with in
terms of like cloud storage and these big internet companies
is would be storage right would be like all of
these server farms i guess so this is not the
same as that this is has a different goal maybe
they can be combined or these purposes can be linked
(05:14):
and they can be on the same site or in
the same facility but when we say data centers we're
not talking about just hard internet capacity storage for people's
photos and you know all of that kind.
Speaker 2 (05:25):
Of stuff no we're talking about processing exactly which again
it is funny because it is a giant Venn diagram
when it comes to the data centers that exist and
have existed for a long time and these new things.
But there is quite a bit of history when it
comes to mass computing centers as computers have been around
(05:47):
for a time.
Speaker 1 (05:48):
Oh yes.
Speaker 4 (05:48):
Yes, sir. At its most basic level, a data center
is what we're describing. It's just a physical location where,
It has servers and computer systems and networking equipment. And
like you were saying, Matt, this idea dates back to
the 1940s. And in the 1940s, the machines that were
(06:09):
part of data centers were basically kaiju. They were like
mech size. They were huge. What was that excellent film
about Alan Turing, The Imitation Game? They have a great
depiction of just the enormity. Of these different machines. These
different machines.
Speaker 5 (06:29):
Clockwork type machines in a way. And then I think
they would move on and evolve to the punch card
type systems. Or maybe there's some crossover between those two eras.
But in my mind, the earliest days that you're talking
about here, I mean, we're talking almost like these giant
mechanized metal monstrosities.
Speaker 3 (06:47):
Yeah. Not to be too alliterative.
Speaker 4 (06:49):
Yeah. As we know, technology has a bit of a
recurrent pattern. So miniaturization gets into the game, right? This
reduces the amount of real estate we need. Then we
have these smaller computers. We call them servers. For a while,
civilization thought miniaturization was, you know, nothing new and data
(07:12):
centers are going to continue the same historical pattern of
so much other technology. But there was that plot twist.
cloud computing, right? Early 2000s, like you were saying, it
rocks the game.
Speaker 5 (07:23):
Well, Ben, I mean, just because we got everything shrunk
down a little, maybe logic might dictate that, oh, that
means we don't need as much space for them.
Speaker 3 (07:30):
Not so.
Speaker 5 (07:31):
We need space for them forever because we need continuous growth,
the ability to continually upscale and hold more stuff and
process more blips, you know? And yeah, sure, the things
are getting smaller and smaller, but then demand is getting
larger and larger.
Speaker 2 (07:49):
And let's talk about that cloud computing just quickly, just
to make sure we all understand. One of the biggest
benefits of that stuff is that your machine that you
are using as an end user no longer has to
do the massive number of computations per whatever second floppy.
It's now this huge data center is doing all that stuff.
(08:10):
Your computer is just accessing that data. data center, that information,
that cloud.
Speaker 5 (08:15):
I do want to clarify, though, because, I mean, to me,
when I think of cloud, I think of cloud storage. Like,
I don't know a situation where I'm offloading the processing
of my machine to the cloud.
Speaker 2 (08:26):
Video games are doing that all the time. Certain apps
that you're accessing are doing that.
Speaker 4 (08:32):
All the time.
Speaker 3 (08:32):
See, I didn't know that exactly. That's fascinating.
Speaker 4 (08:35):
And it's huge for companies, too. Like you're saying, Matt,
we have, with these cloud services now, Let's say you
start a company and you can't afford your own warehouse
of computers. Now you have a virtual parking space and
you can pay per use. You can scale up. You
(08:56):
can scale down. You don't have the headache of maintenance.
You don't have the constant need to keep your tech
current because you're always going to be racing against obsolescence, right?
You're always going to need that new update. It's crazy.
You know, it makes me think, guys, this is this
is an old, old bit of how stuff works lore.
(09:18):
It makes me think of our little server room in
our old office where there's a true story, folks. Occasionally
we had to bring in industrial fans if something went
wrong and we had to hold the door open. and
have a ton of fans in the room just to
stop everything from melting down.
Speaker 2 (09:37):
Well, we didn't have solid state drives at the time.
We had two Pegasus systems that were just mirrored on
each other. And these are regular old hard drives that
still spin, you know, and they hold, they held a
number of gigabytes, not a number of terabytes or anything
like that. And they held the entirety of our video
library as a video team.
Speaker 4 (09:59):
And thus, our future. Yes.
Speaker 2 (10:02):
But then all of those servers were running HowStuffWorks.com, which
is crazy to think about.
Speaker 4 (10:07):
Yes. Yeah. And there were these cooling systems, power supplies,
backup generators, fire suppression systems. I just, I keep remembering
those times where we would, We would all look at
each other and then look at our legendary IT guy, Izzy,
the way that you would look at a general in
(10:29):
the front lines of a war. And the guy was like,
we need more fans.
Speaker 2 (10:33):
Well, dude, and that whole, oh, crap, Comcast went down,
so we had to switch over to AT & T
for the rest of the day or vice versa and
all that kind of weird stuff. It's a massive undertaking
for the human beings that make any– size server room
or data center function. And just from a hardware standpoint
(10:56):
and God, it's the cooling that ends up being the
thing that is most important to humanity's future and to
the data centers themselves.
Speaker 3 (11:05):
Yeah.
Speaker 4 (11:05):
And we literally could not make this show without data centers, folks.
We got to be transparent here. You probably couldn't watch
this show without those. I do suppose technically that, Guys,
we could go back to old school audio tape or Vital.
We could distribute the podcast that way, mail out the cassettes.
It would take a while.
Speaker 2 (11:27):
Amazon Web Services all day. Thank you, Amazon, I suppose.
Speaker 4 (11:33):
Oz, AWS.
Speaker 3 (11:34):
A very trepidatious thank you.
Speaker 4 (11:36):
Yes, a civil tip of the cap to thee. So
what we're saying is these things are not inherently bad. Indeed,
they are necessary for civilization as we know it. But
the controversy comes into play when we ask what a
particular data center is doing and how it affects the
communities and environment around it, right, in its continual function.
(12:02):
So that's what you're going to be hearing about in
the headlines today. overwhelmingly, the controversy is centering on data
centers and artificial intelligence, which is not a real thing.
We're talking about LLMs.
Speaker 5 (12:16):
So unlike some of these more conventional centers of yesteryear
that we've been talking about through the evolution of these
facilities and the Venn diagram of like, is it cloud
storage versus cloud computing or When we're talking about AI,
we're definitely, I think, talking about a combination of both,
because obviously all of that stuff that you enter into
(12:37):
AI and upload has to go somewhere. So it has
to live in a storage facility of some sort, but
then it has to be processed using software. I think
a lot of times they're NVIDIA or what you might
consider graphics card type processors stacked up in racks.
Speaker 3 (12:55):
Exactly. Thank you, Ben. They're the most powerful ones out there.
Speaker 5 (13:00):
And I think we're going to get to it, but
just the monopoly that companies like NVIDIA have over these
types of things and how beholden to companies like that
this whole– AI.
Speaker 3 (13:10):
Boom, uh, bust bubble, uh, is.
Speaker 5 (13:13):
Um, so we are talking about large language models that
learn by constantly analyzing petabytes of text and code and images. Again,
those require high capacity storage. Uh, this requires also extreme
computing power to sift through these massive, massive data sets
(13:33):
at scale, meaning mega, mega fast. Um, as well as,
like I was talking about, specialized hardware that you can't
just bolt onto your typical PC or laptop. So an
AI data center is very different from the typical data
centers we were talking about because the chips that they use.
Speaker 3 (13:53):
Are much, much, much more powerful.
Speaker 4 (13:56):
Yeah, yeah. Your CPU, that's going to have sequential tasks,
and it's going to have what we call cores, a
few powerful main character cores. But a GPU, a graphics
processing unit, has a bunch of complicated, repetitive math that
it does all the time. I keep doing, I guess
I'm doing Benedict Cumberbatch references because it's like a beautiful mind.
(14:20):
That's what your GPU is doing.
Speaker 5 (14:22):
No, an imitation game, right? Benedict Cumberbatch is an imitation game.
Russell Crowe is in a beautiful mind.
Speaker 3 (14:27):
Oh, yeah.
Speaker 4 (14:27):
Those are two different people. Just like a CPU and
a GPU. Yes, we got there. These repetitive complex calculations
are happening all at once, simultaneously, back and forth, distributed
across thousands of other smaller cores. And the one rule
(14:48):
is this stuff can never stop. So these, we call
them frontier AI models, right? They got thousands of processors.
They're running 24 hours a day, seven days a week,
months at a time. And it's so much energy, you guys.
It is an astounding and disturbing amount of energy required here.
(15:10):
A single, quote-unquote, AI server rack is using something like
60 kilowatts, just as some coffee table math. That is
about six times higher than the energy requirements for a
standard server rack. This means that if you put all
these AI servers in one place, depending on the size
(15:32):
of the compound, it can consume as much power as
hundreds of thousands of human homes.
Speaker 5 (15:39):
Well, and we did a story recently on Strange News
about a whole region or a municipality where I think
it was like the electric company was like saying, we're
no longer going to serve customers. We're only going to
serve AI. And that just is like evidence of how
insane the energy consumption is for these facilities. So it's
(15:59):
a twofold problem. We're mainly addressing I think water today,
but it is also a resource problem in terms of
power and like, you know, like maybe depriving regular folks
of energy. Yeah.
Speaker 4 (16:11):
But also, you know, if you're ever cold and, maybe
you could, in theory, run to a data center because...
Warm your hands.
Speaker 5 (16:19):
Eating a bean, a single bean, with your fingerless gloves
as you warm your hands over the hot, hot GPU.
Speaker 4 (16:26):
Tell stories of when water was free. It's like you
were saying there, Matt. It goes back to cooling. These
things generate so much heat because of all the work
that they're doing. Part of the heat problem is compounded because...
The way that these models work requires you to have
(16:47):
thousands of computers, not just operating simultaneously, but operating pretty
close together with a super fast network. So you don't
get bottled traffic jams. Now you get a bunch of
pretty hot things on their own that are hanging out
and being hot together. I'm sure there's a reality show
reference we could make there.
Speaker 5 (17:06):
That would be a great name for a reality show.
Just being hot together, you know, being hot together.
Speaker 3 (17:13):
It's about influencers, maybe. Rather than being hot apart.
Speaker 5 (17:18):
Yeah, I think that would work. That could be the
subtitle of Love Island.
Speaker 4 (17:21):
I like how we're going for the LCD, the lowest
common denominator there. Find us for more reality show pitches. So,
keep going back to this. Matt, as you were saying,
engineers are still trying to figure out the increasingly difficult
craft of keeping all this stuff cool and it's more
(17:43):
and more uh difficult every day.
Speaker 2 (17:46):
You should we go over the types of cooling that
have been used and are currently being used for cooling
like in the way water is used in.
Speaker 5 (17:53):
Those and i just remember when i was younger building
pcs i mean everything's fan cooled then.
Speaker 4 (18:00):
Right ginormous hvac setups or uh one of the the
big primary ones right now would be Evaporation cooling, uh,
which we've got to get into, and then what they
call closed loop cooling. Uh, we also see facilities that
have HVAC systems and cooling stuff at multiple points.
Speaker 5 (18:24):
And you can get a closed loop cooling system, like
for home, you know, I mean, people build these types
of GPU rack systems in their homes, uh, to mine cryptocurrency,
for example. That was the first time I ever heard
GPUs being used this way. And I guess, correct me
if I'm wrong, but isn't this sort of a repurposing?
Like they are graphic processors. So they were originally designed
(18:45):
just for like gaming and like really fast, you know,
responses for, you know, high quality monitors and stuff. Or
not monitors necessarily, but yeah, I guess it is monitors.
It's like, you know, you have to have a certain
speed of graphics card to run a certain type of
game on PC. And then they sort of repurposed these
or were they always kind of meant for.
Speaker 3 (19:05):
All of the things.
Speaker 4 (19:06):
I like repurpose. Let's go with evolved as well. What
we're saying here, folks, is these are thirsty, thirsty dudes
and critics are concerned. A growing number of critics is
a weirdly bipartisan issue. A growing number of critics are
concerned that the more water these things take, the less
(19:26):
water there will be for humans. We'll be right back.
Here's where it gets crazy. Let's break down some general concepts, guys.
We talked about hot hanging out together. How do we
get cool? Like the AI, the water, we all know
(19:49):
these stats. by this point. A hundred word email written
by AI uses the same amount of water as a
freshwater water bottle, right? They're not drinking water the same
way a living organism would. They're taking it for cooling.
And we were talking a little bit off air. We
got to get into the cooling.
Speaker 2 (20:10):
Yeah, and that's if they're using evaporative cooling, right? That's
when a 100-word email uses that amount of water. That's
evaporating water out of the usually municipal drinking water system
because most of these, or at least a major percentage
of them, are hooked up to local water supplies. And
it just evaporates it. Because it's literally evaporating the water
(20:32):
to cool down the equipment.
Speaker 4 (20:33):
Yeah, cooling towers, right?
Speaker 2 (20:35):
Yeah, that's one way to look at it. But it's
often written as evaporative cooling if you're seeing it in
one of these documents that's being brought up by, let's say,
a contractor that is constructing a new data center. You'll
see evaporative cooling in there to describe how they're going
to cool off their stuff. You might also see closed-loop
water cooling, which is the one that's described as the
(20:56):
coils that run through the whole system. with usually cold water, again,
from a municipal system in the same way. These are
good because they can reduce the amount of water you're evaporating. Well,
they circulate.
Speaker 5 (21:09):
So it's not like it goes away, right?
Speaker 2 (21:13):
Yes, you can cut consumption, as we're talking about there,
evaporation by up to 70% compared to evaporative cooling with
that system. However, you're introducing things into the water potentially
with that. And then the water does eventually at some
point get flushed, usually back into the system where it
would go to like a wastewater treatment plant, which has
(21:35):
ups and downs as we're going to discuss. You get
the water back, but now there's a lot of stuff
in it. There's also the air cooling.
Speaker 3 (21:43):
Yeah, exactly.
Speaker 2 (21:44):
There's also air cooling, free cooling. That's another term you'll see.
That's just bringing in cold air from an environment. Let's
say if you're in the north of the US or
in Europe somewhere, often you just pump in cold air.
And the last one, which is really interesting, that I
haven't seen much. I don't know if you guys have
seen much of this happening in new data center construction.
It's an emerging technology, but it's called direct liquid cooling
(22:08):
and immersion. You might see that written. That is where
the systems themselves are actually immersed in a liquid that
is non-conductive. It's like a coolant that the GPUs and
stuff live inside. very much like the matrix with humans
living inside those little things.
Speaker 4 (22:27):
Or like dunking, dunking someone in a bucket and not
letting them swim up to the surface.
Speaker 1 (22:34):
Yeah.
Speaker 4 (22:35):
I'd love for us to talk a little bit more
about that evaporating technique because the evaporating water can, can
in fact accelerate the cycle of thirst because it's, it
leaves behind, as we're saying, it becomes a water plus,
if you will, because it leaves behind in the evaporation
(22:56):
process bacteria, minerals, other chemical contaminants that could corrode or
clog equipment, which means that if you are a data center,
you have to constantly flush out that old water, put
it somewhere, and add in more fresh water to prevent
long-term damage over time. And like you were saying, Matt,
(23:17):
we know that Emergent technology aside, most data centers now
use water in some form. I found some stats that
said 75 to 90 percent of data centers, not just
in the U.S., but across the planet, use some kind
of water-based cooling as their primary method, which means that
(23:40):
right now, like, 10 to maybe 25% are completely water-free.
Speaker 2 (23:46):
And yeah, Ben, and I saw that too. And it was,
it was mostly based on location.
Speaker 4 (23:50):
Yes.
Speaker 2 (23:50):
Temperature of where you're at. I also saw Ben from
the university of Georgia here, the UGA, their college of
agricultural environmental sciences is they have a big part in
water usage with data centers, at least in our state.
And according to them, Around 57% of all data centers
(24:14):
use water from a potable water source. That means water
that you can drink.
Speaker 4 (24:21):
New York is drinking it. New World Water, California is
drinking it. I got to stop singing that song.
Speaker 3 (24:26):
But yes.
Speaker 2 (24:26):
57%?
Speaker 5 (24:26):
Mm-hmm.
Speaker 2 (24:30):
It seems like you could use water that is not
drinkable for this stuff.
Speaker 3 (24:34):
Gray water, right?
Speaker 2 (24:35):
Poop water. Yeah.
Speaker 3 (24:36):
Well, I mean, truly. No, but gray water. We always
talk about that.
Speaker 5 (24:38):
We used to talk about that.
Speaker 3 (24:40):
And I just remember there was.
Speaker 5 (24:42):
An eco push for a while with new construction where
you could get tax subsidies for recycling gray water and
using it for not drinkable water. Again, gray water, by
the way, is like water that comes from the shower,
not the toilet. Like it's like not the poop.
Speaker 3 (24:57):
Water and it's not pure clean water. It's somewhere in between.
Speaker 5 (25:01):
And you can use it for, I think, even like
watering plant crops.
Speaker 3 (25:06):
And things like that.
Speaker 5 (25:07):
But I do wonder if there's a world where that
could be incorporated into this system.
Speaker 1 (25:13):
Yes.
Speaker 3 (25:13):
No one talks about gray water anymore.
Speaker 4 (25:14):
Well, that's part of the issue. The reason it has
to be potable or freshwater is because you have to...
Part of the efficacy of the cooling comes from the
purity of the water. So if it has... If it
has the wrong kind of additives, we'll talk about the
chemical additives in a second. I thought water.
Speaker 5 (25:33):
Would be water when it comes to its ability to
be cool.
Speaker 3 (25:37):
You know, that's interesting.
Speaker 4 (25:38):
Yeah, if it has the wrong kind of additives, it
can spell disaster for the data center, even if the
people are going thirsty. Before we continue, though, guys, our pal, Tennessee.
That's very good. Dylan, you just laid this awesome bar out.
Could you say it on air?
Speaker 3 (25:54):
If you're making slop, you should use slop water. Exactly.
Mop water, slop water.
Speaker 4 (26:00):
Mop water, slop water, not your mom and pop water.
Speaker 3 (26:03):
Well, it is your mom and pop's water. Yeah.
Speaker 4 (26:08):
This is the issue, right? These conventional data centers are
already taking a lot of work to be pooled, to
be powered. They're drawing as much electricity as 10,000 to
25,000 households, according to often it's like the International Energy Agency,
which kind of knows what they're talking about. The issue
(26:31):
is that the newer guys, the data centers you're reading
about in your Washington Post or your Futurology subreddit or
what have you, those hyperscale data centers that are LLM-focused,
they can use as much power as 100,000 homes. Shout
out to Meta.
Speaker 2 (26:51):
I'm going to give that 10 f * * *
that's on the f * * * that scale.
Speaker 4 (26:57):
Yeah, let's give it two Louisianas, or two New Orleans
as well.
Speaker 2 (27:02):
Oh, yeah, two full New Orleans.
Speaker 3 (27:04):
You got a lot of gray water in New Orleans.
Speaker 2 (27:07):
This is super important. There is a sliding scale that
you can look at when it comes to a data
center's water usage and electricity usage, and you can see
that often... As electricity increases, water usage decreases a little
bit or vice versa. And you can really see the
(27:29):
relationship between those two things. And either way, a massive
data center that is going to use the equivalent of
100,000 homes. is going to also be using quite a
bit of water, but maybe not as much. It just
depends on which system they go with. If they go
with some of the closed looped things and the evaporative systems,
(27:52):
which is, again, the evaporative stuff is a little older,
but it is still a major player right now. It
is just the thirstiest, most power hungry thing that currently
exists or is even imagined on this planet.
Speaker 4 (28:07):
Yeah, it's the sin of gluttony. writ large, right? There
is no satiation for the data center. A data center
is not going to say, I've had enough to drink.
Or it's not going to say, I'm good. You can
turn the lights off. It's more and more.
Speaker 2 (28:22):
I'm thinking of characters I've seen in sci-fi movies that
are like the huge, just bloated creatures that are.
Speaker 5 (28:31):
Hexus from uh fern gully the last rainforest or the
kind of goo monsters in princess mononoke uh yeah and
and like you know in that i think it was
spirited away spirited away as well but princess mononoke had
a bit of a eco theme to it you know
about conservation i guess and there were those things that
were just kind of eating everything like eating the animals.
Speaker 3 (28:54):
In the woods and All of that stuff. And as
you feed this thing, and it's just, give me more.
It goes, oh, here's a picture of a cat. Oh,
here's a picture of a cat. He's a hungry boy.
Speaker 4 (29:09):
It reminds us also of that line from Apocalypto, where
people are talking about how humans have a hole inside them.
Speaker 3 (29:20):
The Matt Gibson movie?
Speaker 2 (29:22):
The Bible talks about that.
Speaker 4 (29:24):
But despite that, it's a good film.
Speaker 3 (29:26):
I never saw it.
Speaker 5 (29:27):
Yeah, that's the one where they don't speak. They speak
an indigenous language.
Speaker 4 (29:32):
It's a banger. Is it good? It's worth a watch.
I enjoyed it thoroughly. So we know these comparisons are.
by the way, folks, all on point, and we didn't
spoil anything. Check out those references. The reason we're saying
we're rating it two New Orleans in addition to 10
f * * * that's on the f * * *
(29:53):
that scale, which I love, Matt, we're referencing Meta's Hyperion.
There's another meta data center in Wyoming that at the
time of its construction was projected to use more electricity
than every home in the state of Wyoming put together.
Speaker 2 (30:15):
Yeah, Project Cosmo is, is, was, it still is messed up.
You can read all about a crazy thing they did
where they allegedly, hey, we have to say this, they
allegedly introduced some pretty gnarly bacteria into the municipal water system.
You can read all about it in The Guardian from
(30:36):
this past July.
Speaker 5 (30:38):
Where do these additives get introduced?
Speaker 3 (30:42):
I'm confused.
Speaker 5 (30:43):
Are they pumping stuff into the water to make it
more cool? Is it just stuff that's picking up along
the way because of the pipes or whatever? I'm very
confused as to why is it coming back with so
much gook in it?
Speaker 1 (30:56):
Oh, yes.
Speaker 4 (30:57):
Yeah. That goes back to the idea of the water footprint.
So we set that up real quick. We know that, okay,
so we know that the centers themselves, like we're saying,
are thirsty. They need a lot of water. They also
need a lot of electricity. And what doesn't often get
reported as much as it should is that electricity is
(31:17):
coming from fossil fuel plants or nuclear power plants. And
those folks are also infamous for needing tons of water, right,
for their own cooling systems or to generate steam. So
the water footprint changes. gets bigger and bigger and bigger
uh and the water for the living things uh gets
(31:39):
tougher and tougher to find well i.
Speaker 2 (31:41):
Mean nestle's gotta have its water so you know look
man if the data centers don't suck it up then
then nestle gets it that's how the contract that's what
it states i think it's like section 10 part a i.
Speaker 4 (31:56):
I can't i can't believe we got this far without
referencing without referencing, there will be blood. Obviously, they're drinking milkshakes, right?
This depletes the local aquifers. It could dry out wells.
And Noel, you raise such an excellent point. It is
not just pure water coming in and then pure water
(32:17):
rolling out. The stuff going in, depending on the cooling system,
is going to be mixed with chemicals like chlorine dioxide, bromine,
sodium hypochlorite, which I was unfamiliar with. And that's good
for the center. Because it mitigates or stymies the growth
of algae and mold and bacteria. But what's good for
(32:39):
the data center is not always what's good for the
community in which that center is based.
Speaker 2 (32:44):
The way I think about it, Noel, is if you've
got a river, you've got flowing water that is just
continually moving, right? And because it's continually moving, you might
have buildup of things like bacteria in certain areas of
that moving water or that lake.
Speaker 5 (32:58):
I mean, yeah, it's carrying materials. And, you know, if
you don't filter that out and or treat it, it
can start gunking up your own works. And that is
not efficient for these machines that need to be cool
24-7 forever and all time.
Speaker 2 (33:12):
But in the case of Project Cosmo, you introduce some
of that water into a closed system, even though you're
introducing a whole bunch of things into it. You're still
working with pipes and systems and pumps that are, you know,
there are areas that will be exposed. differently, which may
build up, let's say bacteria. And now that bacteria doesn't
(33:34):
matter how much bromine and crap you're pumping into it.
You've just got quite a bit of bacteria in your
wastewater system, or maybe even your wastewater container that you're
holding that stuff on site. Um, and then you flush,
if you just hit the button and flush it back
into the system, into a river, into a municipal system,
guess what? You've got bacteria now.
Speaker 4 (33:55):
Yeah, check out our earlier episode on the Turkey Point
salt plume. It's for an example of how that water
can get out of control. And don't think we're off
the hook closed-loop cooling systems. You guys use propylene glycol,
adethylene glycol, which, surprise, surprise, also can end up in
(34:17):
a community's water. Big shout-out to PFA's. fluorinated F gases,
whatever those are without being chemist, we can assure you
they're not great for people. That's why they're not, you know,
in the, uh, in the food pyramid, which is also.
Speaker 2 (34:33):
Well, and as you're saying, build up in the systems,
like heavy metals, right? The stuff that we looked at
in Flint, Michigan and small towns across the world with
lead pipes and other issues, you know, with the potable
drinking water for the humans.
Speaker 4 (34:48):
Copper zinc as well yeah because it's uh it's it's
just oh it's bad guys.
Speaker 5 (34:57):
Remember remember that jeopardy category potent potables.
Speaker 4 (35:00):
I do i do man that's all i got sorry
no that's great i always wanted to go on my first.
Speaker 5 (35:06):
Exposure to that word potable and in this case it's
just talking about a thing you drink yeah ones that
are strong yeah.
Speaker 2 (35:12):
Remember that other category Existential threats to all of humanity.
Speaker 3 (35:17):
Yeah, that was a banger.
Speaker 5 (35:19):
Always a real good time on Jeopardy. Shout out to
Ken Jennings, by the way, who we used to work with.
It's so crazy seeing that guy. He's the new Alex Trebek.
Speaker 3 (35:28):
Good job, man.
Speaker 4 (35:29):
We can assure you Ken's a super nice dude as well.
Speaker 5 (35:32):
I think his podcast is still out there that we did.
I made the theme for it, ripping off Doctor Who
pretty aggressively.
Speaker 3 (35:40):
Doctor Who?
Speaker 4 (35:42):
Yeah, exactly.
Speaker 3 (35:43):
Very good.
Speaker 4 (35:44):
Very good.
Speaker 5 (35:45):
But it's out there and it's very good. It's him
and this guy, John Roderick, talking about weird stuff. And
they did stuff like about like how Starlings made it
here because of Shakespeare and stuff like that.
Speaker 4 (35:55):
Yeah, that's a wild story. We also have to tell
you that the problems go past water with these with
these construction projects. And we talked about this a little
bit on Strange News or Listener Mail recently, which you
can find in our audio feed, folks. Similar to how
the prison industry and now the concentration camp industry targeting immigrants,
(36:20):
similar to how those industries will find small, economically struggling
towns and then come to them the same way that
guy sold monorails in that episode of The Simpsons. They'll say, hey,
let us build this thing.
Speaker 3 (36:36):
Right.
Speaker 4 (36:36):
Cast aside your NIMBY. Let us build this thing and
we will give you so many jobs. This will be
such an economic boom.
Speaker 5 (36:46):
I know I've brought this up in the past, so
please forgive me for repeating myself, but it is kind
of like a firsthand observation of the way this works.
Speaker 3 (36:51):
When I was.
Speaker 5 (36:52):
Younger and I worked for public radio after the housing collapse,
I covered this small town near where I worked in Augusta, Georgia.
I'm totally forgetting the name of the small town, but
their almost entirety of their local economy was tied to
manufacturing carpet and housing kind of stuff.
Speaker 3 (37:11):
Whatever it was.
Speaker 5 (37:12):
And so basically they had something in the neighborhood of
a 100% unemployment rate. Like it was really, really, really unusual.
And they had this private prison company that had them
on the list of sites that they would maybe build
this prison. And I went and covered this town hall,
like the monorail guy, where the folks from Corrections Corporation
(37:34):
of America, that I think has rebranded at this point,
basically pitched this thing to the townsfolk and talked about
all the different types of jobs it would generate, janitorial,
corrections officers, all this stuff. And everyone was just so like,
it was like this lifeline of, And then they built
it somewhere else.
Speaker 4 (37:51):
Right.
Speaker 5 (37:52):
I mean, but this is the same kind of thing.
Only I think in this case of the prisons, it
would have made some jobs. This one, I'm not so sure.
What are the jobs outside of the actual construction that
I think it's a bit of a bill of goods.
Speaker 4 (38:06):
Yes.
Speaker 2 (38:07):
Just on a wider scale there. Think about the complexity
of that problem. You just described Noel. We, these humans are,
living in a rather remote area, need jobs so badly
that they are willing to imprison human beings.
Speaker 3 (38:23):
They were stoked about it. They were stoked about it.
I mean, I'm sorry, you know, yeah.
Speaker 2 (38:28):
But that's what all of us, especially here in the
United States, just kind of go along with every day.
And there's not much we can do about it. The
prison industrial complex ends up being a massive part of
how the USA can make its GDP.
Speaker 3 (38:45):
Well, and let me just rephrase.
Speaker 5 (38:46):
They weren't stoked about participating in locking up humans.
Speaker 4 (38:50):
Of course.
Speaker 5 (38:50):
But it was just so removed from that, what that
is on a like large scale that it was just
like job.
Speaker 3 (38:58):
This equals job. We are stoked about job.
Speaker 5 (39:01):
The consequences or whatever the thing is, be damned.
Speaker 2 (39:05):
Because human must have job, right? That is, that's the thing.
Human must have money. Eat for eat, for food, for water,
for everything.
Speaker 5 (39:12):
To ignore the gnarly knock-on consequences because you're so desperate
that you're like, please, yes, in my backyard. And let me,
you know, be the guy that runs the freaking whatever
it is, pulls the switch.
Speaker 2 (39:25):
It's so similar to the existential threat that... the United
States in some way believes it faces with the need
for these AI data centers. If not AI data center,
then we will be destroyed by others with AI data center.
So must build AI data center. more AI data center,
most powerful AI data center, or death.
Speaker 5 (39:47):
Well, there's a space race element to it. And it almost,
you know, it really truly is an arms race as well.
It's not only about progress and about like societal, you know,
whatever superiority.
Speaker 3 (39:58):
Ben, I'm really curious your thoughts on this.
Speaker 5 (40:00):
It is a kind of arms race because of what
these things power and what they mean in terms of
the evolution, what they will do down the line.
Speaker 1 (40:08):
Yeah, yeah.
Speaker 4 (40:08):
Let's The arms race aspect is is a wonderful thing
for us to look into further. We also have got
to tell you, Bill of Goods is the perfect way
to describe the disparity between the jobs that are promised
and the jobs that are delivered. And no, you are
absolutely right that there is a lot of front end
(40:31):
employment in the construction industry. And a lot of politicians,
a lot of real estate developers make their vigorous. So
much opportunity to grip.
Speaker 3 (40:40):
So much opportunity to grip. So much.
Speaker 4 (40:42):
And often, as we'll see in specific cases, the public
of a community might not know the full extent of
what's going on. But we found this fascinating research that
dates back to 2024, where the investigators found that throughout
the United States, as few as 23,000 people held permanent
(41:08):
jobs at data centers. As few as 23,000, that is
about 0.01% of all employment across the country. And at
the same time, this industry consumed more than 4% of
all the electricity used in the United States. And those
numbers are becoming more increasingly distressing. The electricity usage is rising.
(41:34):
Automation means that you have fewer and fewer people, actual people,
who have to work at these places.
Speaker 2 (41:42):
Well, and the tech billionaire's dream is to only increase
that 23,000 people working in data centers by, you know,
just a teeny bit, just as much as they need, right?
Speaker 3 (41:51):
Yeah, their goal is to decrease it.
Speaker 2 (41:54):
Well, no, it's to increase it with the number of
new data centers, just a tiny bit, and then Get
rid of all the other jobs.
Speaker 3 (42:01):
That's what I'm saying.
Speaker 5 (42:02):
Guys, I mean, really quickly, if no one wants to
hear spoilers for the recent Wuk Chan Park film, No
Other Choice, just fast forward like 30 seconds. But it
does such an incredible job of spelling out this conundrum
where it's about this legacy industry, paper manufacturing that in
Japan is like this art. You know, there's like the
old school guys like tap on the rolls of paper
with a stick. And it's this whole almost like built
(42:22):
into the tradition of fine Japanese rice paper making by
hand and all this stuff.
Speaker 3 (42:27):
And then automation comes in. And then everyone gets laid off.
And then the.
Speaker 5 (42:31):
Whole story is about all of these qualified people fighting
for the one job at the automation center. That is
the guy that oversees what once would have been done
by 100, 150, 200 employees on a line. It's just
now dudes pacing through the facility in the dark because
they're saving money on the lights and just checking boxes
(42:52):
on an iPad. And I'm not even going to give
you the plot because in order to even get to
that place, some real gnarly shenanigans take place.
Speaker 2 (43:00):
Yes.
Speaker 5 (43:01):
Buying for those jobs, pitting, you know, colleague against colleague
to just scrap and, you know, ultimately, you know, do
some horrific things in order to be that guy.
Speaker 4 (43:11):
And speaking of horrific things, let's pause for a word
from our sponsors and then get into some specific cases. Folks,
we have returned and not to be too, you know,
self-centric here. but we are referencing our home state of
(43:33):
Georgia for a reason. There's an excellent and profoundly disturbing
example of how corruption can come into play over in Fayetteville, Georgia.
There's a great article you've got to read from Politico
about this. It's a fantastic summation, and it's the beginning
(43:54):
of dystopian sci-fi cinema. There are these residents in an
upscale subdivision named Annalise Park. And one day, these people,
who are pretty well-to-do, they notice that their water pressure
is not up to par. It's running incredibly low. And
these are people who are accustomed to things working out
(44:15):
for them because they are well-to-do. So they go to
the local authorities. Some of them are the local authorities.
And they're going like, hey, this is in Flint, Michigan.
What's up with our water? The county utility folks go
to investigate what's up with the water. And then they
happen upon the local data center, which on the books
(44:37):
has one industrial scale water hookup. But in practice, it
has two hookups of the same size.
Speaker 3 (44:45):
They sneaked in the second one?
Speaker 4 (44:47):
They just did a double. They did a twofer, right?
Oh my gosh.
Speaker 2 (44:51):
I remember we talked about this in May this year
on Strange News. The big deal is that nobody, Nobody
in the county knew, right? That was the big deal.
Speaker 3 (45:00):
The utility company didn't know.
Speaker 2 (45:02):
Nobody knew that they just snuck in a second thing.
Speaker 4 (45:05):
Yeah, yeah. Shout out to Quality Technology Services. What do
we say about innocuous names, folks? Quality Technology Services. There
are two-for-oners here. And they were able to get away
with this without being billed, without it being on record,
just full-on free water for quite some time. A better
(45:26):
way to say it, I would argue, is water theft.
They eventually owed, when they got caught in their grift
because they got too thirsty, they owed something like $ 150, 000
for using more than 29 million gallons of water. And
think about this, folks. Think about the money they were
(45:48):
making while they're stealing this water. Paying $ 150, 000. What a bargain. Yeah,
that's peanuts. That's a puddle.
Speaker 5 (45:56):
Well, So that's the question I have. I don't remember
the way this panned out. And I'm excited to read
the Politico piece, but guys, I can only imagine this
did not lead to any kind of genuine consequences for
this company.
Speaker 2 (46:07):
No. No consequences. I remember this. I remember writing all
this down, guys. There was a quote from somebody from
the city side who said, quote, they're our largest customer
and we have to be partners. It's called customer service. Yeah.
Speaker 3 (46:26):
That's one way of putting it.
Speaker 2 (46:27):
Just talking about how they were the most important thing,
but you know, we, and we joked on that episode
about the, the actual customers, you know, of the city
of the town of Fayetteville.
Speaker 3 (46:40):
Yeah.
Speaker 4 (46:40):
That's insane. Yeah. This Fayetteville facility is one of the
largest developments in the country. It's like 615 acres. Georgia
is home to more than 200 similar data centers.
Speaker 2 (46:55):
Oh, Ben, this is why the QTS campus was projected
to generate $ 150 to $ 200 million annually in property tax revenue.
Speaker 3 (47:06):
That's why.
Speaker 4 (47:06):
Yeah, that's why. That's the customer service. So think about it, folks.
If you want to go in with us on a
$ 150, 000 investment to make millions, how would you say no
to that? I also, we should mention that Meta Center
in Wyoming is in addition to the concerns about wastewater,
(47:27):
just to put a, put a bow on this one,
a bacterial bow, uh, we're talking about something called Cupria
Vetus Gilardi. Cupria Vetus Gilardi. Excuse my trivia voice there,
but it was, uh, it came into the, uh, the
Cheyenne water system because entirely because of allegedly, uh, meta's
(47:51):
activities at that data center. And they had to, the,
The thing that panicked the leadership of Cheyenne was that
they had to shut down parts of the water system
for the people, but not really shut down the data center.
Speaker 3 (48:10):
Customer service, baby.
Speaker 2 (48:11):
There's some messed up stuff in there. And I would
just point out, if you read anything, I've been, guys,
I've been seeing the best stuff in the Guardian just
because it is still independent. I can still access it,
but you can see the names of contractors that the
companies that are going in to build the data centers,
like Goat Systems LLC, that was a part of Project
(48:32):
Cosmo there in Cheyenne, or Fortis, a general contractor there,
and just how these companies are making so much money
from the construction part, and then also seeing in Congress
how the construction part ends up becoming one of the
major water pollution flashpoints in for local residents where these
(48:56):
data centers are going in, because you're talking about all
kinds of crazy manipulation of the land you have to
do to get those flat services to make an 800,000
square foot facility, 1.2 million square foot facility, and how
some of those explosions, some of that stuff ends up
getting into the water system where private wells are, especially
(49:19):
in rural rural areas.
Speaker 3 (49:22):
People have wells.
Speaker 5 (49:23):
It goes down to like, how do you own the water?
And they don't own the land. It's that milkshake argument, right?
They're like drinking it from places that they don't own.
It doesn't belong to them.
Speaker 4 (49:33):
Yeah, like we say, it's draining the local aquifers. The
wells are going dry. There was a fantastic BBC piece
about that, which, again, not to be too Georgia-centric, is
a story from a person based in Georgia.
Speaker 2 (49:47):
Well, to keep it Georgia-centric, AOC back in May held
up two jars of water. We talked about it on
Strange News. This nasty-looking, had all kinds of black sediment
and just brown poop-looking water. And it came from two
private wells in Morgan County here in Georgia. And it
was because of the construction of a data center, not
(50:08):
the use of an operational data center. And, you know,
there's a lot that has been made about that. Ben,
that's where the UGA thing came in. We find out
that the University of Georgia Cooperative Extension Office handles broader
private well testing throughout Georgia. But the EPA said that
private wells are not regulated under the Federal Safety drinking
(50:32):
water act and Georgia does not regulate private well water
at the state level. So if you do live in
a rural area where one of these things is coming
and you have a private well, you're out of luck.
You just got to hope, cross your fingers and hope
that the water you drink and bathe yourself with and
feed to your children and your pets is going to
(50:54):
be okay.
Speaker 4 (50:55):
Oh, and by the way, on that note, Folks, we
know these are incredibly divisive political times here in the
United States. There is at least one issue with broad
bipartisan support of the people in power, regardless of where
they are on the political spectrum. Broadly, the establishment political
(51:19):
parties love data centers. Did you know, folks, fun fact,
both the Biden and the Trump administrations went 10 toes
down on data centers.
Speaker 2 (51:31):
Gee, I wonder why, Ben. Why would so many people
be so into data centers, especially those who, I don't know,
get donations for their campaign or exist on ad dollars?
Why would everybody be so into these AI data centers?
Speaker 4 (51:51):
So again, that's what we have to acknowledge. Ocasio-Cortez, Bernie Sanders,
for instance, they're considered, in the Overton window of discourse,
they're considered very far left and anomalous for not supporting
these data centers. The money is rolling in, right? And
campaign donations, as you said, and pork barrel projects, as
(52:13):
we used to say. Also, it helps with the creation
of a surveillance state. It's true. That's a big part
of it is we're building the eye of Sal Rotten again.
Speaker 2 (52:26):
Let's be real about the advertising industry or the industries
that are supported by advertising. Just, guys, how many AI-related
ads have come our way that we've had to vet?
Speaker 5 (52:43):
Without being too inside baseball, it is an initiative. It
is a thing that is almost like, you know, with
companies like ours or folks that are subsidized by advertising,
it's going to be a conversation internally about guys are
going to be seeing more of this kind of stuff
and you got to have a position on it. Don't
want to get left behind, you know?
Speaker 2 (53:04):
It has to be because we've mentioned before that circular
investment that's happening, the tune of hundreds of billions of dollars,
that stuff gets cycled through marketing too. And it just,
it becomes a lifeline. Like a vital part of a
company that's based on advertising. Those companies now, right now,
all of them need it.
Speaker 5 (53:26):
As traditional advertising maybe wanes a little bit and is
at dollars or maybe not what they used to be
and it's a little tougher because the pie keeps getting
divided and divided and subdivided. Yes, of course you're going
to grab onto that with both hands if you're a
CEO or the head of a marketing agency or whatever
it might be.
Speaker 4 (53:44):
Or beholden to ill-informed shareholders, perhaps. I would also point
out that a similar thing applies to some of these
construction companies. Some of these contractors are entirely dependent upon
projects like this for their continued survival. So a lot
of the people involved, or a lot of the entities
(54:07):
involved here, do feel this is a matter of survival.
And so does the public, by the way. This is
a bipartisan issue for us, the non-powerful.
Speaker 3 (54:18):
It is, but can't you guys see it?
Speaker 2 (54:19):
We are all residents of that small town you described, Noel.
Speaker 3 (54:24):
All of us.
Speaker 2 (54:25):
We are all residents of that small town going.
Speaker 3 (54:27):
Please, please, bring it here. Protect our jobs. Yes.
Speaker 4 (54:32):
All of us.
Speaker 5 (54:33):
Even in an industry like we're in, we're not immune
from concerns around AI replacing us. Or aspects of the
production of our show. And you know what I mean?
And it's like you're almost like you're having to give
in to the AI to protect yourself from the AI.
Speaker 4 (54:52):
Right?
Speaker 3 (54:53):
It's weird.
Speaker 4 (54:53):
Burn the village to save it.
Speaker 3 (54:55):
Kind of. Yeah.
Speaker 5 (54:56):
You got to play ball with them to get their
dollars because you might not be making your dollars because
they've taken your job.
Speaker 3 (55:04):
I mean, it's a weird or a burrus, right?
Speaker 4 (55:06):
Yeah, it's a real burristic for sure. I do want
to give us a bit of good news here, folks.
We said it's weird that this is a bipartisan supported
issue by the establishment, by the people in power or
the powers that be. The flip side of this, which
I find quite inspiring, is that we're also seeing a bipartisan,
(55:30):
not unanimous, push against it from the people who are
not in power. The public is fighting back. Some politicians
are listening. Please check out an excellent article by a
guy named Andrew Howard, a journalist named Andrew Howard for Politico,
talking about how this just recently was largely a local issue, right?
(55:53):
And if you're running for a political office, you probably
wouldn't mention it a bunch, but now it is a
top-button debate in some of the most sensitive races in
American politics, Ohio, Michigan, and so on.
Speaker 5 (56:09):
Politico is writing about Fayetteville. I mean, it's definitely on
the radar in a big way.
Speaker 4 (56:16):
And isn't it inspiring to realize that there are people
in this nation who will never agree on so many things,
but here people are shaking hands across the aisle and
The disagreement, the opposition to data centers and all their
consequences is one of the biggest cross-partisan issues in modern history.
(56:42):
And then, you know, the Epstein files. So again, I'm
saying the public versus the powerful right now.
Speaker 5 (56:49):
It also goes to show that the powerful are represented
on both sides of the aisle. That's why it's so
bipartisan because this is an unimpeachably beneficial thing to, you know, folks.
Speaker 2 (57:00):
Ben, I have to just say you are absolutely right.
And it is crazy to see that we do have
so much power. Like in Hall County in Gainesville, there
was this thing called Project Turbo. Not kidding, guys. That's
not a joke.
Speaker 3 (57:12):
A $ 1.
Speaker 2 (57:12):
2 billion data center that was going to consume 225,000
gallons of water per day to run its data center.
It was going to be huge. It was going to
be a massive thing for Gainesville. The reason why the
folks there in Hall County stopped it is because human beings,
people who lived in Hall County, went to the meetings
(57:34):
and shouted enough, said no enough, and they stopped it.
They stopped it from happening. That was a year ago.
A year ago in August, it was proposed, and then
by December, the human beings spoke up enough that it
killed it.
Speaker 4 (57:50):
Yes, yeah, there are more and more inspiring stories. We're
even seeing journalism describe a paradigm shift, a sea change.
As more people learn about this, you'll see articles talking
about organizations making a U-turn on the construction of data centers.
And we want to give you a tip. I'd love
(58:10):
to give you a tip. Folks, write to us if
you play along at home and tell us how this
works out. I imagine that if you can think of
someone who, disagrees with you politically on every other issue,
go to them and ask them what they think about
data centers. Tell them you think it's * * * *
(58:30):
and then tell us what they said because it might
be common ground for you to save your friendship or
make a new rapport because fighting against the common cause forever,
it has united societies and this might be the battle
of the eight.
Speaker 2 (58:49):
Dude, we could see humanity just investing billions and billions
of dollars into this thing. And what does it really
do for us? There are a couple of cool scientific
advancements that are possible through this stuff. But most of
it is crap and a money-making opportunity. And that's all
it is. It's a scheme. It's a popular scheme. to
(59:10):
make money which is being turned into the threat to
scare you of oh AI it could bring down the
whole world because a bad actor might be able to
crack all of our systems and all this stuff it's
fear it's fear for the money makers the controllers the
owners that's all it is.
Speaker 5 (59:30):
I mean, yeah, and devil's advocate, I mean, it's with
the degree to which we've experienced it just exponentially improving
in what it's capable of. Like I do see there
being something of an investment in this idea of, well,
if it's gotten here this fast, where is it going next?
But in order to get to the next phase of it,
it requires more and more investment, more and more data centers.
(59:54):
It's a weird one because it becomes this argument of
like we can't let– be left behind by history. Yeah.
Speaker 4 (01:00:01):
It also, it's flirting with the, we're right now as
a global civilization, we're on third base of sunk cost fallacy, right?
You've walked far into a dark room. You don't know
the full direction. There is this idea of keeping up
with the Joneses globally, a new kind of, as we
said earlier, an arms race, a bag of digital badgers
(01:00:24):
to pickle for the modern age. And really the, I
don't want to sound like a Luddite, but the smart
thing to happen would be a heavier regulation and maybe
even just turn them off for a second. Turn them off,
take a breath, touch some grass.
Speaker 3 (01:00:40):
Touch some grass.
Speaker 2 (01:00:42):
They're trying to scale so that everybody can use it
for everyday things. And that's the only way they can
make money is if everybody is using these things every
day for everyday purposes.
Speaker 3 (01:00:53):
Let's add it to Google Earth. That's what I'm saying.
That's the only way they can.
Speaker 2 (01:00:59):
The only way they can be profitable is to scale
in those ways. But the only way it's going to
be good for humanity is if it's scaled down and
used for very specific purposes.
Speaker 3 (01:01:09):
Yes, that's exactly right.
Speaker 4 (01:01:10):
Like you make new medicines, you predict.
Speaker 2 (01:01:13):
Yes, that's one thing. That's what you have to do.
That's what we have to do. That's what we have
to make them do. Because right now, they are going
to kill us all.
Speaker 3 (01:01:22):
Seriously.
Speaker 2 (01:01:23):
They're going to kill us all. Because they're greedy and
they want more money. And that's what's happening right now.
Speaker 5 (01:01:30):
Yeah.
Speaker 3 (01:01:30):
I'm with you.
Speaker 4 (01:01:31):
Some extreme critics love the idea we just proposed a
second ago of turning around the ship, turning stuff off
for a second, retreating from AI, finding more focus or
scoped in uses. But it's highly unlikely that that's going
to happen. The research, the construction is going to pause, right?
(01:01:52):
And the public is increasingly opposed as they learn more
about it. But Pausing doesn't mean fully stopping. And it's also,
it's troubling. It's frightening, folks, that we as a civilization
do not know the full consequences of this reality. We
have to ask ourselves, are we going to be able
(01:02:12):
in this dark room, right? We don't know the size
of the room. Are we going to be able to
understand what is in that darkness? Are we going to
be able to comprehend the problems on the horizon? Or
will we really only learn about them once it's too
late to fix?
Speaker 5 (01:02:26):
Well, this is the kind of thing that I think
when we first started hearing about it, it was only
trickling out into the public kind of consumer spaces after
being dicked around with in exactly the way that we're
probably proposing that it should go back to being dicked
around with, like, you know, in labs, you know, in
high level computing facilities that are devoted to things like SETI,
(01:02:48):
you know, or various forms of research and analytics, right?
And then, like anything, someone had the genius idea, but
what if we sold it to everybody?
Speaker 4 (01:02:58):
Let's unscrew the lid on Pandora's jar. We can't wait
for your thoughts, folks. We know we went a little
bit long on this one. Thanks for riding with us.
Cheers to everybody. Enjoy a nice glass of water while
you can. Reach out and tell us your thoughts. We
can't wait to hear from you.
Speaker 5 (01:03:17):
Yeah, you can sure do that on the internet by
reaching out to The Handel's Conspiracy Stuff or Conspiracy Stuff Show, depending.
Speaker 2 (01:03:23):
We have a phone number. It is 1-833-STD-WYTK. See, look,
it's written right here. Just call this number. That's just
for the video team to see if they'll do it.
If you want to send us an email, we are.
Speaker 4 (01:03:38):
Let's double down on Maybe you can see that on
the bottom third, too, other Chiron.
Speaker 3 (01:03:42):
Ugh, making more work for these poor guys.
Speaker 4 (01:03:44):
We are the entities who read each piece of correspondence
we receive with great interest and be well aware sometimes
the void rides back, back, back. We'll see you here
in the dark. Conspiracy at iHeartRadio.com.
Speaker 2 (01:04:16):
Stuff They Don't Want You To Know is a production
of iHeartRadio. For more podcasts from iHeartRadio, visit the iHeartRadio app,
Apple Podcasts, or wherever you listen to your favorite shows.
Stuff They Don't Want You To Know News
Hosts And Creators
Matt Frederick
Ben Bowlin
Noel Brown
Show Links
RSSStoreAboutLive ShowsPopular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
The Bobby Bones Show
Listen to 'The Bobby Bones Show' by downloading the daily full replay.