I'm William Brangham and this is "Horizons."

Artificial intelligence data centers are popping up across America.

They're spurring innovation, but also fierce backlash because of their need for huge amounts of water and power.

Is the AI revolution worth the price communities are paying?

Artificial intelligence data centers What we know about the impact of data centers, are popping up across America.

coming up next.

are popping up across America.

♪ worth the price communities are paying?

Narrator: Support for "Horizons" has been provided coming up next.

♪ by Steve and Marilyn Kerman ♪ and the Gordon and Betty Moore Foundation.

♪ Additional support is provided by Friends of the News Hour.

♪ ♪ Additional support is provided by Friends of the News Hour.

This program was made possible by contributions ♪ to your PBS station from viewers like you.

♪ Thank you.

♪ From the David M. Rubenstein Studio ♪ at WETA in Washington, ♪ here is William Brangham.

♪ Welcome to "Horizons."

This program was made possible by contributions We're going to do a deep dive today into the controversies to your PBS station from viewers like you.

around the explosion to your PBS station from viewers like you.

of artificial intelligence data centers.

Thank you.

If one is or has been built in your community, at WETA in Washington, you likely know what I'm already talking about.

here is William Brangham.

Hundreds of billions of dollars are being spent right now Welcome to "Horizons."

constructing massive facilities We're going to do a deep dive today into the controversies to house and run the computers around the explosion that are needed for the spiking demand of artificial intelligence data centers.

for artificial intelligence.

If one is or has been built in your community, These centers are being built across the country you likely know what I'm already talking about.

with a particular concentration in Virginia, Hundreds of billions of dollars are being spent right now Texas, California, and Illinois.

constructing massive facilities The biggest ones, so-called hyperscale plants to house and run the computers like Meta's Hyperion plant in Louisiana, for artificial intelligence.

could have a footprint nearly as big as Manhattan.

These centers are being built across the country The leading tech giants, Meta, Google, Amazon, and OpenAI, Texas, California, and Illinois.

among others, are in a breakneck race The biggest ones, so-called hyperscale plants to get their centers up and running, The biggest ones, so-called hyperscale plants each hoping to seize the advantage like Meta's Hyperion plant in Louisiana, and become a dominant supplier of AI.

could have a footprint nearly as big as Manhattan.

Hey, hey, ho, ho, data centers have got to go!

among others, are in a breakneck race Brangham: But as those centers are being built, to get their centers up and running, communities across the country are in open revolt, each hoping to seize the advantage concerned about the large amounts of water and energy and become a dominant supplier of AI.

that these centers require.

Hey, hey, ho, ho, data centers have got to go!

And it's a bipartisan backlash, with citizen activists Hey, hey, ho, ho, data centers have got to go!

pushing back at the speed of construction, Brangham: But as those centers are being built, the concern over pollution and rising utility bills, communities across the country are in open revolt, and the often opaque deals and tax breaks concerned about the large amounts of water and energy given to companies by their local officials.

that these centers require.

According to one poll, seven in 10 Americans And it's a bipartisan backlash, with citizen activists would oppose a data center being constructed in their community, the concern over pollution and rising utility bills, and the majority would support a national moratorium and the often opaque deals and tax breaks on building more.

and the often opaque deals and tax breaks Given all of this, we wanted to examine the claims given to companies by their local officials.

swirling around data centers According to one poll, seven in 10 Americans to try and get a little clarity about their real impact.

would oppose a data center being constructed in their community, And for that, we turn to Michael Webber.

and the majority would support a national moratorium He's a chair at both the LBJ School and the majority would support a national moratorium of Public Affairs on building more.

and the Department of Mechanical Engineering, Given all of this, we wanted to examine the claims both at the University of Texas at Austin.

swirling around data centers He served as an officer and consultant to try and get a little clarity about their real impact.

for several tech and energy companies, And for that, we turn to Michael Webber.

including some in the AI business.

He's a chair at both the LBJ School And he's the author of the book "Power Trip, of Public Affairs "The Story of Energy," which was also made and the Department of Mechanical Engineering, into a multi-part series on PBS.

both at the University of Texas at Austin.

Michael Webber, so good to have you on the program.

for several tech and energy companies, I want to get into all these pros and cons including some in the AI business.

and accusations and defenses about these data centers.

And he's the author of the book "Power Trip, But first off, just establish a baseline for us.

"The Story of Energy," which was also made What are these data centers?

"The Story of Energy," which was also made What are they for?

into a multi-part series on PBS.

Who's building them?

into a multi-part series on PBS.

That's a great question.

Michael Webber, so good to have you on the program.

So data centers, first of all, have been around for a while.

Michael Webber, so good to have you on the program.

They're not new, but the pace I want to get into all these pros and cons of their construction and their size and accusations and defenses about these data centers.

is really capturing the attention and accusations and defenses about these data centers.

because the last few years they've gotten a lot bigger But first off, just establish a baseline for us.

and we're starting to use the tools in a more obvious way.

What are these data centers?

By tools, I mean, like these AI tools.

Who's building them?

You might've used Claude or ChatGPT That's a great question.

or something like that.

So data centers, first of all, have been around for a while.

And these tools are calling these data centers, They're not new, but the pace which are these buildings that exist of their construction and their size oftentimes outside of cities.

of their construction and their size And they've got big warehouses, big tilt wall construction.

and we're starting to use the tools in a more obvious way.

And so they're just large facilities, and we're starting to use the tools in a more obvious way.

maybe it looks like a distribution center By tools, I mean, like these AI tools.

or something, but inside is filled with computer racks, You might've used Claude or ChatGPT just go on and on and on, these rows and rows And these tools are calling these data centers, of racks of computers that are stacked up and down.

which are these buildings that exist And those data centers fill these computers oftentimes outside of cities.

are answering our queries as we use these tools.

And they've got big warehouses, big tilt wall construction.

We ask it like, "Make me a logo" And they've got big warehouses, big tilt wall construction.

or "Look something up for me" or "Help me analyze this data."

maybe it looks like a distribution center That question we type into our computer or something, but inside is filled with computer racks, is going over fiber optics towards that data center, just go on and on and on, these rows and rows calling all those computers to do a calculation of racks of computers that are stacked up and down.

and then giving this back some sort of reply.

of racks of computers that are stacked up and down.

So those data centers are the place where those today, are answering our queries as we use these tools.

where those computations are made.

are answering our queries as we use these tools.

I see, okay.

We ask it like, "Make me a logo" I mentioned earlier that this 7 in 10 poll, That question we type into our computer people supporting a moratorium on building data centers.

calling all those computers to do a calculation I want to put up this recent Quinnipiac poll, and then giving this back some sort of reply.

which showed that I believe it was 65% of people where those computations are made.

would oppose a data center in their community.

I mentioned earlier that this 7 in 10 poll, And when asked why, 72% electricity costs, people supporting a moratorium on building data centers.

64% said water use, I want to put up this recent Quinnipiac poll, 41% said noise.

which showed that I believe it was 65% of people So I want to go through each of those individually.

which showed that I believe it was 65% of people Let's start up with electricity.

would oppose a data center in their community.

Data centers do gobble up a lot of energy.

And when asked why, 72% electricity costs, That is quite clear.

And when asked why, 72% electricity costs, A lot of people argue data centers mean 64% said water use, in my community that my utility rates are going to go up.

So I want to go through each of those individually.

What's the truth to that?

Let's start up with electricity.

I mean, that's the concern and that's a fair concern.

Let's start up with electricity.

People are already nervous about electricity rates.

That is quite clear.

Rates have been going up for years That is quite clear.

because of having to harden the system for hurricanes A lot of people argue data centers mean and wind storms or replace the system after wildfires in my community that my utility rates are going to go up.

in the Mountain West.

in my community that my utility rates are going to go up.

So costs have already been going up for years.

What's the truth to that?

And now we have data centers who want a lot of electricity.

I mean, that's the concern and that's a fair concern.

So people are rightly concerned that rates will go up.

People are already nervous about electricity rates.

And I think there's a lot of political Rates have been going up for years or policy concern that retirees and grandmothers because of having to harden the system for hurricanes will have to pay higher rates and wind storms or replace the system after wildfires to make really rich companies even richer.

and wind storms or replace the system after wildfires So there's a lot of concern and sensitivity in the Mountain West.

about electric rates.

So costs have already been going up for years.

There has always been concern about electricity rates.

And now we have data centers who want a lot of electricity.

So that's not a new thing, but all the recent inflation So people are rightly concerned that rates will go up.

concerns really show up in our electric bills.

And I think there's a lot of political So there's a lot of concern.

And I think there's a lot of political And the reasons why that's a valid concern or policy concern that retirees and grandmothers is because these data centers consume a lot of electricity.

to make really rich companies even richer.

And so if you have a big new consumer So there's a lot of concern and sensitivity come in who's going to take a lot of electricity about electric rates.

and maybe they can outbid you for it, There has always been concern about electricity rates.

they might bid the prices up.

So that's not a new thing, but all the recent inflation So that's a valid concern.

So that's not a new thing, but all the recent inflation Those same data centers are owned or operated concerns really show up in our electric bills.

or used by really rich companies.

So there's a lot of concern.

So maybe we can convince them to invest And the reasons why that's a valid concern in making the grid better, which would drive down is because these data centers consume a lot of electricity.

the rates for everybody.

And so if you have a big new consumer And the challenge is, we don't know which way And so if you have a big new consumer it's going to go right now.

come in who's going to take a lot of electricity Are the data centers and the hyperscalers and maybe they can outbid you for it, and these data companies, are they going to invest So that's a valid concern.

to make electricity more affordable for all of us?

Those same data centers are owned or operated Or are they going to compete with us for electricity or used by really rich companies.

and drive our rates up?

So maybe we can convince them to invest That's the concern I think that they're stepping into in making the grid better, which would drive down with a lot of these local communities.

the rates for everybody.

I know that some of these data centers, it's going to go right now.

when they're being built, are saying, "Look, and these data companies, are they going to invest "because of these concerns and these data companies, are they going to invest "and maybe because of local efficiency reasons, Or are they going to compete with us for electricity "we will build our own onsite and drive our rates up?

"power plants that we will put with a lot of these local communities.

"turbines and generators onsite to supply our own electricity."

I know that some of these data centers, Does that alleviate this problem?

"because of these concerns It solves one problem, but introduces other problems.

"and maybe because of local efficiency reasons, So if you're going to build what's called onsite power "we will build our own onsite or behind the meter power generation, "we will build our own onsite the meter being where the building occurs "power plants that we will put with the local utility, "turbines and generators onsite to supply our own electricity."

they might say, "Look, we'll build our own power plant.

"turbines and generators onsite to supply our own electricity."

"So we're not going to charge you "turbines and generators onsite to supply our own electricity."

"or cause your rates to go up or anything like that."

Does that alleviate this problem?

That might solve the problem It solves one problem, but introduces other problems.

of competing with all of us at peak times.

It solves one problem, but introduces other problems.

And maybe we should just take a moment to say So if you're going to build what's called onsite power that peak times tends to be afternoons in August the meter being where the building occurs in the summer when everyone has their air conditioners on.

We actually have an abundance of electricity "So we're not going to charge you almost the entire year, "or cause your rates to go up or anything like that."

but there are a few hours each year "or cause your rates to go up or anything like that."

where we don't have an abundance.

That might solve the problem And if they have onsite power generation, of competing with all of us at peak times.

maybe that data center can disconnect from the grid that peak times tends to be afternoons in August or partially disconnect and use their own power plant.

that peak times tends to be afternoons in August So they're not competing with us at peak times.

in the summer when everyone has their air conditioners on.

And that's desirable in a variety of ways.

We actually have an abundance of electricity The data center companies for the most part almost the entire year, are actually trying to solve a different problem.

where we don't have an abundance.

They're trying to solve a speed to power problem And if they have onsite power generation, where they want to get connected to electricity maybe that data center can disconnect from the grid as soon as possible because they feel like they're in a race.

or partially disconnect and use their own power plant.

And when they talk to the local utility, So they're not competing with us at peak times.

the utility might say, "Look, it's going to take us And that's desirable in a variety of ways.

"four or five years to connect you to the grid."

And that's desirable in a variety of ways.

They'll say, "You know what?

The data center companies for the most part "We'll build our own power plant cause that's faster."

are actually trying to solve a different problem.

So they're often trying to solve a speed to power issue.

They're trying to solve a speed to power problem They can use that same equipment to reduce competition as soon as possible because they feel like they're in a race.

with all of us at peak times as soon as possible because they feel like they're in a race.

in those afternoons in August.

And when they talk to the local utility, And that solves a couple of problems, the utility might say, "Look, it's going to take us which is kind of nice.

the utility might say, "Look, it's going to take us But now all of a sudden you got a power plant "four or five years to connect you to the grid."

in your backyard and you might not want the noise They'll say, "You know what?

or the fumes or the trucks delivering the fuel "We'll build our own power plant cause that's faster."

depending on what the fuel is.

So they're often trying to solve a speed to power issue.

So there are other problems that come So they're often trying to solve a speed to power issue.

with the onsite power generation, They can use that same equipment to reduce competition but it does perhaps alleviate some pressure on electric rates.

in those afternoons in August.

Right, that is another one of the issues And that solves a couple of problems, that you just touched on, this idea of local pollution.

But now all of a sudden you got a power plant Because those, if people haven't seen or the fumes or the trucks delivering the fuel what a big electrical turbine is, So there are other problems that come it's like a jet engine and that there can be but it does perhaps alleviate some pressure on electric rates.

dozens of them, if not more on some of these big plants.

Right, that is another one of the issues That is something that maybe when you bought that you just touched on, this idea of local pollution.

your property out in a rural stretch of the country, that you just touched on, this idea of local pollution.

you weren't anticipating having a couple dozen jet engines what a big electrical turbine is, firing up all the time in your backyard.

it's like a jet engine and that there can be Yeah, a couple of dozen jet engines it's like a jet engine and that there can be or hundreds of these diesel generators.

dozens of them, if not more on some of these big plants.

And diesel generators are especially noisy and fumy That is something that maybe when you bought and all that kind of stuff.

your property out in a rural stretch of the country, And you can imagine hundreds of them lined up you weren't anticipating having at your fence line, do you want that?

you weren't anticipating having No one wants that.

a couple dozen jet engines In fact, diesel generators have been outlawed in California firing up all the time in your backyard.

and other places, because they're so dirty.

firing up all the time in your backyard.

And this is the challenge, like, "Hey, we'll build Yeah, a couple of dozen jet engines "behind the meter or onsite generation, or hundreds of these diesel generators.

"but is it going to be an onsite generation And diesel generators are especially noisy and fumy "at your fence line that you're comfortable with?"

and all that kind of stuff.

That's part of the conversation.

And you can imagine hundreds of them lined up Right.

Okay.

at your fence line, do you want that?

The next item on that Quinnipiac poll No one wants that.

was people's concerns about water.

and other places, because they're so dirty.

Data centers, the NOCs use a ton of water, why?

"but is it going to be an onsite generation What are they using the water for?

"at your fence line that you're comfortable with?"

If you picture a data center as like a big white box, Right.

Okay.

maybe it's got, power comes in, electricity comes in, was people's concerns about water.

and then two things go out from the chips.

Data centers, the NOCs use a ton of water, why?

Photons go out, so you're taking electrons to photons.

Data centers, the NOCs use a ton of water, why?

The chip is converting the electricity into photons, What are they using the water for?

consider photons data or information packets, If you picture a data center as like a big white box, and generating heat, so the chip gets really hot in there.

maybe it's got, power comes in, electricity comes in, A lot of these chips and these computer servers maybe it's got, power comes in, electricity comes in, interacts in these data centers.

maybe it's got, power comes in, electricity comes in, So electricity comes in, photons and heat goes out.

Photons go out, so you're taking electrons to photons.

And the heat's a real problem.

Photons go out, so you're taking electrons to photons.

You have to manage the heat The chip is converting the electricity into photons, so you don't destroy the equipment The chip is converting the electricity into photons, or damage the facility or anything like that.

consider photons data or information packets, And so water is used for water cooling, and generating heat, so the chip gets really hot in there.

for the most part, to cool these systems.

A lot of these chips and these computer servers If you're inside a data center, interacts in these data centers.

there's actually a lot of air cooling.

So electricity comes in, photons and heat goes out.

They're blowing air over the computers So electricity comes in, photons and heat goes out.

and the racks and the chips, And the heat's a real problem.

but more and more as the chips become more powerful, you need water cooling.

or damage the facility or anything like that.

You can't cool them fast enough with just air, or damage the facility or anything like that.

so you have water cooling inside the facility.

And so water is used for water cooling, You also have to cool that water cooling.

for the most part, to cool these systems.

And that water cooling loop, If you're inside a data center, imagine like a radiator in your car, there's actually a lot of air cooling.

which has a closed loop of just radiator fluid They're blowing air over the computers going around to cool your engine.

but more and more as the chips become more powerful, It's that radiator's air cooled as you drive on a highway.

you need water cooling.

If you take that water cooling loop inside your radiator, so you have water cooling inside the facility.

you have the same kind of thing in your data center.

You also have to cool that water cooling.

Sometimes you use a lake or a river or water You also have to cool that water cooling.

to cool that water loop.

imagine like a radiator in your car, And that's part of the tension.

which has a closed loop of just radiator fluid Like, "Wait, how much water are you going to use which has a closed loop of just radiator fluid "to cool your facility?

going around to cool your engine.

"Is it a one-time charge where you're loading up the water It's that radiator's air cooled as you drive on a highway.

"in your closed loop cooling and then you're done?

It's that radiator's air cooled as you drive on a highway.

"Or are you going to use water every day If you take that water cooling loop inside your radiator, "or every hour of every day If you take that water cooling loop inside your radiator, "to cool the water loops for the data center?"

Sometimes you use a lake or a river or water So it's really around cooling the chips.

And this is where there is a lot of opportunity to cool that water loop.

for innovation and research.

And that's part of the tension.

But if you go to a community that doesn't have much water, "to cool your facility?

I'm thinking like West Texas, not much water, "Is it a one-time charge where you're loading up the water a lot of competition for that water, "in your closed loop cooling and then you're done?

and you say, "We're going to build a data center, "Or are you going to use water every day "we're going to create all these jobs "or every hour of every day "and we're going to take all your water," "or every hour of every day this is a real sensitive point for people "to cool the water loops for the data center?"

because it's irreplaceable.

So it's really around cooling the chips.

So there's a lot of pressure and pushback on data centers.

for innovation and research.

Like, "We need you to use air cooling wherever possible.

But if you go to a community that doesn't have much water, "We need water efficiency, thermal efficiency, I'm thinking like West Texas, not much water, "reduce your water or do other things."

a lot of competition for that water, And we can talk about what some of the other things are, a lot of competition for that water, but water is a very sensitive point.

"and we're going to take all your water," And I would say from Texas, this is a real sensitive point for people water is a more sensitive point than electricity probably.

because it's irreplaceable.

We have pretty cheap electricity in Texas, So there's a lot of pressure and pushback on data centers.

we'd like to keep it that way, Like, "We need you to use air cooling wherever possible.

but the water feels even more scarce.

"We need water efficiency, thermal efficiency, So that's a real sore point.

"reduce your water or do other things."

Right, I mean, one of the last programs And we can talk about what some of the other things are, we just did here on "Horizons" but water is a very sensitive point.

was about the Colorado River running dry And I would say from Texas, and all these states in the American Southwest We have pretty cheap electricity in Texas, sucking more and more out of a diminishing resource.

but the water feels even more scarce.

And if you're suddenly putting data centers but the water feels even more scarce.

on top of that, you can see why So that's a real sore point.

there's a lot of resistance to that.

Right, I mean, one of the last programs Yeah, but the same kind of comment with electricity, was about the Colorado River running dry the data center companies are very rich.

and all these states in the American Southwest Maybe they can invest in more water efficiency or water production systems sucking more and more out of a diminishing resource.

or wetlands restoration or desalination.

And if you're suddenly putting data centers They have the money and the desire on top of that, you can see why to invest in water systems.

there's a lot of resistance to that.

They might end up improving the water system Yeah, but the same kind of comment with electricity, if we ask them to and they agree and everything else.

Yeah, but the same kind of comment with electricity, So it doesn't just have to be that they're going the data center companies are very rich.

to take all the water, it can be their part of the solution.

Maybe they can invest in more water efficiency I mean, you're touching again on a point or wetlands restoration or desalination.

that really does come up a lot of times They have the money and the desire where local communities to invest in water systems.

look at the contracts that are signed.

They might end up improving the water system Oftentimes the companies will ask So it doesn't just have to be that they're going for non-disclosure agreements So it doesn't just have to be that they're going so that the deals themselves to take all the water, it can be their part of the solution.

that the data center construction companies I mean, you're touching again on a point or the data center operators sign are really opaque.

where local communities And people think, "What on earth is going on here?

look at the contracts that are signed.

"Why are they speeding this through?"

Oftentimes the companies will ask Like, are there examples of places where communities so that the deals themselves have said, that the data center construction companies "We want you to do this in the way that the data center construction companies "that doesn't harm our community?"

or the data center operators sign are really opaque.

Are there effective examples that you would point to or the data center operators sign are really opaque.

to say, "Hey, here's an example And people think, "What on earth is going on here?

"where we did one of these right?"

"Why are they speeding this through?"

There is actually a lot of good examples.

Like, are there examples of places where communities And I think those are useful for us to know that, have said, okay, we're not trapped in this negative path.

"We want you to do this in the way We can look for a better opportunity.

"that doesn't harm our community?"

It's not the standard yet.

Are there effective examples that you would point to It tends to be one-off examples.

to say, "Hey, here's an example Meta has a water positive facility to say, "Hey, here's an example they're planning outside El Paso.

"where we did one of these right?"

Google has done a really clever project There is actually a lot of good examples.

with renewables in Minnesota.

Microsoft switched to air cooling And I think those are useful for us to know that, outside of San Antonio.

okay, we're not trapped in this negative path.

So you can think of the big hyperscalers.

okay, we're not trapped in this negative path.

These are brand names We can look for a better opportunity.

we've all worked with or use or know.

It tends to be one-off examples.

There are examples where they've done it in a really good way.

Meta has a water positive facility It's not the default, it's not the standard, they're planning outside El Paso.

and so I think that's part of all of us Google has done a really clever project working together to say, "Hey, the minimum standard Microsoft switched to air cooling "should be quite good."

outside of San Antonio.

And then let's make it better as time goes on outside of San Antonio.

and we get more innovative.

These are brand names But the challenge is that where the locations are, we've all worked with or use or know.

they're not always within a jurisdiction There are examples where they've done it in a really good way.

where there's really clear authority There are examples where they've done it in a really good way.

about who regulates what.

It's not the default, it's not the standard, In fact, a lot of the development It's not the default, it's not the standard, are in some of these unincorporated areas or maybe you have county commissioners or something, working together to say, "Hey, the minimum standard but you don't quite have like a city council "should be quite good."

and a mayor or something.

And then let's make it better as time goes on So there's a lot of ambiguity sometimes But the challenge is that where the locations are, about who's in charge or how to respond.

But the challenge is that where the locations are, And there's a lot of stealth that happens.

where there's really clear authority It's not always for malicious, malign behavior.

In fact, a lot of the development Sometimes the stealth is they're trying to get the land are in some of these unincorporated areas to build the data center and they don't want are in some of these unincorporated areas their competitors to know.

or maybe you have county commissioners or something, So there is stealthy behavior as always happens in business.

So there's a lot of ambiguity sometimes And that arouses a lot of suspicions So there's a lot of ambiguity sometimes because people are like, "What's going on?

about who's in charge or how to respond.

"What are you going to build?

Where, why, why didn't I know?

And there's a lot of stealth that happens.

"And do we get to be part of the process It's not always for malicious, malign behavior.

"to help you make sure it's going to be It's not always for malicious, malign behavior.

"sort of harmonious with society?"

Sometimes the stealth is they're trying to get the land So there are a lot of examples of bad actors their competitors to know.

that have done it in a sort of too quiet, So there is stealthy behavior as always happens in business.

too malicious way, maybe or dishonest way.

And that arouses a lot of suspicions And then some good examples.

because people are like, "What's going on?

And I would say, well, let's push for the good examples.

because people are like, "What's going on?

We know it's possible.

"What are you going to build?

Where, why, why didn't I know?

Let's make that normal.

"And do we get to be part of the process Let's not make that the one-off exception.

"to help you make sure it's going to be Right.

One of the other concerns that we have seen "sort of harmonious with society?"

this in a lot of local news reporting that have done it in a sort of too quiet, is the noise of these facilities.

too malicious way, maybe or dishonest way.

I mean, they are just quite enormous.

And then some good examples.

And sometimes, if they are in proximity And I would say, well, let's push for the good examples.

to where people live, Let's make that normal.

that is just a cacophony that people do not want.

Right.

One of the other concerns that we have seen What do we know about how loud these things are?

this in a lot of local news reporting They're loud.

is the noise of these facilities.

But there are a couple of different sources is the noise of these facilities.

of the noise to pay attention to.

I mean, they are just quite enormous.

One is if you're inside the data center.

And sometimes, if they are in proximity When I walk through data centers And sometimes, if they are in proximity and I've been to many, to where people live, you're often given ear protection, that is just a cacophony that people do not want.

like earplugs or ear mitts or earmuffs or something that is just a cacophony that people do not want.

of some sort to cover your ears What do we know about how loud these things are?

because it's so noisy inside the data center What do we know about how loud these things are?

from all the airflow.

They're loud.

They have to push so much air through the facility of the noise to pay attention to.

to cool the chips and computers.

One is if you're inside the data center.

It's quite noisy.

When I walk through data centers So you hear the constant hum of the fans moving the air.

you're often given ear protection, That noise inside, you don't really hear outside though.

like earplugs or ear mitts or earmuffs or something There's sound insulation for the most part.

of some sort to cover your ears Outside, you have a lot of generators.

because it's so noisy inside the data center You mentioned those diesel generators, from all the airflow.

maybe those like aeroderivative jet engines They have to push so much air through the facility that are bolted to the ground to give you power on site.

to cool the chips and computers.

Those diesel generators or gas generators So you hear the constant hum of the fans moving the air.

or gas turbines, if those are all noisy, That noise inside, you don't really hear outside though.

you'll hear that as well.

That noise inside, you don't really hear outside though.

And then there's all the truck noise and other things.

There's sound insulation for the most part.

There's people coming and going.

There's sound insulation for the most part.

Certainly during construction, there's a lot of noise.

Outside, you have a lot of generators.

And so if you move to the rural countryside maybe those like aeroderivative jet engines to get away from the city noise, that are bolted to the ground to give you power on site.

and all of a sudden you've got a bunch of trucks that are bolted to the ground to give you power on site.

and generators and fans and cooling towers, Those diesel generators or gas generators it's noisier, you can hear it.

or gas turbines, if those are all noisy, That can be solved usually by distance, And then there's all the truck noise and other things.

by getting more land and having a little more space There's people coming and going.

between you and the fence line, Certainly during construction, there's a lot of noise.

that would help if you have the space to do it.

And so if you move to the rural countryside And that's one reason why they don't build to get away from the city noise, in urban areas is the noise is, and all of a sudden you've got a bunch of trucks well, the land's expensive, but the noise is also a problem.

and generators and fans and cooling towers, But there's several sources for the noise.

That can be solved usually by distance, How much of this, Michael, That can be solved usually by distance, do you think is a sort of mismatch between, we're doing all of this massive construction in urban areas is the noise is, for the promise of the AI revolution, well, the land's expensive, but the noise is also a problem.

which right now, yes, we can use ChatGPT and all of that, But there's several sources for the noise.

and I can figure out a recipe But there's several sources for the noise.

or get an answer to some question How much of this, Michael, that my son is asking me about.

do you think is a sort of mismatch But the real revolution is still in the future between, we're doing all of this massive construction and the present day madness for the promise of the AI revolution, is in people's backyards.

which right now, yes, we can use ChatGPT and all of that, Without a doubt.

which right now, yes, we can use ChatGPT and all of that, I think we're always suspicious of new stuff anyway.

or get an answer to some question That's like kind of a human condition.

that my son is asking me about.

But we've got this new stuff showing up.

But the real revolution is still in the future It's very visible, very loud.

But the real revolution is still in the future It might be expensive for electricity.

But the real revolution is still in the future It might take our water.

But the real revolution is still in the future So we've got all these concerns and the present day madness and we don't even know what it's good for.

is in people's backyards.

Maybe it's better cat videos, this kind of thing.

is in people's backyards.

And that's the challenge is the benefits of AI I think we're always suspicious of new stuff anyway.

haven't really shown up in visible ways yet, I think we're always suspicious of new stuff anyway.

but we should get better cancer treatments, That's like kind of a human condition.

we should be able to accelerate fusion reactor design.

It's very visible, very loud.

We should be able to optimize the grid.

It might be expensive for electricity.

Like there are all these really good things So we've got all these concerns we can get out of AI, and we don't even know what it's good for.

but those haven't shown up in a very obvious, Maybe it's better cat videos, this kind of thing.

visible way yet.

Maybe it's better cat videos, this kind of thing.

So we feel the impacts today, the benefits come later.

And that's the challenge is the benefits of AI That mismatch in time is a huge challenge but we should get better cancer treatments, because the benefits are kind of like, "Oh, we should be able to accelerate fusion reactor design.

"maybe I get a better search engine."

we should be able to accelerate fusion reactor design.

There are benefits from the economic development We should be able to optimize the grid.

of the jobs, but not all of us have a job Like there are all these really good things at those companies or get those economic benefits.

we can get out of AI, So there's a mismatch in who suffers, visible way yet.

people in the rural countryside So we feel the impacts today, the benefits come later.

who now have a noisy or loud or whatever it is, That mismatch in time is a huge challenge data center, very visible, ugly ones, for example, "maybe I get a better search engine."

versus who benefits, which are people later on, There are benefits from the economic development or maybe some shareholders.

There are benefits from the economic development That mismatch in who benefits at those companies or get those economic benefits.

and who feels the impact and when we get the benefits So there's a mismatch in who suffers, and when we feel it, that's a huge problem.

people in the rural countryside The best example today we're getting from AI who now have a noisy or loud or whatever it is, that people might understand is these self-driving cars data center, very visible, ugly ones, for example, like Waymo, which are dramatically safer versus who benefits, which are people later on, than human-driven cars.

versus who benefits, which are people later on, They're like 80% reduction in accidents That mismatch in who benefits and fatalities, other things.

That mismatch in who benefits So that's all enabled by autonomy and who feels the impact and when we get the benefits and artificial intelligence.

and when we feel it, that's a huge problem.

That is a great example of AI making traffic safer.

The best example today we're getting from AI And that's one of the benefits we'll get.

that people might understand is these self-driving cars So that's one we kind of see now at small scale in some cities, than human-driven cars.

but we really want all those other big benefits They're like 80% reduction in accidents and then it will transform the world.

They're like 80% reduction in accidents But meanwhile, we're suffering the impacts of the fumes, So that's all enabled by autonomy the water, the electricity, the land, the noise, and artificial intelligence.

the light, the heat, everything else.

That is a great example of AI making traffic safer.

And I think too, I want to put this poll up.

And that's one of the benefits we'll get.

This is a Pew Research poll about people's views of AI.

but we really want all those other big benefits According to Pew, half of Americans are, quote, But meanwhile, we're suffering the impacts of the fumes, "more concerned than excited But meanwhile, we're suffering the impacts of the fumes, "about the increased use of AI in daily life."

the light, the heat, everything else.

And the same poll, a slightly smaller percentage And I think too, I want to put this poll up.

think it will make it harder to think creatively, This is a Pew Research poll about people's views of AI.

that it'll worsen our relationships, This is a Pew Research poll about people's views of AI.

that it'll have a negative effect on the economy According to Pew, half of Americans are, quote, and on education.

According to Pew, half of Americans are, quote, Do you think that that, again, "more concerned than excited is also part of the equation here, "more concerned than excited that people seem to think, "about the increased use of AI in daily life."

"I'm going to get laid off by AI.

And the same poll, a slightly smaller percentage "It's going to make my kids dumber.

And the same poll, a slightly smaller percentage "Everyone's cheating on their homework with it.

think it will make it harder to think creatively, "Why are we building these massive data centers that it'll worsen our relationships, "to support all these things that I think of as negative?"

and on education.

Without a doubt, all these human condition concerns is also part of the equation here, or ethical concerns, like what is AI?

"I'm going to get laid off by AI.

Why is it good?

How is it going "I'm going to get laid off by AI.

to change my relationship with my fellow neighbor "Everyone's cheating on their homework with it.

or my family and everything else?

"Everyone's cheating on their homework with it.

These are real concerns.

"Why are we building these massive data centers We see it at the universities.

"Why are we building these massive data centers We already have to figure out how to deal with it, "to support all these things that I think of as negative?"

with cheating and everything else.

"to support all these things that I think of as negative?"

So there are some downside risks, without a doubt.

Without a doubt, all these human condition concerns And I think that's true with every new technology.

or ethical concerns, like what is AI?

We said some of the same things with the arrival of the internet when I was in graduate school.

to change my relationship with my fellow neighbor And Wikipedia, was it cheating to use Wikipedia?

or my family and everything else?

People would argue just the arrival of computers We see it at the universities.

decades before was cheating.

We already have to figure out how to deal with it, You should have been able to do math with cheating and everything else.

with an abacus or your own by hand So there are some downside risks, without a doubt.

or a slide rule or something else.

And I think that's true with every new technology.

So we've had new tools come along.

We said some of the same things with the arrival We had the same kind of fights every time.

of the internet when I was in graduate school.

Socrates was concerned when writing was developed.

And Wikipedia, was it cheating to use Wikipedia?

He thought if we used writing People would argue just the arrival of computers instead of committing things to memory, People would argue just the arrival of computers it would make us weak as learners.

decades before was cheating.

So they were afraid of the new technology, writing.

with an abacus or your own by hand And so this is kind of a constant thing.

or a slide rule or something else.

We're always afraid of the new thing So we've had new tools come along.

because we don't really understand the benefits yet.

We had the same kind of fights every time.

And there are downside risks for sure of all this.

Socrates was concerned when writing was developed.

I think that like any tool He thought if we used writing that could be used for good or bad, instead of committing things to memory, you can use a hammer for good or bad.

it would make us weak as learners.

It's a tool.

it would make us weak as learners.

You can use AI for good and bad.

So they were afraid of the new technology, writing.

And for the most part, humans will use it for good.

And so this is kind of a constant thing.

And some people will use it for bad.

We're always afraid of the new thing That's a real ethical concern.

because we don't really understand the benefits yet.

But if you're worried about it's going to lay me off And there are downside risks for sure of all this.

and make my kids stupid And there are downside risks for sure of all this.

and take all our water I think that like any tool and make our electricity more expensive, I think that like any tool it's like, yeah, this is a big set of concerns.

that could be used for good or bad, And I think that's on us to figure out It's a tool.

on us or everyone who's developing AI to start, You can use AI for good and bad.

"Okay, is this a tool we can live with?

And for the most part, humans will use it for good.

"If so, do we need rules?

And some people will use it for bad.

"What are those rules going to be to ensure But if you're worried about it's going to lay me off "it's safe for our children and everybody else?"

and make my kids stupid So I think that's part of the conversation and a huge part of the concern.

and make our electricity more expensive, And kind of rightly so, And I think that's on us to figure out I think we'll get through it because we'll start to see on us or everyone who's developing AI to start, some good use cases, on us or everyone who's developing AI to start, but it's really hard right now.

"Okay, is this a tool we can live with?

Do you think the companies, again, "What are those rules going to be to ensure you and... we both have been saying how "it's safe for our children and everybody else?"

they are all in a headlong rush and a huge part of the concern.

to get to be first to market, And kind of rightly so, to get these centers up to speed.

I think we'll get through it because we'll start to see Do you think that they appreciate but it's really hard right now.

and see this burgeoning opposition Do you think the companies, again, all over the country you and... we both have been saying how and they recognize, you and... we both have been saying how one, they have to address it more they are all in a headlong rush and maybe make the case more publicly to people to get to be first to market, as to here's why we're imposing on your communities.

to get these centers up to speed.

And two, here's what the possible benefits are.

and see this burgeoning opposition Do you think that they got that message and see this burgeoning opposition and that they are doing that?

and see this burgeoning opposition I think they got the message and they're starting now.

and they recognize, I also think they're very late.

one, they have to address it more They did not start with that.

and maybe make the case more publicly to people So I feel like they have moved ahead longer.

as to here's why we're imposing on your communities.

This is in a race.

as to here's why we're imposing on your communities.

It feels like an arms race that there's only going to be as to here's why we're imposing on your communities.

a few winners and they And two, here's what the possible benefits are.

want to be one of the few winners.

And two, here's what the possible benefits are.

So speed matters more than niceties and that they are doing that?

or getting along well with your neighbors.

I think they got the message and they're starting now.

And now they're finding out, actually getting along I also think they're very late.

well with your neighbors is part of the condition They did not start with that.

for having speed because your neighbors So I feel like they have moved ahead longer.

can slow you down.

So I feel like they have moved ahead longer.

So I think they're figuring it out kind of late.

This is in a race.

And this reminds me of the hydraulic fracturing a few winners and they or shale revolution that happened want to be one of the few winners.

about 15 to 20 years ago So speed matters more than niceties where a lot of these oil and gas companies or getting along well with your neighbors.

looking to produce shale And now they're finding out, actually getting along moved into these rural communities And now they're finding out, actually getting along with this new technology well with your neighbors is part of the condition that would unlock a lot of domestic oil and gas for having speed because your neighbors from these shale formations.

can slow you down.

And they thought they'd be treated as heroes So I think they're figuring it out kind of late.

liberating us from the yoke And this reminds me of the hydraulic fracturing of Middle Eastern oil and gas.

And this reminds me of the hydraulic fracturing And instead people got mad because of all the noise about 15 to 20 years ago and trucks and everything else.

about 15 to 20 years ago The same kind of story.

where a lot of these oil and gas companies And they figured out a few years in looking to produce shale like, "Oh, we need to go talk to the neighbors with this new technology "and really sort this out that would unlock a lot of domestic oil and gas "because they see us as a part of the problem And they thought they'd be treated as heroes "rather than part of the solution."

liberating us from the yoke AI companies see that now too.

of Middle Eastern oil and gas.

They thought they'd be heroes.

And instead people got mad because of all the noise "We're bringing in new tools, we're going to make you rich, And instead people got mad because of all the noise "we're going to employ all these people, and trucks and everything else.

"we're going to beat the bad guys."

The same kind of story.

And now they're finding out people perceive these companies like, "Oh, we need to go talk to the neighbors as the bad guys.

So they're having to play some catch up.

"and really sort this out And that's pretty typical.

"because they see us as a part of the problem That often happens when there's a sort of a land rush "rather than part of the solution."

or gold rush, so to speak, AI companies see that now too.

where they move quickly, make mistakes, "We're bringing in new tools, we're going to make you rich, then figure out, "Oh, we need to go back "we're going to employ all these people, "and manage this better."

"we're going to beat the bad guys."

So they've got the message, they're working on it.

And now they're finding out people perceive these companies I don't think they've solved it yet.

And now they're finding out people perceive these companies And this reminds me of the shale revolution.

as the bad guys.

So they're having to play some catch up.

I've seen this before for oil and gas.

And that's pretty typical.

There are, in this conversation about data centers, or gold rush, so to speak, you inevitably will hear Elon Musk, Google, then figure out, "Oh, we need to go back others talk about this, "and manage this better."

that, okay, So they've got the message, they're working on it.

maybe there are limits to what we can do I don't think they've solved it yet.

on the land with regards to these data centers and we ought to go I've seen this before for oil and gas.

to outer space and build them up there.

There are, in this conversation about data centers, Explain the rationale why that might be feasible you inevitably will hear Elon Musk, Google, and why they would want to do that.

that, okay, Outer space or subsea, maybe there are limits to what we can do offshore and below water, right?

on the land with regards to these data centers So it strikes me as an engineer is just so difficult.

and we ought to go Offshore is so much harder than onshore to outer space and build them up there.

because you got to deal with the waves and the wind Explain the rationale why that might be feasible and the saltwater, the corrosion, Explain the rationale why that might be feasible the water tightness, everything else.

Explain the rationale why that might be feasible And then that's easy compared to space.

and why they would want to do that.

Space is really hard.

and why they would want to do that.

Like these are really hard places to make things work.

offshore and below water, right?

To be space rated means something different So it strikes me as an engineer is just so difficult.

than to be suitable for fuse on land or offshore.

Offshore is so much harder than onshore So the idea that you might go offshore because you got to deal with the waves and the wind and below water or out into outer space the water tightness, everything else.

and out of the atmosphere And then that's easy compared to space.

is driven by some sort of necessity, Space is really hard.

I think is what's happening, where they feel Like these are really hard places to make things work.

like it's just too slow to build the grid the way we need it.

To be space rated means something different It's too hard to get the power we need.

than to be suitable for fuse on land or offshore.

There are too many neighbors who complain.

than to be suitable for fuse on land or offshore.

Let's go where there aren't neighbors who can complain, So the idea that you might go offshore offshore or outer space.

and below water or out into outer space And so you can get a couple of things.

and out of the atmosphere Fewer people complain is driven by some sort of necessity, because you're out of sight, out of mind.

is driven by some sort of necessity, You get your cooling for free.

I think is what's happening, where they feel Deep space is quite cold, the ocean's quite cold.

like it's just too slow to build the grid the way we need it.

So it gets a lot of cooling.

It's too hard to get the power we need.

So that reduces your energy needs.

It's too hard to get the power we need.

But also you might be able to get solar power Let's go where there aren't neighbors who can complain, in space more easily than building Let's go where there aren't neighbors who can complain, a huge solar farm on the ground.

Now, I'm skeptical.

offshore or outer space.

The engineering of that all seems really, really hard.

Fewer people complain But if you believe that the cost to go to space You get your cooling for free.

will drop, the cost for solar panels will drop, Deep space is quite cold, the ocean's quite cold.

that the cost to expand the grid on ground will double, But also you might be able to get solar power then it starts to look like space might make sense.

in space more easily than building If you think neighbors are always getting in your way, a huge solar farm on the ground.

Now, I'm skeptical.

going offshore for the cooling water The engineering of that all seems really, really hard.

and to get away from the neighbors The engineering of that all seems really, really hard.

might make sense as well.

But if you believe that the cost to go to space That strikes me as really hard, frankly.

But if you believe that the cost to go to space But really smart people who are developing this, will drop, the cost for solar panels will drop, who are thinking about it, they're all looking at it.

that the cost to expand the grid on ground will double, So I might be wrong on this, then it starts to look like space might make sense.

but I would say it's going to be easier then it starts to look like space might make sense.

to build it on earth, on ground than in space.

going offshore for the cooling water But SpaceX and other companies are trying to drive down might make sense as well.

the lift costs to make it cheaper to go to space.

That strikes me as really hard, frankly.

Maybe it will make sense.

But really smart people who are developing this, Just in the last 30 seconds we have or so, who are thinking about it, they're all looking at it.

you do believe though, you're optimistic, So I might be wrong on this, that we have enough agency to do it right here.

to build it on earth, on ground than in space.

Let's do it right.

I mean, we can make these data centers the lift costs to make it cheaper to go to space.

power flexible, water positive, carbon negative, Maybe it will make sense.

and really harmonious with the land.

Just in the last 30 seconds we have or so, Let's just do it.

Let's demand it of our partners.

I say, let's do it that way.

that we have enough agency to do it right here.

Michael Webber, it is always great to hear from you, that we have enough agency to do it right here.

but especially on such a controversial Let's do it right.

I mean, we can make these data centers and confused and convoluted topic.

Let's do it right.

I mean, we can make these data centers You're such a clear voice on this.

power flexible, water positive, carbon negative, Thank you so much.

power flexible, water positive, carbon negative, Michael Webber at the University of Texas.

and really harmonious with the land.

Great to talk to you.

Thank you.

Let's just do it.

Let's demand it of our partners.

Thank you.

I say, let's do it that way.

Michael Webber, it is always great to hear from you, And that is it for this episode of "Horizons."

but especially on such a controversial You can watch us on YouTube and confused and convoluted topic.

or listen wherever you get your podcasts.

and confused and convoluted topic.

If you like what you hear, You're such a clear voice on this.

please subscribe and give us a rating.

Michael Webber at the University of Texas.

That single signal will help more people see this program.

Thank you.

Thank you so much for watching.

We'll see you next week.

And that is it for this episode of "Horizons."

Narrator: Support for "Horizons" has been provided or listen wherever you get your podcasts.

If you like what you hear, by Steve and Marilyn Kerman please subscribe and give us a rating.

and the Gordon and Betty Moore Foundation.

That single signal will help more people see this program.

Additional support is provided by Friends of the News Hour.

We'll see you next week.

♪ Additional support is provided by Friends of the News Hour.

This program was made possible by contributions ♪ to your PBS station from viewers like you.

♪ Thank you.

♪ ♪ ♪ You're watching PBS.

♪