How AI boom is impacting Latin America and the US

DW (English)

Plan for data centers have led to protests in Uruguay, Chile and US. Should server farms be built in already dry regions? What are the alternatives?

Transcript

00:00Plans to construct data centers often meet a lot of pushback.

00:05As in other parts of the world, Latin America has seen protests.

00:10In 2023, severe water shortages gripped the country of Uruguay.

00:15It was the worst drought in 75 years.

00:18At this particularly dry time, Google showed up.

00:21They wanted to use 7.6 million liters of water per day, enough for 22,000 people,

00:30and during a drought, an extreme water crisis.

00:32Daniel Pena, a scientist and activist in the capital Montevideo,

00:37says the community wasn't consulted.

00:42There was no public dialogue and no public hearing.

00:46The project was classified as category B, considered low to moderate risk.

00:52As a result, public involvement was minimal.

00:57Daniel Pena is concerned about the impact of the Google Cloud project

01:02on the Santa Lucia River, which is important for Montevideo.

01:07The battle over water has already begun here.

01:10When there's no clean water from the tap, bottled water is the only option left.

01:15Data centers, the backbone of AI, need massive computing power.

01:19Power runs hot and cooling takes huge amounts of water.

01:27These huge centers are often built in some of the world's driest regions.

01:32More on that later.

01:35Google says it's investing more than US$850 million in the Uruguay center.

01:40This is what the building site looked like in summer 2025.

01:45Construction in Canelones, which is near Montevideo, has been underway for a year.

01:51Municipal authorities have welcomed the project.

01:54We hope that once it's up and running, they can generate synergies that can attract other

02:01types of industries and activities associated with technological innovation.

02:06The data center industry's biggest global players are Amazon Web Services or AWS, Microsoft Azure and Google Cloud.

02:19In many server hubs, most cooling water simply evaporates through towers. It's a standard practice.

02:25In fact, evaporation is the most efficient way of removing heat. And that's why we, for example,

02:35we sweat. When we get too hot, that evaporation of our sweat off of our skin is an effective way of

02:41keeping us cool and keeping us from overheating. Globally, the water consumption of data centers is

02:46set to double by 2030. After recent protests, Google spoke of a commitment to responsible water use.

02:53It now plans to cool its data center in Uruguay with air.

02:58Sounds simple, but it's not. Water cools more effectively.

03:02Air cooling requires more energy, which often comes from fossil fuel plants.

03:08It is possible or likely that many waterless cooling systems are less energy efficient

03:16compared to systems that utilize evaporation. And as a result,

03:21you end up just having a higher water footprint upstream at the power generation source as opposed

03:27to at the data center site itself. We've seen this before in Chile, for instance,

03:33where Google's proposed computing facility also prompted protests over excessive water use.

03:40Santiago's environmental court partially revoked the project's approval.

03:44Now Google has to revise its plans and will likely switch to air cooling.

03:51So why build so many data centers in hot, dry regions with scarce water?

03:57The main reason for looking for dry areas is both an efficiency reason and an avoidance of corrosion.

04:06Some experts are calling for these digital facilities to be built in cooler regions for environmental reasons.

04:14Like here in Sweden. In Faulun, about 200 kilometers from Stockholm, there's no shortage of water.

04:21And the electricity it uses is 100 percent renewable.

04:30During wintertime, we have cold temperature. We have like 20, normally 20 degrees below zero,

04:36which gives us free cooling, actually.

04:40The Swedish computer center is also built from wood.

04:44That's a common local resource that locks in carbon.

04:55Now we're up to 95 percent of wood.

04:57And we find out that this is the most perfect way of building a computer center.

05:04In the US, another massive data center recently opened in Memphis, Tennessee.

05:08Tech multi-billionaire Elon Musk is behind the AI's supercomputer facility, which was completed in just 122 days.

05:18Unusually fast for a project of this scale.

05:21Many residents are asking what its arrival will mean for their most vital resource.

05:27Industries do need water for their process. Businesses do need water to make their revenue.

05:33But that should never be put in front of the drinking water needs of residents.

05:38Its electricity comes from methane gas turbines.

05:41Residents in this predominantly black, marginalized area worry about air and groundwater pollution

05:47and ignore environmental laws.

05:50Globally, data center transparency is a big problem.

05:53Tech giants like Amazon, Microsoft and Google stand accused of failing to disclose

05:58key data on water and electricity use.

06:01Google declined our request to visit their site in Uruguay and turned down an interview too.

06:07Google is a big problem.

06:09Reliable data is extremely hard to come by.

06:13Most of these other aspects of water consumption are not being indexed, are not being measured,

06:20so it's really very intransparent.

06:23I have my doubts, yes, because until now what we can only do as researchers is try to use open source.

06:32We really don't know how much water they really consume.

06:36These tech giants say their goal is to be water positive by 2030, meaning they'll provide

06:42more water than they use. But they don't say how water positivity will work.

06:48So what can we do?

06:49Every one of us can take small steps to ease the computing load.

06:56For example, use a calculator instead of artificial intelligence for simple maths.

07:02Stick to a simple search engine instead of asking AI.

07:05Or make less resource-intensive videos.

07:15But that alone won't make much of a dent.

07:17AI use and its data demand is only growing.

07:21The real challenge is making tech more environmentally sustainable.

07:26mmm

07:322

07:334

07:384

07:405

07:444

07:464

07:495

Category

Transcript

Be the first to comment

Recommended