Getting to the Truth about AI Water Use

   

It might have been Mark Twain who said, or at least repeated, that a lie will get halfway around the world while the truth is still putting on its pants.  And that expression finds no better example than the way water usage by data centers has been portrayed over the last couple of years.

 

I live in Texas, which has seen some of the most rapid growth of data centers in the U. S.  There has even been a local controversy here in San Marcos, with citizen groups organizing rallies and protests against the construction of new data centers.  Every day when I go on my bike-exercise route, I pass two or three signs saying "No Data Centers" showing a big faucet with a single drop of water coming out.  Unfortunately, that sign represents about the typical level of discourse that prevails in discussions about the environmental concerns raised by data centers, specifically their water use. 

 

Last May, a book by Karen Hao, Empire of AI, hit the market and received a wildly enthusiastic reception.  Amazon currently ranks it the No. 1 best seller in International Economics and No. 3 in General Technology and Reference, a very broad category.  It won the National Book Critics Circle Award in 2025. 

 

And yet it contains numerous factual errors, many of which blogger Andy Masley pointed out in detail in his Substack blog over a period of months.  In discussing plans Google had for a data center in a town in Chile, Hao wrote that the data center would use 1,000 times more water than the city of 88,000 used. 

 

Masley, a close reader of the technical stuff in such a book, thought this looked strange, although it is not out of line with similar claims by others for the vast quantities of water that data centers will supposedly use.  So he looked into the problem.

 

When he took the number that Hao said was the city's total water usage and divided it by the population of 88,000, he got the result that each citizen was using less than a half-cup of water (0.2 liters) per day.  That is nonsense, of course.  When he found the original data, it turned out that the city's consumption was stated in cubic meters, not liters.  As later confirmed by Hao, someone on her staff of assistants mistook the cubic-meter consumption for liter consumption. 

 

As every student of the metric system knows, one cubic meter contains 1,000 liters, hence the error by a factor of 1,000. 

 

What Masley then wanted to know was how such an egregious error made it past the fact-checkers acknowledged in Hao's book, and the more than 1,000 reviews on Amazon, the adulatory comments in prominent media outlets such as the New York Times and The New Yorker, and the people who decide on the National Book Critics Circle award.

 

The only answer he could come up with is that the people writing the reviews don't give a flip about numbers.  And the people who care about numbers don't write book reviews, except for him, evidently.

 

Hao has since admitted that the error happened in the manner Masley suspected.  But that's a little bit like apologizing for not wiping your feet on the welcome mat after you burn your host's house down.  Masley has spent what amounts to a fulltime job calling out and correcting this and innumerable other errors in the vast outpouring of what may well be called propaganda opposing the construction of new data centers.  His words are finally gaining some traction in the media, which is why I now know about his work.

 

But something is way out of whack with public discourse if one side of a controversy can put out tons of material that is chock-full of factual errors, and these errors can go unchallenged for months or years before one person with the willingness to do the math and find and expose the errors gets enough attention to be heard from.

 

This story reveals laziness and groupthink among those who control prominent media outlets, plus something that could be called math avoidance.  You would think that at a time when mathematics has never been as vital an underpinning to civilization as it is now, that most people would be at least somewhat used to doing math when some issue that vitally involves math comes up.

 

But this story about liters and cubic meters shows that a lot of people who should do their math homework before blatting the glories of a book, or taking sides in a controversy, simply don't.  They find an expert whose vibes agree with their own and enthusiastically fall in line without the slightest regard for accuracy or healthy skepticism. 

 

I don't know what the future of data centers holds.  Their wild growth is starting to look to me like some sort of bubble, much as we had a wild growth of fiber-cable capacity in the early days of fiber-optics, and for years the industry contended with the problem of "dark fiber"—too much capacity.  Dark fiber cables don't use water or power—they just lie there.  But data centers need both, and it's possible that we may find ourselves awash in surplus data-center capacity and the boom will become a bust. 

 

But so far, there seems to be no limit to the appetite for "compute," as the experts call it, that AI wizards have.  The only limit at this point will be economic.  If the people who build the capacity can't figure out how to make money with it, they'll eventually quit adding capacity.  But as long as the public and corporations are willing to get more AI results—and so far, they are—the demand will drive the supply and more data centers will be built.

 

That is, unless people keep on believing the nonsensical claims of the more extreme opposition to data centers, and do something like getting a nationwide ban on new construction passed.  Then we'll get to see what a data-center shortage looks like, and it might be worse than you think.

 

Sources:  Numerous news outlets have covered Andy Masley's debunking efforts directed at correcting errors in the claims of data-center opponents.  The details of the particular Empire of AI error described herein can be found at his Substack post https://blog.andymasley.com/p/empire-of-ai-is-wildly-misleading. 

Whither the U. S. Postal Service?

  

We don't usually think of mail as a technology.  But if we define a technology broadly as any system engineered for the accomplishment of a practical purpose, the U. S. Postal Service is not only a technology, but a vital one.  Like any technology that doesn't go extinct, it has to change with circumstances or die.  And those two alternatives are becoming more obvious by the day as competition from electronic media bring intense pressures on so-called "snail mail" services, not only in the U. S. but worldwide. 

 

This column is brought on by an incident which nonetheless may be symptomatic of wider problems in the system.  Ever since my father taught me how to use a checkbook, I have paid for many monthly bills by mailing checks.  Until last month, this was a reliable way to pay things like utility bills.  But in February and March of this year, four checks I mailed simply disappeared, including the payment for the electric, water, and sewer bill. 

 

These incidents have forced me to join most of the rest of the world in switching to electronic payments for those bills.  But it also made me wonder how the U. S. Postal Service is doing in general, and the answer is:  not well.

 

At its inception under the guidance of the first U. S. postmaster—some dude named Benjamin Franklin—the Post Office, as it was known then, became a powerful nation-binding force as it put even the most remote state or territory in contact with the rest of the country by both private letters and favorable rates for periodicals such as newspapers and magazines.  It was operated as a Cabinet-level department and not expected to show a profit.  That remained unchanged until thousands of postal workers struck in the largest wildcat (=unauthorized by union leadership) strike in U. S. history in 1970.  That led directly to the passing by Congress of the Postal Reorganization Act of 1970, which President Nixon signed as a way of giving the postal unions the right to collective bargaining, though it still made strikes illegal.

 

The Act did more than authorize unions, however.  It changed the name of the organization to the U. S. Postal Service and set it up as a quasi-independent corporation that was expected to be self-supporting without government subsidies.

 

In 1970, the highly-remunerative monopoly on first-class letter carrying enjoyed by the USPS was more than enough to allow it to make a profit.  But the era of electronic communications was just around the corner.  If you are old enough to remember all the things that used to be done by mail that are now done by means of the internet, that's a lot of mail that has simply disappeared, and most of it was first-class mail.  Bills, checks, legal documents, and the whole volume of commercial first-class mail that used to support the old Post Office—virtually all that has now turned into bits transmitted on fiber cables. 

 

Some analyses by Elena Patel of the research institute Brookings show that the declining volume of first-class mail has led to the USPS showing a deficit every year since 2007.  By law, it can borrow money only from the U. S. Treasury, and there is a cap on its total indebtedness, which it has already hit.  When it can't borrow any more, it has to rely on its cash reserves, and these days those are running out too.  So we face the near-term prospect of the U. S. Postal Service going bankrupt unless its governing laws are changed.

 

This would have happened even earlier if the volume of package deliveries had not increased in a way that has partly compensated for the huge loss of first-class mail, which the USPS had a monopoly on.  But in package delivery, the USPS faces stiff competition from fully private businesses such as FedEx and UPS that operate on slimmer margins than a quasi-government service like the postal system, which has built-in labor costs and obligations to serve every single post office in the U. S. 

 

I don't know if all these adverse circumstances are directly responsible for my checks getting lost, but they didn't help.

 

So what should be done?  This problem of electronic-media competition upsetting the fiscal status of mail service is worldwide, not just in the U. S., and different countries are dealing with it in various ways.  In some places, the national government simply absorbs the losses and regards the mail service as a necessary part of national infrastructure.  That's the way our old Post Office began—as a nation-binding service that was simply paid for out of government funds—but by law the current USPS can't operate that way. 

 

Some might feel that the practice of laboriously carrying little pieces of paper around and physically delivering them is an outmoded practice that should be allowed to die a natural death.  But one of the Brookings studies shows that postal services form an important part of the economy of certain areas of the country, especially where population is sparse but people can still operate businesses with nationwide clientele through the postal service.

 

I don't have any brilliant solution to these problems.  But it's clear that things can't go on the way they're going, with the laws governing the USPS assuming economic conditions that simply no longer exist.  The first-class-mail monopoly that formerly subsidized everything else the USPS did has vanished as a source of profit.  And unless the federal government recognizes that what the postal system does is important enough to pay for with taxes, we will sooner or later hit a crisis resembling what recently struck the Transportation Security Administration, which wasn't funded in the latest Congressional budget.  This caused snarled air transportation as TSA workers increasingly showed a reluctance to go to work without being paid. 

 

Maybe a lot of young people wouldn't miss the postman (post-person, these days).  But one way to find out the significance of a technology is to imagine that tomorrow you woke up and all of it has vanished into thin air.  If the USPS doesn't get its finances straightened out by Congress soon, we may find out what that thought experiment looks like in reality.  And the results won't be good.

 

Sources:  I referred to two Brookings studies by Elena Patel on postal systems at https://www.brookings.edu/articles/postal-systems-worldwide-confront-the-same-financial-pressures/ and https://www.brookings.edu/articles/the-us-postal-services-fiscal-crisis/,

and the Wikipedia articles "1970 United States postal strike" and "Postal Reorganization Act."

 

US Fears AI, Uses It Anyway

  

A recent report in National Review summarized opinion polls about what U. S. residents think of artificial intelligence (AI) and how much they are using it.  Paradoxically, the more people use AI, the more they fear it.

 

A poll by NBC News showed that 46% of those queried had a negative opinion of AI versus only 26% positive.  Other polls show that citizens expect mostly or entirely negative effects on society from the widespread use of AI, and believe it will lead to serious job losses.  Over half the Americans polled by Democratic research firm Blue Rose feared that AI will lose them their job or a relative's job. 

 

At the same time, polls asking about AI use show that most people queried have used an AI tool in the past month, and a fourth say they use it every day.  So the old saying "familiarity breeds contempt" may be a guiding principle in how AI is viewed by the general public.

 

In a way, none of this matters.  If a new technology gets widely used and the companies providing it make money, who cares what people think about it?  Another technology that spread rapidly in only a few years, and also had profound effects on society, was television.  In 1950, only 9% of households had a TV, but by 1955 over half did.  And while there may have been a few voices raised in opposition to its growth, I think it's fair to say that the only groups that looked on the spread of TV with disfavor were industries threatened by it:  Hollywood, for instance.  And Hollywood has long ago made peace with the advent of television.  Your average person in the early 1950s was just waiting to see when TV sets got affordable enough to buy, and any negative consequences of TV use were not noted much in public before the 1960s.

 

One difference between the advent of TV and the advent of AI is that TV didn't threaten jobs like AI does.  And one job sector that is already seeing big effects from AI is computer science and computer programming.  The thing about public perception, regardless of whether it's accurate or not, is that it can easily become reality.  I work at a university, and I have heard in the last week that enrollment in computer-science programs is dropping across the board, after years of steady growth.  The reasons for this are not entirely clear, but one factor may well be that students fear spending four or more years getting a degree and then finding that all the entry-level positions are now being done by a few senior people writing AI prompts. 

 

On the other hand, one of the most enthusiastic proponents of AI I know is an 80-plus professor of biology who has been using ChatGPT in his research for the last year or two.  He says it helps him write papers more clearly and to organize his thoughts, and claims it's the greatest thing that's happened to him research-wise in a long time. 

 

Many of the polls mentioned were commissioned by political interests with a view toward forming policies about AI.  Currently, the Trump administration favors few if any regulations on the technology, and wants to keep states from enacting a patchwork of legislation that would encumber the field.  Historically, this approach has worked well for computer- and network-intensive industries themselves, allowing them to create vast new economies and profit mightily therefrom.  But it has also led to a number of real and lasting problems, ranging from the maleficent effects on politics of social media and the quantified and well-known harms to children and teenagers whose lives are distorted by the use of smartphones. 

 

The crystal ball of predicting how technologies will affect society is always more or less cloudy, and I will not venture to say what the future effects of AI's negative polling will be.  Even if AI were universally detested, it's not clear that Washington could get its act together enough to pass meaningful regulatory legislation, especially when Big Tech and the federal government sometimes seem to blur into each other.  On the state level, if the feds don't stop them, some states may pass laws attempting to regulate AI, but it's a little bit like trying to nail Jell-O to the wall.  When the thing you are trying to regulate is so protean and shape-changing, it's hard to decide what regulations to pass, let alone to figure out if they've been violated.

 

Some of the anxiety the public feels about AI is simply due to the breathtaking speed with which it has advanced and improved.  Arthur C. Clarke's principle that any sufficiently advanced technology is indistinguishable from magic applies here.  Real magic is scary if it happens, and I still feel a kind of queasiness when I type commands into a chat box and the program comes back with "I did this and that."  It's understandable that millions of teenagers use AI chatbots as a substitute friend, and it's also very creepy.

 

Going to extremes, a few people believe AI will engender the end of civilization as we know it.  Other hyper-tech-optimists such as Ray Kurzweil look forward to being uploaded to an eternal cloud and think it will be heaven on earth.  The truth probably lies somewhere in between.  What we can do as individuals is to keep reminding ourselves that AI systems are not human beings, and that human beings are not machines.  But both of those truths may become harder to keep in mind as time goes on. 

 

Sources:  The National Review website carried James Lynch's article "The More Americans Use AI, the More They Fear It" on Mar. 25, 2026 at https://www.nationalreview.com/news/the-more-americans-use-ai-the-more-they-fear-it/.