Robby Soave and Christian Britschgi discuss the brewing Texas showdown between Texas Attorney General Ken Paxton and Democratic Senate candidate James Talarico. Then, they break down Rep. Nancy Mace’s (R–S.C.) proposal to give boomers a property tax break and Pope Leo XIV’s latest encyclical on artificial intelligence. Finally, they wrap up with some lighter debates over The Legend of Zelda, Nicolas Cage movies, retro-futurism, Jill Biden’s latest remarks, and whether President Donald Trump’s political influence will ever fade.

0:00—Heretics and hypocrites in Texas

14:30—Talarico takes back his former wokeness

19:10—If you can’t take it, don’t dish it

32:25—Coal mines are cool?

34:00—Mace’s boomer luxury communism

39:20—The pope’s views on AI

47:40—Why does anyone play video games?

58:59—Nicolas Cage is a good actor

1:05:57—Retro-futurism

1:10:26—Jill Biden’s latest remarks

1:18:36—Will Trump’s influence ever fade?

The post James Talarico vs. Ken Paxton, the Pope on AI, and Caves appeared first on Reason.com.

Stewart Brand has spent decades shaping how we think about technology, the environment, and the future. He first came to prominence in the 1960s as co-creator of the Whole Earth Catalog, the counterculture bible that helped inspire personal computing, the hacker ethic, and the modern environmentalist movement. Since then, he’s launched the Long Now Foundation, championed nuclear power and de-extinction, and pushed us to think in 10,000-year time spans.

In his new book, Maintenance: Of Everything, Part One, Brand argues that the real work of civilization isn’t flashy invention but the long, patient care of complex systems. In March, he spoke with Nick Gillespie about what that means—and whether his vision of planetary stewardship conflicts with libertarian values of individualism, creative destruction, and decentralized power.

Reason: Your new book argues that maintenance is the hidden foundation of everything. What do we miss when we focus on innovation, creative destruction, and disruption and forget about checking that everything is tied down the right way on a daily basis?

Brand: I don’t think they’re opposed. A lot of innovation comes out of maintenance. People who figure out how to improve a thing are often the ones who are stuck with keeping it going and realizing how difficult that is. “Gee, we could make it easier this way or that way. Or what if we just throw this stupid thing away and get something better?” Which is all part of the process of keeping something going.

We often think of maintenance in terms of preventive maintenance. Repair is such a big hassle when something breaks. It’s a trauma to you and to the system that the thing is part of. We spend some of our time doing the very unrewarding thing of changing the oil and brushing your teeth so that your teeth don’t fall out and your car doesn’t blow up. But really maintenance is the whole complete process of keeping the thing going. For example, right now, I’m writing on the history of agriculture, because if you’re an animal, you’ve got to keep it fed. We are animals and we have to keep ourselves fed. The process of doing that has been one innovation after another.

You write about how interchangeable parts made it easier for people to fix things rather than throw the whole thing away, and how necessity was the mother of invention—living miles away from your neighbor, you had to figure out how to fix things yourself. Do you feel that, as a society, we still have that ethos, or have the machines we use to live and prosper become mysterious to us?

The Model T was designed to be maintained and tailored any old which way. Henry Ford grew up on a farm in the Midwest. He knew that farmers and ranchers were very good at fixing their own stuff, so he counted on that. The Model T stayed the Model T, and millions of them were made and used, and got old, went in the junkyards, and then got completely pillaged in the junkyards for the parts—because 20 years later, a part from a really old Model T would fit in your brand new Model T.

I’m glad you brought up the bit about interchangeable parts, because it’s probably the most anti-libertarian, anti–Reason magazine section of the book, in that it was government people who for 40 years in the War Department, the Ordnance Department specifically, spent millions of dollars, which now would be a lot more than that, trying to get manufacturers in the U.S. capable of making interchangeable parts. It turns out you have to get down to about a 50th of a millimeter accuracy in order for that to actually work. They did it. And that’s why America wound up taking the lead in the manufacturing part of the Industrial Revolution.

We led the way, a thing called the “American system.” It turned out that the way to get really good interchangeable parts was to basically automate the machine tools that made them.

Ford never allowed anybody to use a file on the assembly line, because as soon as you took a part and filed it to fit better, that would suddenly stop the assembly line. With guns, if your gun failed and you’re military and you’re in the field and you need to fix it, you’ve got to find a gunsmith, which is not going to be anywhere near the battlefield, and they’ll file a part to get it put in there or they’ll make it from scratch. Once they’re interchangeable, like the AK-47, any piece of any AK-47 will substitute just fine and they’re very roughly designed and built.

I see that process in the 19th century as being what led to [the Defense Advanced Research Projects Agency]. The researchers I counted on call it military enterprise.

In the book, you write about how the AK-47 was created by the Soviet Union and is the weapon of choice for armies and insurgents around the world because it is relatively simple, easily fixed, and there are lots of parts for it. You compare that to the M16, developed in America, which was rolled out en masse in Vietnam and was terrible for that. It has twice as many parts as the AK-47. It’s more subject to corrosion and rust and gunk getting in there. It seems like we’re always going between simpler products that might be more limited vs. something more sophisticated. Is there a sweet spot somewhere in between?

I don’t think so. I think you wind up being rewarded by going in both directions. AK-47s were cheap and M16s are not cheap. AK-47s operate in the mud. They operate in the sand. They operate in humidity. You can pick one up out of the mud and fire it, and as you fire, the mud flies out. [An] M16 basically has to be surgically cleaned to function. When it functions, it’s absolutely fantastic—at 500 meters you can put high-velocity small rounds through a helmet.

There’s always a place in not just the market, but a range of situations and mindsets, for things that are cheap, fast, and just barely in control, and things that are exquisitely attuned to perfection if they’re in relatively perfect circumstances. Exploring both of those, as a good direction to explore the whole field of design, is always worth doing.

Your work on, among other things, the campaign to have NASA release a photograph of the whole Earth helped inspire the first Earth Day. You helped create the environmentalist mindset. When you say they went too far, that carries a lot of weight.

Greenpeace were anti-technology. Romantics in general are anti-technology. John Henry is always going to get the song. The steam drill never does. But the steam drill is actually a better way to do that kind of work.

The French Revolution—where the idea of interchangeable parts for guns, or muskets then, was actually started—the French Revolution said, “Don’t do that, because these nice gunmakers will be out of work.” They actually shut down the guy named [Honoré] Blanc who was building interchangeable parts for muskets. France went, in the course of a generation, from having the best muskets in Europe to having the worst muskets in Europe. There was no taking account of “Was this beneficial for the customer?”—in that case, soldiers. And that keeps happening.

The anti-technology romantic is almost always: 1) wrong, and, 2) mistaken in his arguments. The way to critique a new anything—AI now—is to embrace it and experiment with it, make mistakes with it, see if you can break it, red-team it, do all these things to sharpen it. And if in the course of that you decide, “You know what, this is a blind alley and we need to back out of it,” the critique from people who have done that is a valid critique.

People are just imagining problems. You can always imagine more problems than you can imagine more ways of things actually going right.

What do you think is the strongest critique of AI as it’s being rolled out now? What are the things that we should be testing it on to see if it’s something that we’re not really going to embrace?

I don’t think of that. But I do know a lot of the people who are going to think of that, and I trust them to investigate the ways it could go wrong.

They’d be able to head off some of them. They won’t be able to head off all of them. Some of them will go probably pretty calamitous, and then we negotiate. The use of AI in weapons is clearly going to be played out very quickly, because the militaries always grab new technologies and turn them into weapons, and rightfully so from their standpoint. It could be a thing like the use of gas that gets tabooed after a while. The use of nuclear weapons, even tactical ones—that’s been tabooed for quite a while.

Everything has to do with a threat. And the massive amount of testing that went on in the ’50s and ’60s was negotiating with threat. But once you had mutual assured destruction with the second strike capability, that actually put a stop to it. That was, in a way, why it was developed. Some problems turn out to be nonexistent. Some problems turn out to be easy to recognize and solve, and some problems are really hard to solve and it takes a while. We’ll go through all of that with AIs.

In the late ’60s, there was more political violence than there is now and it was undergirded by people who actually believed that political violence was the answer. The country seemed to be coming apart. There was no consensus about things. We seem to be back in something like that, where there’s a lot of polarization, demonization, and political violence. Is the political system robust enough to keep things in check or is it just not working anymore? How do we create a consensual government system where we get most of the benefits with fewer of the harsh costs?

This is actually an interesting time to be alive right now, because all of that is up for proof, up for grabs. There’s a lot of grabbing going on. The system was designed to be ungrabbable, but it’s been grabbed. How far down that path of having been grabbed does it go before it becomes stuck there? That’s what we’re in the process of finding out.

There’s lots of reasons to find its way back to some kind of balance. It’ll be different than before. The systems will be different than before. We may have different political parties with different names, but the basic apparatus in the towns, to a large degree in the states, is pretty intact. We’ll find a way to find a new center of balance.

The Whole Earth Catalog‘s statement of purpose says, “We are as gods and might as well get good at it.” The first issue came out in 1968. We’re coming up on 60 years. Do you think we’ve gotten pretty good at it?

The big test will be climate. We’ve certainly gotten better at a whole lot of godlike powers—god with a small g. These are great god things, not the Almighty.

We haven’t really attempted anything in terms of maintenance of the planet at planet scale, at planet pace. With climate change, we’re dealing with a big, deep, slow process. There’s no instant cures, although some are better than others. Geoengineering is one that we’ll come to just because the cost of continually rising temperatures and rising oceans will make it seem like this is the low-cost way to buy enough time to really convert all of our energy systems to basically noncombustion. That will be a different planet, a different society, different global civilization, because it’s not economic. There’s a global economy, but there’s no such thing as a planetary economy. The things that matter at planet scale are not measurable in dollars and cents.

The great thing of the advance of science is that we have lots and lots of capabilities of sensing that something is going wrong and what exactly is going wrong in that thing. In terms of maintenance, the ability to do that kind of sensing is crucial. They call it predictive maintenance. Before the thing breaks, you have indications that would like to break, and that’s when you try to head it off. Often, you can’t. So far, we’ve not fully succeeded in doing that with climate. We’ve gone a long way, much farther into solar than I would ever have thought. We were pushing solar 60 years ago in the Whole Earth Catalog in a big way. Among our crowd, we thought, “This is obvious,” but it took a while for it to become cheap and easy.

If you’re a farmer and you have 100 acres of your farming, you can get a certain amount of food out of it. Even with precision farming, it’s going to be just a certain amount. If you let some company put a whole bunch of solar collectors on your 100 acres and you lease it to them, you get 10x to 100x the money and none of the hassle. The cheap, abundant source of energy is increasingly becoming the sun. And there’s quite a lot of sunlight. There’ll be even more in orbit as these guys are trying to figure out how to start having major data centers in space.

Do you feel like the world is growing up a bit about nuclear energy? You rankled a lot of people in the environmentalist movement when you claimed that nuclear power makes obvious sense if you want to reduce various kinds of emissions and minimize impact on the planet. Do you feel that message is about to be fully accepted around the world?

The opposition got outlived. Basically, they were not able to convince younger and younger generations to buy into what turned out to be a false fear. And then very wealthy young people want to do AI or want to do crypto or whatever; that takes a lot of energy. They look directly at the advantages of nuclear power. They’re not looking at it through the history of nuclear weapons, Hiroshima, Nagasaki, and Three Mile Island and all that. They’re looking at “Can this be made safe?” Yes, totally. “Is it something that can really scale?” Yes. So nuclear scaling up and solar continuing to scale up look like they’re both going to happen.

What are you doing to maintain your legacy? How does one go about maintaining their legacy while they’re still around to do it?

I tried to write a memoir at one point, and as soon as I wrote a line of it, I hated it. I was just bored with myself. But I didn’t have to because some guys came along and wanted to make a film about me and this documentary called We Are As Gods was made, and it’s good. John Markoff came along and wanted to write a biography, and he did, and it’s good. All of the Whole Earthwork we did over 30-plus years is now online at wholeearth.info. My legacy has gone ahead and somehow established itself. It’s not something I’m concerned about. I’m one of the really very lucky people in that respect.

After The Population Bomb author Paul Ehrlich died, you wrote, “I’ll speak up for Paul. He was wrong about discounting the ‘demographic transition’ in human population, but he co-authored (with Peter Raven) one of the most cited papers in biology, on co-evolution.” Given his large role in stressing people about overpopulation and his influence among governments in reducing births, how do you assess his overall contribution to science and society?

Remember, it was Dave Brower at Sierra Club who asked Paul to write that book and to write it not as science but as a polemic. Overpopulation was an environmentalist obsession before Paul lit the fuse on his book. (These were the same self-named “ecologists” who couldn’t tell a trophic level from a Tyrannosaurus.)

Peter Raven is a botanist (whose Wikipedia bio says, “Raven is possibly best known for his work ‘Butterflies and Plants: A Study in Coevolution,’ published in the journal Evolution in 1964, which he coauthored with Paul R. Ehrlich”). Paul was a zoologist, a lepidopterist focused on checkerspot butterfly populations. He noticed that supposedly identical butterflies dined on completely different plants in different regions. Over coffee, he and Raven complained to each other that zoologists treated plants as just so much edible plastic, whereas in evolutionary reality, plants pay just as much active attention to animals as animals do to them. Naming that attention “coevolution” was a thunderclap in evolution theory, because it forced biologists to notice that most of evolution is in fact coevolution—living things devote most of their adaptation to dealing with other living things, who are busy adapting right back at them.

It’s a profound idea that reframes everything. That’s why I named a magazine for it—CoEvolution Quarterly. For me, it far outweighs Paul’s exaggerations about human population numbers.

In a sense, I’m attempting something similar with “maintenance.” It’s not just a persistent nuisance. It’s so essential that it’s what most living things have to spend most of their time and attention on tending to.

This interview has been condensed and edited for style and clarity, as well as augmented by questions answered over email.

The post Stewart Brand on Fixing Stuff, Modern Environmentalism, and the Nuclear Future appeared first on Reason.com.

“ChatGPT advised the FSU shooter that a mass shooting would get more attention from media if it involved several children,” NBC deputy tech editor Ben Goggin posted on X yesterday.

“Advised” is a funny way to put it, implying that the artificial intelligence system recommended this course of action or helped the shooter—then-20-year-old Phoenix Ikner—plot details of how he would carry out his attack. In fact, ChatGPT seems to have provided neutral information in response to questions that were not obviously asked with murderous intent.

That attack, which took place in April 2025 at Florida State University, left two people dead, including Tiru Chabba. Chabba’s widow, Vandana Joshi, is now suing ChatGPT maker OpenAI in federal court, alleging negligence, battery, defective design, failure to warn, and wrongful death.

After chatting with the shooter, ChatGPT “either defectively failed to connect the dots or else was never properly designed to recognize the threat,” the suit alleges. OpenAI “failed to create a product that would refrain from participating in discussions that amounted to it co-conspiring with Ikner” and “failed to create a product that would appropriately alert a human that investigation by law enforcement may be necessary to prevent a specific plan for imminent harm to the public.”

But treating the conversations between ChatGPT and Ikner as grounds for legal liability is misguided, no matter how understandable it may be that the victims’ loved ones would want to assign blame here.

In this case, ChatGPT allegedly provided Ikner with information on basic features of certain guns, on what times the FSU student union was crowded, and on what sorts of mass shootings received attention.

Knowing what Ikner eventually did, it may be easy to view this as damning. But asking about what times a campus is crowded is not at all weird in itself. Asking how a gun works could be simple curiosity, or related to hunting or self-protection. And researching the common features of prominent mass shootings is something one might do for all sorts of harmless reasons—academic research, media criticism, or gun violence prevention efforts, to name a few. ChatGPT providing neutral information on the kinds of shootings that receive attention does not amount to (as the suit alleges) “advice” or “recommendations.”

And just because Ikner asked about all three things does not mean he did so simultaneously, in one session, in a way that might trigger alarms. It’s possible for people to use AI tools in ways which would make “connect[ing] the dots” between any dispersed conversations difficult.

It wasn’t as if Ikner talked with ChatGPT about nothing but mass shootings. Joshi’s complaint alleges that ChatGPT “helped him with his homework and his work-out routines, gave him tips on getting girls and relationship advice, and suggested to him how to dress and style his hair.” They chatted about everything from loneliness and being bullied to video games, Nazis, Christian nationalism, Donald Trump, and mental health.

It also allegedly advised him to seek help. “Ikner described his depression to ChatGPT, who confirmed some of his symptoms and advised him to seek out a therapist,” states the lawsuit. When Ikner asked about suicide, ChatGPT provided “information of effects of suicide on others and twice directed him to a suicide prevention hotline.

Joshi’s complaint suggests the suicide talk in conjunction with other chats—including ones in which Ikner asked about the assassination attempts on Trump and one in which he asked about the aftermath of shootings—constituted a big red flag. Again, we don’t even know that ChatGPT had historical memory of any of these supposed red-flag conversations by the time another one came up. But even if it did, it’s unclear why these queries should have raised alarms. Most people who contemplate suicide don’t become mass shooters. It’s natural for people to want information about assassination attempts on the president. And a question about what would happen after a mass shooting at FSU could easily be something that someone afraid of school shootings would wonder.

“ChatGPT provided factual responses to questions with information that could be found broadly across public sources on the internet,” OpenAI spokesperson Drew Pusateri, told NBC, “and it did not encourage or promote illegal or harmful activity.”

It’s important in emotionally charged situations like these to think about the alternatives—alternatives to Joshi getting this information from ChatGPT and alternatives to the way ChatGPT and OpenAI handled things.

It seems silly to imagine that if Ikner had not got any of the objected-to information from ChatGPT, he wouldn’t have been able to carry out his planned shooting. All of the information he gleaned could have been obtained easily from a basic internet search or other sources.

ChatGPT could be trained to refuse to answer questions about certain topics, including guns or the history of mass shootings. But this could limit its general usefulness and prevent it from providing information to people seeking it for neutral or even beneficial reasons—and for what ultimate purpose? A motivated criminal isn’t going to give up just because ChatGPT won’t answer his question.

OpenAI could be more aggressive in reporting people to authorities over their chat topics. But this seems unlikely to go well for anyone. It would almost certainly make people more wary of using ChatGPT. AI detractors may imagine that as a good thing—until people start turning to other AI tools, including those outside the United States and unsympathetic to any U.S. law enforcement requests.

And authorities would be overwhelmed by useless reports. Following up on all of these could take time away from more important pursuits. It could also lead to all sorts of negative encounters between innocent individuals and police, putting people’s civil liberties and even their lives at risk.

If tech companies are potentially on the hook for murder because their AI products chatted with a murderer, we can expect to see them reporting anyone who asks about mental health, guns, historical violence, and much more. This would inevitably draw a lot of innocent people into encounters with police, child welfare agencies, and other authorities.

Each new entertainment and communications tool gets its turn being blamed—in the public imagination and in court—for people’s bad acts. Before AI, we saw people blame social media; before social media, we saw people blame video games; before video games we saw people blame violent TV and movies, and so on.

People want some simple answer to horrible events—just ban violent video games, or put ratings on TV shows, or make AI companies file more police reports. But expecting AI companies to stop shootings won’t lead to fewer shootings. It’s just going to create new problems.

IN THE NEWS 

Texas app store act blocked: “A federal judge in Austin has once again blocked a state law from taking effect that would regulate minors’ access to content on Google Play and Apple’s App Store,” notes the Austin American-Statesman: 

Judge Robert L. Pitman previously blocked the App Store Accountability Act from taking effect on Jan. 1 by issuing a preliminary injunction while the law’s constitutionality is considered in court. He declined to lift that injunction Wednesday afternoon.

SB 2420, signed into law by Gov. Greg Abbott in 2025, would require app stores to ensure users are over 18 or obtain parental consent before allowing them to download or purchase an app.

Texas Attorney General Ken Paxton wanted Judge Pitman to permit enforcement of the law as the case played out. But Pitman has said the law raises serious concerns for free speech.

On Substack 

“Instead of fracturing our shared reality, this handful of AIs seems to be piecing it back together,” writes Jerusalem Demsas at The Argument. She argues that artificial intelligence is a centralizing rather than decentralizing technology.

Public conversation tends to treat chatbots as the next in a long line of digital communications technologies that have decentralized truth.

The internet, smartphones, and social media all made the production of information cheap and significantly decentralized who could produce it. AI is making the production of information extremely expensive and centralizing who can produce it.

And while, yes, AI hallucinates, the direction of its errors is toward mainstream consensus, not fringe positions. When ChatGPT gets something wrong, it tends to do so in a confused-Wikipedia-editor-misremembering-something-they-once-read kind of way, not in a QAnon-forum-poster-high-on-ketamine kind of way.

The open question is who will get to control the centralizing forces of AI.

Read This Thread 

There are so many insane wildly misleading stories coming out about data centers almost every day now that I’m mostly having to give up on commenting on them to focus on actually getting blog posts out, but it feels like a tsunami. I’ll share one from just today as an example.

— Andy Masley (@AndyMasley) May 10, 2026

More Sex & Tech 

• Prostitution has “been called the oldest profession, and it seems like if there is a willing seller and a willing buyer between adults, the government has no business getting involved,” Rhode Island state Rep. Edith Ajello (D-Providence) told The Providence Journal. Ajello is the lead sponsor of House Bill No. 8057, which would decriminalize prostitution in the state. In April, the legislature held the measure for further study.

• A Foundation for Individual Rights and Expression poll conducted in April 2026 found that only 26 percent of respondents trust the federal government to oversee social media use for minors. But most people—69 percent—said they trusted parents to do so.

• Lawmakers in Portland, Oregon, want to make it easier to crack down on hotels where prostitution takes place. But “shutting down a venue doesn’t make [sex work] go away,” Emi Koyama, founder of Coalition for Rights & Safety for People in the Sex Trade, told Filter. “It displaces people to other areas, and it becomes more dangerous.”

• “Adult site Pornhub will now allow users in the U.K. to confirm their age using Apple’s verification system, introduced in iOS 26.4,” reports Forbes. Pornhub’s parent company, Aylo, has resisted conducting its own ID checks to verify ages but “announced on May 5, 2026 that Apple’s method—the world’s first operating-system-level age check—meets their rigorous privacy standards.”

ª Chris Ferguson on a new study of cell phone bans in schools: “at least on the surface, this study is very bad news, indeed for cellphone ban fans. It supports the narrative that they are largely ineffective. There are some reasonable criticisms of the study though.”

The post Don't Blame ChatGPT for the Florida State Shooting appeared first on Reason.com.

Self-checkout machines are in the crosshairs. In recent months, numerous states and localities have considered legislation to curtail the use of automated checkout in grocery stores. These bills are often positioned as part of an effort to cut down on retail theft, but it appears the driving force behind them is to create more unionized jobs.

According to USA Today, at least six states have considered rules that would restrict self-checkout machines. The states range from blue Connecticut to red Ohio, but it doesn’t stop there. Two cities in California already have self-checkout limits in place, while New York City is currently considering restrictions as well.

Self-checkout restrictions are often framed as a commonsense crime prevention measure that protects grocery store workers and cuts back against the recent uptick in retail theft nationwide. But when it comes to these bills, the fine print points toward a different motivation.

For instance, the Connecticut bill mandates that stores must have one employee for every two self-checkout machines, in addition to having one manual checkout station for every two automated lanes. Stores cannot go over eight self-checkout lanes total. And any employee designated with the task of supervising self-checkouts is barred from engaging in any other simultaneous duties that could interfere with such supervision.

The various bills percolating in other state capitals and city halls are all largely structured the same. A previous iteration of California’s self-checkout bill specified that any store seeking to implement technology that “significantly affects the essential job functions of its employees” or “eliminates jobs or functions” must conduct an “impact assessment” before doing so, underscoring the real impetus there.

In other words, in the name of reducing theft, these rules would functionally operate to increase the number of clerks that each store must employ at any given time. Given that the grocery industry has historically had higher unionization rates than other retail sectors, this would translate into more unionized jobs.

If any doubt remains, one need only look at the biggest supporters of these bills. In Connecticut, all the legislative testimony submitted in favor of restricting self-checkout came from labor unions, including representatives affiliated with AFL-CIO, Service Employees International Union (SEIU), and the United Food and Commercial Workers (UFCW), which is the largest grocery worker union in the U.S. A CalMatters summary of sponsors and opponents for California’s self-checkout bill likewise shows that the majority of the bill’s boosters are labor unions.

News articles published about these bills also frequently quote UFCW reps touting their virtues. The effort to restrict self-checkout dates back to at least 2019, when unions in Oregon pushed a state ballot measure that would have limited groceries to two self-checkout lanes per store.

To be sure, there is evidence that self-checkout machines can result in higher shoplifting rates. One frequently-cited study found that so-called inventory shrink at grocery stores was 16 times more likely with self-checkout than with traditional cashiers. A LendingTree survey reported that 27 percent of self-checkout users admit to intentionally stealing items in self-checkout lanes, with another 36 percent saying they took items inadvertently.

But unsurprisingly, stores themselves—who have a direct bottom-line incentive to prevent shoplifting—have proven more than capable of responding. Walmart and Target have garnered headlines for dropping or limiting self-checkout at various stores around the country, while Five Below and Dollar General have also curtailed automated checkout in recent years. Technology also offers promise, with various groceries now onboarding smart video and AI to crack down on shoplifting in the self-checkout lane.

In the end, no one wants to clamp down on retail theft more than grocery stores themselves. If that means cutting back on self-checkout, they will do so. But they don’t need the government, aided by unions, to decide for them—and to forcibly inflate their payrolls at the same time.

The post Self-Checkout Is Under Fire Across the Country. Is Theft Really the Reason? appeared first on Reason.com.

The Australian government, which has already imposed strict regulations on American tech firms operating in the country, now expects these companies to pay taxes to support Australian journalism.

On Tuesday, Australia unveiled draft legislation for a “News Bargaining Incentive,” which would require major tech companies, including Meta, Google, and TikTok, to make commercial deals with news organizations or face a 2.25 percent tax on local revenue, reports The Wall Street Journal. Companies would be incentivized to comply by receiving offsets of either 150 or 170 percent, effectively reducing the tax. The legislation would not apply to AI companies.

Prime Minister Anthony Albanese told reporters that the bargaining incentive would bring in an expected 200 to 250 million Australian dollars, “every single dollar” of which “will go back to journalists.”

Australia’s communications minister, Anika Wells, pitches this as a way to fix the country’s old News Media Bargaining Code, which took effect in 2021. Like the legislation introduced this week, that code pressured designated tech companies to pay journalistic outlets for news. Google and Meta initially entered into agreements with news outlets. But when Meta’s contract expired in 2024, the company refused to renew, arguing that just 3 percent of its content was news-related.

A Meta spokesperson criticized Australia’s most recent proposal as a “digital services tax,” writing on X: “News organizations opt to post content on our platforms because they get value from it. We don’t take their news content. Yet the tax applies whether or not news content appears on our platforms.”

Google is also pushing back against the tax, explaining in a statement that the proposal “ignores the fact that Google already has commercial agreements with the news industry, misunderstands how the ad market changed and mandates payments from some companies while arbitrarily excluding platforms like Microsoft, Snapchat and OpenAI—despite the major shift in how people consume news.”

The Australian government has a history of meddling in its country’s information environment. In December, it prohibited people under the age of 16 from using social media platforms. Enforcement has been rocky, and many young Australians have successfully evaded the ban. Instead of accepting that kids are savvy enough to evade restrictions, the Australian government threatened to sue Facebook, Instagram, Snapchat, TikTok, and YouTube for noncompliance.

The mission to save local journalism may appear well-meaning. Who doesn’t want to support original reporting? But the new scheme would not just benefit scrappy reporters at small-town papers doing shoe-leather journalism. While it offers incentives for tech companies to strike deals with smaller organizations, the companies could still reduce their tax burden by making deals with larger operations. Australia’s major news organizations, including News Corp Australia, the Australian Broadcasting Corporation, and Nine Entertainment Co., have been vocal supporters of the code and would likely be its major beneficiaries.

And even if the new code gives a boost to struggling newsrooms, that wouldn’t address the journalism industry’s underlying problems. It would impose a system where Australian newsrooms rely on another country’s tech industry for survival. The more durable, albeit challenging, path forward for newsrooms is to reach audiences and secure funding without a government middleman.

The post A Journalism Tax Is a New Front in Australia's War on American Tech appeared first on Reason.com.

Seven years ago, police in Midlothian, Virginia, sought to identify a bank robber by asking Google to search the records of more than 500 million people who used the company’s “location history” feature. That search identified 19 devices that were in or near the bank around the time of the robbery, which police winnowed down to three people, including Okello Chatrie, the man who was ultimately convicted of the crime.

Depending on your perspective, that use of a “geofence” warrant was either an unobjectionable example of smart police work or an outrageous invasion of privacy. On Monday, the Supreme Court weighed the merits of those contending views in a case that illustrates the threat that two dubious doctrines pose to Fourth Amendment rights now that Americans routinely entrust huge volumes of personal information to tech companies that help them with myriad quotidian tasks.

In 1967, the Supreme Court said the Fourth Amendment applies only when you have a “reasonable expectation of privacy.” In subsequent cases that involved bank and telephone records, the Court said you have no such expectation when you voluntarily share information with third parties for their own business use.

The problems with those principles were clear in a 2018 case that addressed the FBI’s tracking of a robbery suspect via location data collected by cellphone sites. Although the majority held that such tracking generally requires a warrant, that conclusion was hard to reconcile with the third-party rule, as Justice Neil Gorsuch noted in his dissent.

Chatrie’s case is broadly similar, except that the information used to identify him was collected by his own phone rather than cell sites. Although Google no longer keeps such data on its servers, many commonly used apps track the whereabouts of their users, and that information is often stored remotely.

In Chatrie’s case, police obtained a warrant. But his lawyers argue that such orders sweep so broadly that they violate the Fourth Amendment, which says a warrant must be based on probable cause and must “particularly” describe “the place to be searched” and “the persons or things to be seized.”

The Trump administration wants the Supreme Court to reject that argument. But it also maintains that geofence inquiries do not require any sort of warrant because people who consent to location tracking have no reasonable expectation that the resulting information will be kept private.

If so, Chatrie’s lawyer, Adam Unikowsky, warned during oral argument on Monday, the same logic would allow the government to search people’s remotely stored emails, photos, calendars, and documents at will. Several justices understandably viewed that prospect with alarm.

Although Deputy Solicitor General Eric Feigin assured the justices that his argument did not extend that far, those other examples likewise involve voluntary sharing of information with third parties, which supposedly makes the Fourth Amendment inapplicable. And even when the information is limited to location tracking, it can still reveal sensitive details of people’s lives.

“The potential for abuse is breathtaking,” Unikowsky notes in a Supreme Court brief. “The government need only draw a geofence around a church, a political rally, or a gun shop, and it can compel a search of every user’s records to learn who was there.”

In addition to arguing that people reasonably view their location histories as private, Unikowsky takes a page from Gorsuch, who in 2018 questioned both that test and the third-party doctrine. Since the Fourth Amendment protects people’s “papers” and “effects” against “unreasonable searches and seizures,” Gorsuch suggested, the relevant question is whether the information sought by the government falls into those categories.

Gorsuch thought it was plausible to argue that people retain a property interest in their records even when they share them with others for specific purposes. That argument, which Unikowsky adopts, seems like a promising way to escape the danger posed by the third-party doctrine, which modern technology has magnified to an extent that would have dismayed the Fourth Amendment’s framers.

© Copyright 2026 by Creators Syndicate Inc.

The post A SCOTUS Case Exposes the Dangers of 2 Misguided Fourth Amendment Doctrines appeared first on Reason.com.

Even before artificial intelligence was a meaningful force in the economy, technologists, politicians, and policy wonks of all political persuasions have endorsed a universal basic income to cope with the mass unemployment that will be caused by the AI revolution.

The familiar case is that an AI-powered economy will be able to automate most economic production, making the economy as a whole much richer, but leaving the average person jobless and destitute. The solution is then to redistribute some of the gains from AI to the public by sending everyone, regardless of income, a check.

Businessman Elon Musk has gone one step further by calling for a “universal HIGH INCOME” to pay for the AI-induced unemployment, which he suggested would be inflation-free thanks to the downward pressure AI will put on prices.

Universal HIGH INCOME via checks issued by the Federal government is the best way to deal with unemployment caused by AI.

AI/robotics will produce goods & services far in excess of the increase in the money supply, so there will not be inflation.

— Elon Musk (@elonmusk) April 17, 2026

Musk is almost certainly right that AI will put downward pressure on prices, as one would expect of any productivity-enhancing technology.

He’s mistaken in believing that this makes a universal income (regardless of whether it’s basic or “HIGH”) a wise policy.

Even in a future in which AI does revolutionize the economy, we will not see technologically driven mass unemployment. In fact, a universal basic income would likely result in more of the joblessness it’s meant to mitigate.

To the first point, the industrial revolution has been outsourcing more and more tasks to labor-saving machines for roughly 300 years now. While this ongoing process has certainly made lots of individual jobs obsolete, it has not made jobs generally obsolete.

Excepting the monthly ups and downs of the unemployment rate, the total number of jobs in the economy continues to rise precipitously in the long run.

If labor-saving technology destroyed the need for labor, we should have fewer jobs today than ever before. We don’t. Even as farms and factories employ fewer people, we keep finding ways to keep ourselves busy.

The AI boosters and doomers argue that this time will be different, because unlike spinning jennies, combine harvesters, and email, AI will eventually be smarter than humans at everything. When there’s nothing that flesh-and-blood humans can do better than machines, we’ll end up doing nothing at all.

These arguments are obviously speculative because we don’t have artificial general intelligence yet. Even when we do, it’s reasonable to assume that humans will continue to have employable comparative advantages, if only because humans prize human interaction.

There are lots of jobs today that could be automated but aren’t. Plenty of people work in offices even if their tasks could be completed remotely. So long as people are social creatures, I can only assume we’ll find something marketable to do with our time.

Outside of speculative future scenarios, here in the real world, the economic impact of AI continues to look similar to the impacts of past productivity-enhancing technological innovations. That’s true even in industries that have been most influenced by AI.

Language translation is something that AI has long been pretty good at, and language translation services have become increasingly automated over time.

When journalist Timothy Lee looked at the impact of AI on the industry in 2023, he found that the technology had caused prices for translation to fall, and more consumption of translation services. Translators themselves were adapting by either specializing in translation of legal or medical texts (which still requires human oversight), using AI to increase their productivity, or dropping out of the industry.

The effects of AI on translators weren’t all positive. But that basic story of falling prices, rising productivity, some jobs disappearing, and others becoming more specialized sounds a lot like every industry revolutionized by technology.

The evidence that AI will finally be the technology that puts everyone out of work just isn’t there.

Economic transitions don’t happen automatically. It will take time for people to find new jobs as AI destroys the old ones.

That’s precisely why a universal basic income or (“HIGH INCOME”) would be so dangerous to adopt.

A pretty robust finding in the research is that giving people unconditional cash grants leads them to work less and even stop working at all if the benefits are generous enough.

Pairing advancing AI with a universal basic income would give people a major incentive not to work, right as many existing jobs are being automated away. Instead of people finding their next comparative advantage in an economy being made more productive but also automated by AI, many would probably just stay home instead.

Far from mitigating the employment effects of AI, a universal income would seem to usher in the jobless dystopia that those convinced of AI’s transformative effects are worried about.

We should have a little faith in humans and technology. For centuries, technological progress has made us richer while creating more jobs. The only way AI will be different is if we use its productivity gains to pay people not to work.

The post Elon Musk's Mistaken Call for a 'Universal High Income' appeared first on Reason.com.

During a tempestuous moment in the ongoing trade war between the world’s two biggest economies, the Chinese government last October threatened to exercise its most potent economic leverage by ending exports of rare-earth minerals.

The bold move got results. The Trump administration, which had earlier threatened tariffs as high as 130 percent on Chinese goods, agreed to cut tariffs and ease tech export restrictions from the U.S. to China. In exchange, China said it would hold off on limiting rare-earth exports for at least a year.

The incident—and the ongoing threat embedded in its resolution—highlighted China’s strategic control over rare earths, the collective name for 17 different heavy metals essential to many aspects of modern computing technology. China mines 60 percent of the global supply of rare earths and processes 80 percent.

In an attempt to break China’s hold on the global market for this vital resource, the Trump administration is now spending millions to spur the development of new rare-earth processing facilities in Texas and California, and is also seeking partnerships with Australian mines.

That response flows from the widespread belief that Chinese dominance of rare-earth minerals is the result of a market failure.

“The market fundamentalists argue that government should not be picking which industries to support,” Marco Rubio, now the secretary of state, explained in a 2019 speech. “But what happens when an industry is critical to our national interest, yet the market determines it is more efficient for China to dominate it? The best example of this is rare-earth minerals.”

That view ignores the government’s own role in sabotaging America’s production of rare-earth metals—a market that the United States dominated until the 1980s.

Permitting is a major problem. It takes seven to 10 years for a new mine to obtain the necessary permits in the United States, compared to an average of two years in Canada and Australia, according to the Essential Minerals Association (EMA), an industry group. An S&P Global study published in 2024 found that it took American mines an average of 29 years to go from discovery to production—the
second-longest period in the world behind Zambia.

Those delays often stem from the fact that a single mine requires approval from multiple federal and state agencies with overlapping and duplicative regulatory requirements. The EMA estimates that permitting delays add more than $1 billion to the development of major mining projects.

As a result, much of America’s abundant supply of rare earths—which are actually not all that rare, despite the name—remains untapped. Round Top Mountain, in West Texas, contains the largest deposit of rare earths in the U.S., but a mine there has been in development since the 1980s and is now aiming to be open by 2028. Other projects in California and Utah are similarly tied up in permitting delays.

Breaking China’s hold on the rare-earth supply chain doesn’t require the federal government to pick winners and losers. Instead, it should just get out of the way.

The post Blame U.S. Regulations for China's Dominance in Rare-Earth Minerals appeared first on Reason.com.

As New Jersey faces a major impending budget deficit in the next few years, Democratic Gov. Mikie Sherrill is asking lawmakers to fund a crusade against social media companies in the name of “online safety.” 

During Sherrill’s first budget speech on Tuesday, in which she said her state is facing a “$3 billion structural deficit,” the governor warned of the ills of technology on kids. Big Tech “isn’t just the Big Tobacco of our era,” Sherrill said, “it’s worse, and it’s exactly the kind of situation where government has a role to play to keep our kids safe.” 

To address this perceived problem, Sherrill is proposing setting aside $125,000 in the state budget for a new Office of Youth Online Mental Health Safety within the Health Department. This office would “research and make recommendations to guide responsible use of social media platforms among youth” in “preparation for New Jersey’s first cell phone-free school year this fall.” She also proposed spending $500,000 on a new Social Media Research Center to “study the impact of digital technology on young people’s mental health.”

It is unclear whether such a research center would be guided by genuine scientific inquiry or by anti-tech ideology. Sherrill already seems convinced that digital technology negatively affects youth mental health. During her speech, she declared, “In New Jersey, we’re not going to rely on Big Tech to come clean about the harm these technologies cause. We are going to lead the way.” 

While a growing number of state and federal lawmakers are asserting that social media undisputedly harms teens’ mental health, not all research supports this narrative, as Reason‘s Elizabeth Nolan Brown has frequently pointed out. One 2022 Pew Research Center study presented a more “nuanced picture of adolescent life on social media,” finding that most teens say social media can strengthen their friendships, while still acknowledging that there are other pressures that may come from being online. Only 9 percent of the teens surveyed said that social media has had a “mostly negative” effect on them personally, with the majority saying the effect has been neither positive nor negative, and 32 percent of those surveyed described their experience as mostly positive. This is not exactly the picture of doom and gloom painted by Sherrill. 

Even assuming that social media causes harm, government interventions have often led to infringements on online freedom in the name of protecting kids. The U.K.’s Online Safety Act, which requires platforms hosting any sort of material deemed harmful to children to verify users’ ages, has effectively censored swaths of the internet for those unwilling to give the government sensitive data. In complying with the act, social media sites in the U.K. have restricted posts about Ukraine and Gaza, Substack posts, and Reddit forums. The U.K. has even attempted to impose its country’s online safety rules on U.S. companies, so far with little success. Alarmingly, U.S. lawmakers are eagerly trying to implement these policies on their own constituents. Several states, including Utah and Texas, have attempted to enact similar age-verification laws that would restrict access to online speech, prompting legal challenges from free speech groups. 

Although Sherrill has not yet proposed age-verification laws, she signed an executive order on her first day directing state agencies to prioritize promoting online safety. The order calls for preventing “harms such as cyberbullying, deepfakes, online exploitation, and exposure to harmful or addictive content.” 

New Jersey legislators seem ready to back Sherrill’s campaign against Big Tech. In February, a trio of lawmakers introduced a bill that would require certain social media platforms to display a “black box warning” detailing the dangers of social media use whenever a user accesses a platform.  These sites would also be required to monitor user activity “for problematic behaviors,” as determined by the commissioner of health, and would have to give that user resources related to their “problematic behavior.” 

Faced with a budget squeeze, New Jersey lawmakers are gearing up to splurge on government-run crackdowns on social media platforms. In addition to wasting taxpayer dollars, these projects could infringe on New Jerseyans’ rights. 

The post Facing a Budget Squeeze, New Jersey Decides To Go After Big Tech appeared first on Reason.com.

Sen. Bernie Sanders has a problem with data centers. They’re just too good.

In a video posted to social media in December 2025, the Vermont independent complained that billionaire tech moguls are reaping huge profits from their data center investments while the technological innovations these facilities power will automate away countless jobs currently done by human workers. He called for a federal moratorium on data center construction to “give democracy a chance to catch up with the transformative changes that we are witnessing.”

By 2030, companies are projected to invest as much as $7 trillion on building or updating the boxy, server-filled facilities that keep the digital cloud aloft and train large language models to speak intelligibly. About 40 percent of that spending will happen in America.

Sanders is hardly alone in complaining about America’s staggering data center boom.

As 770 data centers enter the development pipeline, a chorus of usual suspects is demanding they be stopped. Environmentalists complain that they are gobbling up virgin land and drinking all of the water. Preservationists say these boxy facilities are ruining the character of local areas. Consumer advocates say their power demands are raising everyone’s electricity bills.

In Sanders’ case, his complaints about data centers tacitly accept the premises of the people investing huge sums in them: that these facilities will be fabulously profitable investments that spur the development of the innovative, labor-saving technologies of the future. But the socialist senator thinks that’s a bad thing. After all, no government bureaucrat has precisely planned where all this economic dynamism will take us.

The rest of us should be able to see the tremendous upsides of the country’s data center boom. Advances in artificial intelligence and robotics could liberate humanity from boring, backbreaking labor. The early profits of data center development are a leading indicator of the increasingly productive economy that awaits us in the years to come.

While the rest of the economy wobbles under the weight of trade wars and other forms of government intervention, data center investment is almost exclusively responsible for driving GDP growth.

Contrary to the claims of the not-in-my-backyard technophobes, all this growth comes with minimal environmental downsides. Data centers consume a tiny portion of the nation’s water. While they’re not the prettiest buildings to look at, they mean less noise, fumes, and traffic than almost any other land use one could care to name.

Their power consumption is gargantuan. But data centers’ electricity demands are also driving secondary innovations in the world of energy efficiency and power generation.

In short, it’s hard to imagine an industry that gets more juice from the olives it squeezes. Data centers are the silent, uncomplaining Atlas holding our dynamic, tech-driven economy on his mighty shoulders. They don’t produce rainbows (yet). But they might as well. We should all stop worrying and learn to love them.

Building Boom and Backlash

The internet isn’t a series of tubes. It’s a series of warehouses filled with computing equipment in Loudoun County, Virginia.

The suburban jurisdiction just outside the Washington Beltway is home to 200 data center facilities with a physical footprint of 47 million square feet. The wider Northern Virginia region hosts some 13 percent of the entire world’s data center capacity, supporting a huge percentage of global internet activity. When Amazon Web Services’ data centers in the area experienced a brief disruption in late 2025, everything from Venmo to the British government’s tax services went down.

It’s fashionable to complain that America is a “build nothing” country. Loudoun County shows how much matter we can add to meatspace when the law allows it.

In 2000, the county’s zoning administrator decided to treat data centers as though they were office buildings, meaning they could be built by-right (i.e., without the need for approval from elected officials) in large swaths of the county. That decision made a lot of sense. Data centers are, after all, effectively just office buildings with more computers and fewer windows.

The subsequent boom in data center construction beggars belief.

As an October 2025 county report states, “The rate of data center growth in Loudoun County over the past 20 years has been among the highest of any community in the world, and that rate has accelerated exponentially since 2022.” From 2016 through 2020, data centers’ square footage in Loudoun more than doubled. From 2020 to 2025, it more than doubled again to today’s nearly 50 million square feet. Another 117 new data centers are currently planned to be built in the county.

Predictably, the ground zero of America’s data center boom is now the epicenter of the backlash to them. In March 2025, the Loudoun County Board of Supervisors voted to revoke data centers’ by-right status in the zoning code. Going forward, new data centers will need individual approval from the Board of Supervisors. (The new rules grandfather in many of the pending data center proposals filed under the old rules.)

It’s understandable why people would rather live next to an open field or a single-family home than a data center. Yet the voluminous press coverage of property owners’ complaints about these facilities reveals just how low-impact such centers are, particularly in the context of their overall economic importance.

Neighbors complain that data centers are unsightly. Their white concrete walls produce a glare. The low, constant hum of their generators and cooling equipment, while well within existing noise limits, can annoy human beings and scare away birds. (The hum is anywhere between 40 and 59 decibels, which is a little louder than a library and a little quieter than a pickleball court.)

Anti-growth activists often complain that new residential development generates traffic, strains local schools, and sucks up resources from local budgets. Data centers do not. McKinsey & Company notes in a research brief that large data centers have around 50 permanent workers. These relatively small staffs mean their impacts on roads and schools are minimal.

“My standing joke about data centers is they’re a very big building with a very small parking lot,” Shane Greenstein, a Harvard professor who studies data centers, told Inside Climate News last year.

Despite the targeted local tax breaks that data centers received, they pay for about 30 percent of Loudoun’s local budget. Perhaps there’s a case for regulations to mitigate some of the centers’ externalities, but they are not smoke-belching factories. Yet data center critics are increasingly winning the local land-use battles. Heatmap News reports that 25 data centers were scrubbed nationwide in response to local opposition in 2025, four times as many as were canceled in 2024. That’s evidence of both the demand for more data centers and the regulatory obstacles they’re up against.

As the neighbors come out against data centers, people with more global concerns are trying to block them as well.

Power Thirsty

If you were simply reading the headlines, you’d assume that data centers were drinking up most of America’s water and polluting whatever they leave behind. That couldn’t be further from the truth. Individual facilities are, all things considered, modest drinkers.

“Most data centers use about the same amount of water or less than an average large office building, although a few require substantially more, and some require less than a typical household,” notes a research report prepared by the Virginia General Assembly’s Joint Legislative Audit and Review Commission. These facilities consume water in two ways: directly, as a means of cooling their equipment, and indirectly, via the water used in generating the power they consume.

Neither makes much of a dent in America’s fresh water supply. Data centers’ water consumption is a tiny portion of overall U.S. water usage.

Data centers directly consumed 46 million gallons a day in 2023, according to a report produced by researchers at Lawrence Berkeley National Laboratory. Their indirect consumption from the power plants serving them added another 576 million gallons of water consumed daily, for a combined daily total of 625 million.

That’s a lot of water. It’s also a very small percentage (0.2 percent) of the 322 billion gallons of water that the United States consumes each day.

California’s almond industry consumes almost 4.2 billion gallons of water a day, according to the California Water Impact Network. The country’s golf courses consume another 1.4 billion gallons a day.

People can, and do, debate the merits of the amount of water that goes to almond milk and putting greens. But neither golf nor almonds were responsible for effectively all of U.S. economic growth this past year.

And the Lawrence Berkeley lab’s estimate of data centers using 625 million gallons a day might be an overshot. That figure includes water lost to evaporation from the reservoirs that store water used by hydroelectric power plants, as the engineer Brian Potter has noted in his Construction Physics newsletter. Potter argues that while dams and reservoirs lead to additional evaporation, they also make additional fresh water available for use in drier seasons. That evaporation will also occur regardless of whether reservoir water is used to generate power.

If you exclude that from data centers’ consumption, their combined water use falls to a lowly 250 million gallons a day. Despite the panic about the industry’s water usage, that’s really just a drop in the bucket.

Shocking Consumption

Data centers do use lots of electricity. Existing facilities consume 4 percent of the nation’s electricity, and U.S. government estimates predict their share of power usage will hit as high as 12 percent by 2028.

This creates a challenge both for data center operators themselves and for the utility companies that supply them. It has also invited a lot of off-base and bad-faith criticism from data center haters.

Senate Democrats, led by Elizabeth Warren (D–Mass.), say data centers are raising everyday consumers’ electricity bills. There is some truth to that: Higher demand and stagnant supply will, all else being equal, lead to higher prices. Yet all else isn’t equal. Where they’re allowed, data center owners are bankrolling the expansion of electrical generation capacity, which often benefits consumers through lower prices.

“States with higher electricity demand growth generally experienced smaller increases in retail prices,” the energy industry analyst Marc Oestreich recently noted in Reason. “In some cases, prices declined outright.”

In fact, data centers are ideal customers for utility grids because their power consumption is predictable and steady. Their demand doesn’t surge on hot days, as when everyone in an apartment building turns on their air conditioners. This is why utility companies with power to spare often actively court data centers to move into the areas they serve.

More often than not, the obstacles to supplying data centers’ power demands are political, not economic.

Voluminous state and federal regulations can make it difficult to build new power plants, sources of renewable energy, and transmission lines. A key component of the much-in-vogue “abundance agenda” is making it easier to do these things.

If Warren is worried about data centers driving up electricity prices, she should tell the Sierra Club, Earthjustice, and all the other environmental groups she normally agrees with to quit filing lawsuits that attempt to stop new power generation capacity.

Meanwhile, data center companies’ demand is also pushing out the innovative frontier of power generation.

In the past few years, startups like Last Energy have attempted to revolutionize the nuclear power industry by designing and building modular nuclear reactors that can serve as a dedicated, emissionless source of off-grid power for data centers.

Another company, Deep Fission, plans on installing small nuclear reactors at the bottom of deep, narrow tunnels, where the free bedrock can serve the same function as expensive surface containment buildings.

These efforts are in their early days. A major barrier to fast rollout is—surprise—federal regulation. Licensing applications run more than 10,000 pages, and annual reactor operating fees can cost millions of dollars.

If We Let Them

Sanders’ demand for a federal moratorium on data centers is radical. It’s also unpopular. Many liberal Democrats declined to endorse it. When asked by Politico about Sanders’ moratorium proposal, even Rep. Jasmine Crockett (D–Texas) said artificial intelligence could ” bring real economic opportunity to Texas.”

Voters themselves aren’t too agitated about data centers one way or another. When polled by Navigator Research in December 2025, 58 percent of respondents said they’d heard either a little or nothing about new data center construction. Most respondents did not have clear views on whether data centers are good or bad.

Even so, a data center moratorium wouldn’t be an entirely unprecedented intrusion of federal regulation into local land-use decisions.

The federal 1996 Telecommunications Act preempted local governments’ ability to use their zoning codes to block cellphone towers. Without that top-down deregulation, much of the modern information economy probably would not exist today.

Sanders’ call for a nationwide moratorium does make one wonder if we still have it in us to keep government out of the way of new and dynamic industries.

Peter Thiel’s famous quip that we wanted flying cars but got 140 characters—a reference to Twitter’s original character limit per post—was meant to highlight how the tech economy had exploded while our ability to improve the physical world had stagnated. The fight over data centers illustrates that this is a false dichotomy. We need computing equipment for AI to detect cancer early, write your emails, decode scrolls charred by Mount Vesuvius, help you plan dinner, guide driverless cars, and so much more. That computing equipment needs to live somewhere. That somewhere is data centers.

Contra their critics, big boxy buildings with servers inside produce vanishingly few externalities. Their water consumption is modest, their emissions are negligible, and their aesthetic merits are a matter of personal taste. Their power demands are considerable, but also eminently serviceable.

On the other side of the ledger, their economic benefits are hard to overstate. A bright future of AI-supported innovations in safety, health, and convenience is only realizable if we get out of the way of data center construction.

The question isn’t whether data centers are a good thing. It’s whether we’ll let them be good at scale.

We don’t know the answer to that question. There are certainly powerful forces arrayed behind data centers’ steady expansion.

Large tech companies obviously view them as profitable investments. The Trump administration’s interest in winning the “AI race” with China has seen it issue a number of executive orders that attempt to ease the regulatory burden on the industry. But tech is not particularly popular on the left or the right. People’s irritation at their own smartphone use has snowballed into a wider bipartisan assault on the industry. In this environment, overheated worries about data centers’ water and power bills are just another excuse to beat up on the Mark Zuckerbergs and Jeff Bezoses of the world.

The same zoning laws that dictate whether homeowners can add a pool house or open a home-based business also give the neighbors an effective veto over new data centers and the techno-optimist future they power. When it comes to questions of who’s allowed to build what and where, the real enemies of the future are often local planning boards who don’t trouble themselves with global concerns.

It needn’t be this way. Data centers will do good things for us, if only we’d let them.

The post The Joys of Data Centers: Debunking the Backlash Against the $7 Trillion AI Building Boom appeared first on Reason.com.