AI and data center services are highly tradable. Unlike physical infrastructure that must be consumed locally, the output can be generated anywhere and accessed globally. A high-regulation jurisdiction can reject hosting data centers, and its citizens will still benefit from AI capabilities—unless they're serving the EU market, where Brussels' regulatory framework forces a physical footprint. For everyone else, workloads simply relocate to Texas, Ireland, or Singapore, and domestic consumers barely notice.
A decade from now, expect a wave of populist-left governments to decry how we "outsourced" the AI revolution to China or Australia—conveniently ignoring that they were the ones who made local hosting politically untenable through zoning restrictions and environmental hearings.
That said, the larger concern is autonomous vehicles. We're expending regulatory capital on static compute when dynamic, physical automation has a far stronger claim to near-term economic productivity. AVs could reshape logistics, labor markets, and urban economics within five years in ways LLMs cannot yet match. If we regulate robo-taxis into oblivion before they scale, we forgo a tangible economic multiplier—and unlike AI, you can't offshore a truck route.
Could it stem from the history of factory towns and their later abandonment e.g. suspicion toward company incentives? Also country people tend to be low in trait Openness, so they really prefer something be long-standing and well-established before trusting in its goodness and stability (stability is of special importance to them). A belief in water/electrical grid strain might just be a stand-in for this general anxiety - making it stronger than the object-level belief itself.
I think very distantly. Hacks like More Perfect Union basically dug up their old anti- factory and anti- mine agitprop, changed nothing, and then laundered it to the public. That's why, talking to normies, they'll bring up pollution (?!), which is literally just not an issue for datacenters
You left out or greatly discounted several negative externalities of data centers
1. Diminished property values— If an existing residence is within eyesight or earshot of a data center its value will go down
2. Peace and quiet wrecked--People who live out in the country often regard the serenity and silence of the environment a great part of the value of living there. An ugly noisy data center means you're no longer living in the environment you chose. You'd better hope that you can Leave your quiet home behind with a minimum of financial loss. Perhaps, rhe data center owner would be willing to buy houses that people no longer wish to inhabit
3. Esthetics are important to many people— But you consider them a mere irrationality.
4, Noise: Of course, the amount of noise generated by a data center depends on many factors, but it can be but it often considerable. Constant noise degrades human health. Google AI summarized the issue in this manner:
"A data center's continuous low-frequency hum is typically audible from half a mile to 1 mile away under normal conditions. Under specific atmospheric conditions or with very large facilities, the low-frequency noise can travel 2 to 3 miles, and rare reports note faint sounds up to 5 to 10 miles away."
"Opposition to building is driven by a mixture of disdain for unsightly builds, demands to preserve community and environmental character, pure resent, and so on. These are generally not rationally motivated concerns, which helps to explain why fact-based messaging doesn’t work that well:"
I've come to understand that you had no intention of addressing all the negative externalities of data centers.
Your focus seems to be limited determining whether people can be induced to see the broader economic good or bribed into accepting the negative externalities.
I simply did not realize your goal was that limited.
Your link to "the Ottoman Empire" was not very informative. A better source would be some article about Ottoman policy towards new technology, as specific as you can find. Also, I'm surprised you didn't mention Nixon's failed nuclear-energy initiative
I considered it well-known that the Ottomans suppressed tons of good innovations like the printing press, and that their Westernization efforts were tepid. I could perhaps write about that later.
Nixon's nuclear initiative is actually not strongly supported as something he wanted to do.
I thought it varied by time a lot. Like they definitely suppressed the Printing Press, but not much later than Katherine the Great forbid (non-state) printing press operation. By the time of railroads and such, I had thought they were trying, if not very well. I do *not* remember the details very well, however.
The data you gave to support your claim of people "not believing the truth" is extremely misleading. The survey you gave as evidence only asks whether or not the arguments given are "convincing" or not. People could very well believe your arguments don't provide a convincing reason to offset the negatives of a data center.
Also, the article clearly mentions providing facts specifically about *the most common counter-arguments*. So, people would think that "water" matters, they'd face the truth, and then not be convinced.
I did not see a single counterargument against "generate constant disruptive noise" and the arguments against the strain on the power grid were also very limited as well as lowering property values. If the only thing you can counter is the water argument then someone may not be convinced entirely as their points were not addressed.
You're implicitly arguing that people fall back to other arguments, and they do so in a way that's 'just so' to not fit with their reported levels of credence in other issues with data centers, given the observed levels of overall support/disdain. I don't like arguing to unlikely coincidence; it's unserious.
Electricity is only rarely an issue, usually the grid can handle it, though I'm honestly pretty ok with forcing datacenters to add grid capacity, and so is Cremieux if you read the full article.
Both of these issues were addressed in the article and the study.
Given public opinion, I was surprised the noise thing wasn't mentioned, but it's not really a serious critique: it's loud within, like 100m max. I think only one data center actually inconvenienced people this way, which was very stupid, but could probably be managed with lawsuits under existing ordinances, followed by like 2 sound-absorbing baffles. Didn't stop every news crew across the country from posing dramatically in front of it, though
It's astonishing that Amazon distribution centers poll so well. I would think they poll terribly given the road traffic they must generate.
I'd wager it's because people connect it to something they love, the convenience of speedy deliveries.
Facts don't work, bribery (implicit and explicit) hopefully will, as you noted.
AI and data center services are highly tradable. Unlike physical infrastructure that must be consumed locally, the output can be generated anywhere and accessed globally. A high-regulation jurisdiction can reject hosting data centers, and its citizens will still benefit from AI capabilities—unless they're serving the EU market, where Brussels' regulatory framework forces a physical footprint. For everyone else, workloads simply relocate to Texas, Ireland, or Singapore, and domestic consumers barely notice.
A decade from now, expect a wave of populist-left governments to decry how we "outsourced" the AI revolution to China or Australia—conveniently ignoring that they were the ones who made local hosting politically untenable through zoning restrictions and environmental hearings.
That said, the larger concern is autonomous vehicles. We're expending regulatory capital on static compute when dynamic, physical automation has a far stronger claim to near-term economic productivity. AVs could reshape logistics, labor markets, and urban economics within five years in ways LLMs cannot yet match. If we regulate robo-taxis into oblivion before they scale, we forgo a tangible economic multiplier—and unlike AI, you can't offshore a truck route.
Could it stem from the history of factory towns and their later abandonment e.g. suspicion toward company incentives? Also country people tend to be low in trait Openness, so they really prefer something be long-standing and well-established before trusting in its goodness and stability (stability is of special importance to them). A belief in water/electrical grid strain might just be a stand-in for this general anxiety - making it stronger than the object-level belief itself.
Look at the support for other things in the first chart for an answer to your history question.
I think very distantly. Hacks like More Perfect Union basically dug up their old anti- factory and anti- mine agitprop, changed nothing, and then laundered it to the public. That's why, talking to normies, they'll bring up pollution (?!), which is literally just not an issue for datacenters
You left out or greatly discounted several negative externalities of data centers
1. Diminished property values— If an existing residence is within eyesight or earshot of a data center its value will go down
2. Peace and quiet wrecked--People who live out in the country often regard the serenity and silence of the environment a great part of the value of living there. An ugly noisy data center means you're no longer living in the environment you chose. You'd better hope that you can Leave your quiet home behind with a minimum of financial loss. Perhaps, rhe data center owner would be willing to buy houses that people no longer wish to inhabit
3. Esthetics are important to many people— But you consider them a mere irrationality.
4, Noise: Of course, the amount of noise generated by a data center depends on many factors, but it can be but it often considerable. Constant noise degrades human health. Google AI summarized the issue in this manner:
"A data center's continuous low-frequency hum is typically audible from half a mile to 1 mile away under normal conditions. Under specific atmospheric conditions or with very large facilities, the low-frequency noise can travel 2 to 3 miles, and rare reports note faint sounds up to 5 to 10 miles away."
Noise and property values are both mentioned in the second chart.
I also linked to my prior post that was partially about the importance of aesthetic concerns for getting things built!
So three issues are not addressed at all, and one is supposedly addressed elsewhere.
There are many links in the article—which one are you referring to? I skimmed the article and I could not find it
All four of your concerns were addressed. Did you actually just not read the article or the other person who replied to you?
"Opposition to building is driven by a mixture of disdain for unsightly builds, demands to preserve community and environmental character, pure resent, and so on. These are generally not rationally motivated concerns, which helps to explain why fact-based messaging doesn’t work that well:"
I've come to understand that you had no intention of addressing all the negative externalities of data centers.
Your focus seems to be limited determining whether people can be induced to see the broader economic good or bribed into accepting the negative externalities.
I simply did not realize your goal was that limited.
Thank you for confirming you did not read the article, or at the very least, did not read it properly.
What about the other three issues?
If you had made an explicit note about that link to esthetic issues in the text, it would have been helpful.
I would never expect my readers to spontaneously research my views outside the essay at hand, nor expect them to intuit the utility of a link.
Cremieux did not address those issues in his analysis. It is his analysis that interests me
Your link to "the Ottoman Empire" was not very informative. A better source would be some article about Ottoman policy towards new technology, as specific as you can find. Also, I'm surprised you didn't mention Nixon's failed nuclear-energy initiative
I considered it well-known that the Ottomans suppressed tons of good innovations like the printing press, and that their Westernization efforts were tepid. I could perhaps write about that later.
Nixon's nuclear initiative is actually not strongly supported as something he wanted to do.
I thought it varied by time a lot. Like they definitely suppressed the Printing Press, but not much later than Katherine the Great forbid (non-state) printing press operation. By the time of railroads and such, I had thought they were trying, if not very well. I do *not* remember the details very well, however.
The data you gave to support your claim of people "not believing the truth" is extremely misleading. The survey you gave as evidence only asks whether or not the arguments given are "convincing" or not. People could very well believe your arguments don't provide a convincing reason to offset the negatives of a data center.
What negatives??
Also, the article clearly mentions providing facts specifically about *the most common counter-arguments*. So, people would think that "water" matters, they'd face the truth, and then not be convinced.
I did not see a single counterargument against "generate constant disruptive noise" and the arguments against the strain on the power grid were also very limited as well as lowering property values. If the only thing you can counter is the water argument then someone may not be convinced entirely as their points were not addressed.
You're implicitly arguing that people fall back to other arguments, and they do so in a way that's 'just so' to not fit with their reported levels of credence in other issues with data centers, given the observed levels of overall support/disdain. I don't like arguing to unlikely coincidence; it's unserious.
"Property values" go up.
Electricity is only rarely an issue, usually the grid can handle it, though I'm honestly pretty ok with forcing datacenters to add grid capacity, and so is Cremieux if you read the full article.
Both of these issues were addressed in the article and the study.
I wrote a longer comment about "noise".
Given public opinion, I was surprised the noise thing wasn't mentioned, but it's not really a serious critique: it's loud within, like 100m max. I think only one data center actually inconvenienced people this way, which was very stupid, but could probably be managed with lawsuits under existing ordinances, followed by like 2 sound-absorbing baffles. Didn't stop every news crew across the country from posing dramatically in front of it, though