Ryan Horne

Trying to live a decent life in an increasingly strange world. 🍊🌴🐊Florida is home. ☦︎

[Ryan Horne @RyanHornex]

@MediaShado75352 @TheAdamWaite Re-read the text, not just builders but also workers who operate in post offices construction are also foreign born.

[Ryan Horne @RyanHornex]

@Robotbeat @TheAdamWaite What do you mean based on absolutely nothing?

[Ryan Horne @RyanHornex]

@ExergyDev Extreme fear mongering*

[Ryan Horne @RyanHornex]

My argument concerns the incremental infrastructure costs associated with unusually large new loads and how utilities allocate those costs across customer classes.

https://t.co/6gPrNo7mEg

This 2026 MIT CEEPR study estimates data-center entry from 2010–2024 increased residential retail electricity prices by about 2.1%

More importantly, regulators themselves have recognized this and changed accordingly. Wisconsin’s Public Service Commission recently required very-large-load customers to pay 100% of certain costs and specifically changed its tariff to prevent transmission costs from shifting onto existing customers.

https://t.co/a1iWnztIdh

[Ryan Horne @RyanHornex]

I won’t necessarily deny that there’s extreme fear monger if outside of what the data shows. That doesn’t deny that the data still shows what’s I’ve stated.

Beyond that One you’ve still conceded your original point entirely. And again the rates are just an average meaning lower and higher rates exist it’s not the universal cap.

Appreciate the concession though have a good evening.

[Ryan Horne @RyanHornex]

You started at “there’s no evidence to back up data centers as the cause of higher rates.” Now that I’ve given you newer causal research finding a single hyperscale facility raises household electricity prices ~4%, and concentrated development raises them ~13%, the argument has become “well, that’s not that much and 10 colocated facilities aren’t common.”

That’s a completely different claim.

And the colocation correction actually reinforces my original point. I was specifically talking about small communities absorbing concentrated hyperscale development where existing infrastructure is inadequate. If the effect becomes substantially larger as data centers cluster within the same utility territory, that is precisely the cost-socialization problem I was talking about.

You can argue 4% or 13% is an acceptable price to pay for the economic benefits. Fine. But you’ve moved a long way from “there’s no evidence” and “QED, the increases must be fabricated.”

[Ryan Horne @RyanHornex]

@FFjunkie_ @TheAdamWaite Yeah, it’s apart of the problem. If the data center itself is meant to benefit the local community if 30%+ of the workforce building and operating it is foreign born it provides almost no benefit to local residents. Leftists used to oppose immigration on this exact point.

[Ryan Horne @RyanHornex]

Now having received your source and also being familiar with it, your own source doesn’t establish your QED. It estimates an average effect from 2015–2024 and explicitly cautions that future supply constraints could reverse that effect. Its proposed mechanism depends on incremental demand being accommodated without binding capacity constraints. That’s a pretty important qualification when we’re discussing the current hyperscale AI buildout and communities being asked to construct enormous amounts of new generation and grid infrastructure.

But since you asked for newer research, here you go.

https://t.co/44bsRDJySV

Scalera, Bosetti and Pecci, published through CEPR five days ago, specifically examine the post-2021 AI-associated data-center buildout. They estimate one hyperscale facility raises household electricity prices about 4%, while a one-standard-deviation increase of ten data centers raises them about 13%, with households disproportionately bearing the burden.

So no, you can’t take one 2015–2024 national estimate, ignore its own stated limitations, declare every larger contemporary rate increase fabricated or unrelated, and slap “QED” on the end. We now have competing causal estimates, and the newer study examining the AI-era expansion finds exactly the cost-shifting effect you’re claiming there’s “no evidence” for.

[Ryan Horne @RyanHornex]

That doesn’t follow at all. The 2.1% figure is an average estimated residential price effect across the study’s entire 2010–2024 sample, not a ceiling on localized rate impacts. The same study finds substantially different effects depending on utility ownership and regulatory structure. More recent research estimates considerably larger localized effects from hyperscale facilities. So you can’t take a national average treatment effect, treat it as the maximum possible local effect, and then conclude every larger observed increase must either be fabricated or unrelated. QED requires the conclusion to actually follow from the premises.

[Ryan Horne @RyanHornex]

@RyanHornex Nobody had a problem before that utility rates were usually lower for commercial vs residential. I keep seeing this line touted but no evidence to back up data centers as the cause of higher rates.

[Ryan Horne @RyanHornex]

@JayDyer Insanity.

[Ryan Horne @RyanHornex]

@ExergyDev The question is who bears the cost and risk of building the infrastructure necessary to serve that customer?

[Ryan Horne @RyanHornex]

But the counter argument still exists around the socialization of the cost of infrastructure absent any kind of guarantee to the public.

If these projects are going to be built infrastructure upgrade cost shouldn’t fall to the taxpayers who largely have little say in these projects but instead should be incurred by the private firms building them.

[Ryan Horne @RyanHornex]

Just gonna skip past the “you don’t understand data centers” point because it’s a nonstarter.

I understand that the national and geostrategic threat posed by the AI “arms race” is very real. However, that doesn’t logically support the subsequent claim that opposition to localized data center development is silly. Why should small, often rural communities be expected to socialize the infrastructure costs required for these data centers without particular guarantees from private firms regarding their long-term presence or support for those communities?

I’ll grant that the line “select corporations and tech bros” is rhetorically loaded, but there’s substantive underlying argumentation worth considering. There’s no denying that the current AI market is extremely concentrated among a small handful of enormous companies. Brookings estimates Alphabet, Amazon, Meta, and Microsoft are planning around $650 billion in capital expenditures this year alone. It still falls back on the initial question. I can even entertain the question of whether a different strategy, similar to China’s emphasis on efficiency, adoption, and physical-world integration alongside raw compute capacity, is worth considering.

As for the Warren/Sanders point, calling opposition to corporate subsidies a Bernie Sanders talking point is a strange argument coming from the right, especially. If these projects are as economically indispensable and profitable as you’re telling me, why exactly do some of the wealthiest corporations in human history need taxpayers to subsidize them? Again, this goes directly back to the original question of socializing costs while privatizing the returns.

I don’t think the argument, although perhaps being expressed less elegantly, is one of categorical opposition to data centers. The argument is that we should build infrastructure where its total benefits exceed its total costs and require the primary beneficiaries to internalize those costs rather than shifting them onto local communities.

Disregarding those concerns isn’t some inherent conservative win like you’re making it out to be.

[Ryan Horne @RyanHornex]

@realmfoster That’s just blatantly untrue, economic sustainability is almost always a factor in any given development whether it’s commercial, industrial, or residential. It’s disingenuous to suggest otherwise.

[Ryan Horne @RyanHornex]

@TheRealBeliveau @connorpklocke Such a silly strawman.

[Ryan Horne @RyanHornex]

The quiet part is just how dependent our modern economy is on abundant, affordable energy.

U.S. crude inventories, including the Strategic Petroleum Reserve, are now sitting near 40–50 year lows measured in days of supply. That leaves considerably less cushion against a major supply disruption or geopolitical shock which is ever present with the ongoing Iranian conflict.

Rebuilding that buffer needs to be an energy-security priority.

[Barchart @Barchart]

U.S. has just 41 days of crude oil inventory left, the lowest level in half a century 🤯 👀

[Ryan Horne @RyanHornex]

@TheAKGuy You are the future of the GOP! Keep up the great work.

[Brandon Herrera @TheAKGuy]

Turns out Texas isn’t willing to embrace a liar who wants to trans children, return to catch and release, and doesn’t even know how many counties are in their district.

We raised more in 45 days than she has her whole career, and before it’s even spent, we’re up by 6 points 🇺🇸

[Ryan Horne @RyanHornex]

@brianeskow Yes, having to operate and function with other libertarians being the biggest downside.